key: cord-287233-srkny5v4 authors: yu, hai-ping; ma, li-li; hung, yun-ying; wang, xue-bin; peng, you-qing; chen, chi; zhuang, hui-ren title: application of ‘mobile hospital’ against 2019-ncov in china date: 2020-04-24 journal: epidemiol infect doi: 10.1017/s0950268820000862 sha: doc_id: 287233 cord_uid: srkny5v4 nan the world health organization (who) has declared the outbreak of the novel coronavirus (2019-ncov) a public health emergency of international concern [1] , which poses a great challenge to china's health system. as of 28 february 2020, the total number of confirmed covid-19 patients had reached 78 961, with 2791 deaths [2] . to effectively control the spread of the epidemic and provide medical services, the government directed general hospitals to rapidly open fever clinics. more than 110 fever clinics are required to provide services to the public in shanghai. however, due to the sudden outbreak of the epidemic and the arrival of the spring festival, most hospitals have been unable to cope with the consequences of the 2019-ncov outbreak. a mobile hospital is a kind of temporary medical institution that can be rapidly deployed [3] , and is often used as a medical resource to rapidly implement treatment in case of an emergency. our hospital has an international emergency medical team, which is certified by the who and equipped with perfect mobile hospital facilities. in this anti-epidemic battle, our hospital adopted the mobile hospital to deal with patients in the early stage of the outbreak in shanghai and as a supplement medical facility to assist the wuhan keting medical shelter. here, we describe the lessons learned during these deployments. as a temporary facility in the early stage of the outbreak on 23 january, we completed the layout of the mobile hospital fever clinic in just 4 h, which not only effectively deals with fever patients, but also saves precious time for our hospital to complete the formal reconstruction of its fever clinic. from 23 to 27 january, 367 fever patients were registered in the mobile hospital, including one confirmed patient, without any medical staff being infected. our experiences are described below. we established a management team, including experts in infection control management, medical and nursing administration, it management, pharmacy, laboratory testing, material procurement and logistics management. the anti-epidemic management team was led by a hospital leader, and everyone performed their duties and cooperated sincerely ( table 1) . the layout of the mobile hospital was completed according to the requirements of infectious diseases, and consisted of 12 folding tents. each tent covers an area of 48 m 2 with separate negative pressure air purification systems. the following areas were set up: triage and waiting room, consultation room, laboratory, pharmacy, treatment room (equipped with six infusion chairs), two isolation wards (equipped with beds and mobile toilets) and rest and dressing rooms for medical staff (figs 1 and 2 ). all staff members were required to wear personal protective equipment. each shift was staffed by two doctors, two nurses, one pharmacist, two laboratory technicians, one cleaner and one security guard. all staff in the mobile hospital were provided with level ii protective equipment (protective clothing, goggles, n95 masks and gloves), and the laboratory technician was provided with level iii protective equipment (as for level ii, except that the goggles are replaced with a face mask). strictly enforced temperature measurements were implemented. to prevent cross-infection and ensure the accuracy of temperature measurements, the nurses used ear temperature guns to measure the temperature of patients, and used disposable earmuffs. if necessary, mercury thermometers were used for accurate thermometry. the patients were required to provide complete personal information and were questioned about their epidemiological history, such as: have you been to an epidemic area within the last 14 days? have you contacted people from the epidemic area? patients also signed a letter of commitment for the information they provided. twenty waiting chairs were set up, 1 m apart. all patients and caregivers were required to wear masks, and the hospital provided masks free of charge to those without masks. the patients were screened by using nasopharyngeal swab, blood sampling and lung computerised tomography (ct) scans. anti-2019-ncov experts in the hospital consulted about the suspected patients, and these patients were reported to the district health supervision institute. the mobile hospital was equipped to carry out samplings that were then sent for nucleic acid detection in the center for disease control and prevention. while awaiting the report, the patient received relevant treatment in the isolation room, and the hospital provided all daily necessities. then, the confirmed patients were transferred to infectious disease hospitals using a special ambulance. this outbreak occurred suddenly and during the chinese spring festival, and few epidemic prevention materials were available. in the cleaning area of the mobile hospital, five cabinets were used for storing protective materials. the nurse manager was responsible and accountable for the management of these protective materials. given the flexibility of the mobile hospital, we were able to quickly and effectively set up three areas and two 'channels'. polluted areas, semi-polluted areas and a clean area were well divided, and two separate channels were built for patients and medical staff (fig. 3) . a quick hand disinfectant facility was set up. coronaviruses are sensitive to alcohol disinfectants, and alcohol hand disinfectants were placed on all tables and next to registration machines. two ultraviolet disinfection lamps were fixed for air disinfection. indoor articles were wiped with 2000 mg/l chlorine, and the ground was mopped with 1000 mg/l chlorine, every 2 h. the vomitus excreta of the patients was treated as class i infectious disease excreta. the excreta were disposed of in the mobile toilet. two layers of infectious storage bags were placed in the pedestal pan in advance. the excreta were pretreated with 50 000 mg/l effective chlorine. after pouring the effective chlorine, the mouth of the bags was sealed, and the bags disposed of using the delivery box after 4 h. the bed articles used by the patients were placed directly into the double-layer infectious storage bags, and chlorine (10 000 mg/l) poured into the bags. double bags and double seals, which were marked, were handed over to the washing unit for treatment. the final disinfection of the confirmed patient's room was done using spray disinfection and a hydrogen peroxide machine. given the inadequate airtightness of the tent, the room was sealed with sealing strips before disinfection. the total amount and time needed to disinfect a room was calculated based on the room volume. for a tent with an area of 48 m 2 , 60 min of disinfection was required, followed by 60 min of ventilation. on 4 february, our international emergency medical team was ordered to assist the wuhan keting medical shelter hospital. the medical shelter was originally reconstructed from the wuhan culture and art center, with a total of 1461 beds. as of 17:00 on 21 february, it had received 1701 patients with mild disease. the clinical manifestations of the patients with mild disease were slight, or they had a fever, respiratory tract problems and other symptoms. the imaging findings showed ground glass, no dyspnoea and chest distress. arbidol hydrochloride tablets combined with chinese oral medicine was the main treatment method. if the patient's condition worsened beyond the capacity of the shelter hospital, staff of the headquarter were responsible for contacting the designated hospital and arranging 120 special vehicles for medical staff to transport patients to the designated hospital. (the designated hospital is the comprehensive hospital closest to the fangcang, for example, the designated hospital for keting fangcang is jin yin tan hospital, 1 km away.) there was 1043 medical staff staying in this shelter, including 151 doctors, 841 nurses, 11 imagers, five pharmacists, two infection control experts and 33 laboratory technicians. given that the shelter was reconstructed from an art centre, the layout of the medical facilities was imperfect. as soon as we arrived, our rescue team built a mobile hospital using 25 tents, including outpatient services, a pre-inspection space, observation rooms, monitoring rooms, a pharmacy, offices, meeting rooms, dressing rooms, restrooms, a canteen and bathrooms and toilets. seven tents were used as the medical channels of the shelter hospital, which greatly contributed to the smooth operation of the hospital (figs 4-6) . however, we also encountered many problems during implementation of the mobile hospital, including: how to supply water and power, how to transfer patients to receive ct examination (because it was impossible to perform ct scans in the mobile hospital) and how to transport patients' specimens and excretory wastes. to solve these problems, the following measures were taken: we chose to build the mobile hospital at a certain distance away from the main building, but close enough to connect a water source and power supply; we set up a special ct unit in the emergency department reserved only for those patients with fever, and also set up a special channel for patient transport to and from this ct unit with transfer wheelchairs (fig. 7) and made use of special specimen-transferring boxes and sealeddischarge-transferring boxes (figs 8 and 9 ). mobile hospitals have some shortcomings that need to be further improved. first, due to the thin material of the tent, the thermal insulation is poor, and given that air conditioning cannot be used (because this would assist in spreading the virus), the medical staff reported that they felt cold working in the tent. also, there are certain security risks associated with the weather. the mobile hospital in shanghai encountered 3 days of rain, and water needed to be cleared and anti-skid matting used. in wuhan, there was heavy snow (fig. 10) , which needed to be cleared. how mobile hospitals can better cope with bad weather should be addressed in future work. in conclusion, mobile hospitals, which are characterised by flexibility, have played a critical role in this anti-epidemic campaign. these mobile hospitals have been rapidly put into use, thereby alleviating the shortage of medical resources in the early stage of the anti-epidemic campaign. available at https:// www.who.int/news-room/detail/30-01-2020-statement-on-the-second-meetingof-the-international-health-regulations-(2005)-emergency-committeeregarding-the-outbreak who. novel coronavirus (2019-ncov)situation report -20 use of an innovative design mobile hospital in the medical response to hurricane katrina cleaning up snow in wuhan acknowledgements. the authors wish to thank the anonymous reviewers, associate editors and editors for their thoughtful reviews and suggestions in this paper. conflicts of interest. the authors declare no conflicts of interest. key: cord-011663-3ggah1y1 authors: haider, najmul; yavlinsky, alexei; kock, richard title: response to ‘evaluation of modelling study shows limits of covid-19 importing risk simulations in sub-saharan africa’ (epidemiology and infection – hyg-le-10513-may-20) date: 2020-06-10 journal: epidemiol infect doi: 10.1017/s0950268820001211 sha: doc_id: 11663 cord_uid: 3ggah1y1 nan response to 'evaluation of modelling study shows limits of covid-19 importing risk simulations in sub-saharan africa' (epidemiology and infection -hyg-le-10513-may-20) najmul [2] . in the letter, the authors state that they obtained 2417 covid-19 cases reported by 40 countries in sub-saharan africa within the 30 days of the first case confirmed in nigeria on 27 february. of the 442 cases with international travel history, only one had travelled to china. we are encouraged by this finding and believe that it validates our modelling approach. the authors also point out that the model did not consider the risk of importing covid-19 cases from other countries. we would like to point out that we submitted the final version of our manuscript to epidemiology and infection on 7 february 2020. at the time, virtually no instances of community transmission were being reported outside of china and thus there was no data available to reliably calculate the risk of case importation from other countries (please see who's situation report-18 on novel coronavirus (2019-ncov): https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200207sitrep-18-ncov.pdf?sfvrsn=fa644293_2). it was also our hope that the public health measures that were being implemented in countries that were at high risk of importing covid-19 cases would be sufficient to prevent further international spread of this disease, which unfortunately did not come to pass. at the time of manuscript submission, the possibility of pre/asymptomatic covid-19 transmission was still a matter of debate within the scientific community, see, e.g. https://www.sciencemag.org/news/ 2020/02/paper-non-symptomatic-patient-transmitting-coronavirus-wrong. however, we accept that the pre/asymptomatic transmission aspect of covid-19 may have played a significant role in the collective failure to halt its spread and prevent it from becoming a global pandemic. evaluation of modelling study shows limits of covid-19 importing risk simulations in sub-saharan africa 2020) passengers' destinations from china: low risk of novel coronavirus (2019-ncov) transmission into africa and south america key: cord-332258-o3u52mhl authors: brlek, a.; vidovič, š.; vuzem, s.; turk, k.; simonović, z. title: possible indirect transmission of covid-19 at a squash court, slovenia, march 2020: case report date: 2020-06-19 journal: epidemiol infect doi: 10.1017/s0950268820001326 sha: doc_id: 332258 cord_uid: o3u52mhl since the beginning of the covid-19 epidemic, there is an ongoing debate and research regarding the possible ways of virus transmission. we conducted an epidemiological investigation which revealed a cluster of five covid-19 cases, linked to playing squash at a sports venue in maribor, slovenia. acquired data raises possibility that the transmission occurred indirectly through contaminated objects in changing room or squash hall or via aerosolisation in squash hall. transmission of covid-19 occurs mainly during close contact between people and through respiratory droplets produced during symptomatic phase, when an infected person coughs or sneezes [1] . however, there is ongoing research to identify other possible ways of transmission, such as transmission through faecal-oral route [2, 3] , indirect transmission through contaminated common objects or with virus aerosolisation [4, 5] . there is also growing evidence of virus spread by asymptomatic carriers [6] [7] [8] [9] . the aim of this report is to present a cluster of cases where transmission seems to have occurred indirectly through contaminated objects or aerosol. epidemiological investigation revealed a cluster of five covid-19 cases linked to playing squash at a sports venue in maribor, slovenia. index patient (person a) was travelling in italy from 29 february to 2 march, where he most likely acquired the infection. he developed symptoms of the disease (tiredness and fatigue) on 4 march during a game of squash. later epidemiological investigations linked four other cases of covid-19 to the same squash hall. person a was playing squash with person b. they arrived at the sports venue a few minutes before 17:30. they changed their clothes in a dressing room, but they were not able to remember which dressing room they used. they began their match at 17:30 in a squash hall number one and played until 18:30, when they returned to the dressing room. person b changed clothes and left soon after that. person a showered and left around 18:45, but no later than 19:00. they were alone in dressing room. person a did not report meeting anyone on his way out. person a developed fever (>38°c) at home later that day. he sought medical help on 5 march, when a nasopharyngeal swab was taken and reverse transcription polymerase chain reaction (rt-pcr) confirmed covid-19 infection. we identified person b as a close contact and provided him with instructions about self-monitoring. he developed symptoms (headache, rhinorrhoea and fever) on 10 march rt-pcr confirmed the infection on 14 march. person c and person d also played squash at the same sports venue on 4 march. they arrived at 19:10, changed their clothes in dressing room number three, and began their match at 19:15 in hall number 1. they played until 20:00, then rested and talked with persons e and f in the hallway. after that they returned to dressing room, where person c took a shower. they both left at 20:30. person c developed symptoms on 7 march and tested positive on 11 march. person d developed symptoms on 8 march and tested positive on 12 march. persons e and f arrived at 19:50, changed their clothes in dressing room number three, talked with persons c and d in the hallway in front of hall number 1 and began their match at 20:00. they played until 20:45, returned to the dressing room, where person e took a shower. they left at 21:00. person e developed symptoms on 8 march and tested positive on 14 march. person f also developed mild symptoms. clinician instructed him to selfisolate, but did not indicate diagnostics and laboratory confirmation (fig. 1) . the time difference between the first pair (a and b) exiting the venue and the second pair arriving (c and d) was at least 10 min, more likely around 20 min. in addition, the second pair began their match in hall 1 approximately 45 min after the first pair ended. person e arrived an hour after person a left the sports venue, and began to play in hall 1 approximately 90 min after the first pair finished. unfortunately, we could not confirm if all three pairs changed in the same dressing room, but all of them reported playing in the same squash hall. none of them shared any sport equipment and had no contact with the receptionist or any other employee at the sports venue. there are five employees at the venue and none of them were symptomatic at the time, neither did they develop any symptoms later. one of them got tested, but rt-pcr came back negative. the same venue also offers different types of sport activities (basketball, soccer, fitness and badminton), but we did not identify cases linked to any other sport activity at this venue. our epidemiological investigation raises possibility that secondary cases in the cluster got the infection through indirect transmission at the sports venue. the only plausible source of infection we were able to identify was player a. all secondary cases developed first symptoms within the incubation period (3-6 days for presented cases). none of the secondary cases had any other epidemiological link to other confirmed cases or symptomatic persons outside of squash hall. outside the squash hall, none of them reported any other contact with one another after that day. we cannot exclude the possibility of transmission through an unconfirmed or asymptomatic employee that works at the sports venue, because rt-pcr was not performed on all employees. however, we reduced this possibility with additional questioning of the patients and epidemiological investigation at the venue. in the time of identifying and managing this cluster, the epidemic in slovenia had just started and there was no community spread. slovenia began testing for covid-19 infection as early as 27 january and the first case was confirmed on 4 march. on 8 march (2 days before last players, cases d and e, developed symptoms), there were only 19 confirmed cases in the country and only one other confirmed case (imported) in maribor region. until 12 march, we did not detect community spread in slovenia, since all cases were either imported or epidemiologically linked to other confirmed cases. we identified the first case in maribor region without a known source of infection as late as 20 march. this shows that it is unlikely that the persons identified in the cluster were infected from different sources. however, we cannot exclude this possibility completely, because undetected community spread could have existed before those dates. after thorough questioning of everyone involved, we concluded that persons c and d had no direct contact with the index patient (person a), who had been mildly symptomatic at the time of playing squash at the sports venue. the last pair of players (persons e and f) also denied any contact with the index patient. because of the time difference between the game of first and last pair of players, it is highly unlikely they had any direct contact. persons c, d and e reported direct contact with each other, but they all developed symptoms on 7 or 8 march. because of very similar date of onset, a common source (person a) seems more plausible than transmission from person c to e and d or vice versa. we concluded that the mode of transmission between the index patient and the secondary cases in this cluster was either through contaminated common objects or virus aerosol, since all three pairs shared the same squash hall, which is a small and confined space with poor ventilation, where strenuous physical activity is performed, during which shedding and aerosolisation of the virus could be increased [4, [10] [11] [12] . moist and warm atmosphere coupled with turbulent air flow generated by intense physical activity could extend the lifetime of virus-bearing droplets and eventually produce residues that may stay suspended in the air for hours [13, 14] . the time difference between first and third pair of players intrigued us; however, some early studies performed in experimental settings show that virus can remain viable and infectious in aerosols for several hours and up to days on surfaces [5] . this could explain our epidemiological findings. clusters with epidemiological evidence of indirect transmission in shopping malls [4] and dance classes [11] have been published, but we did not find reports where transmission occurred indirectly with reported time difference between the presence of an infectious person and their contacts in the same place. however, in our case, the situation is unique because it included a small number of people who visited the venue for precisely scheduled activity (time and place were very well known). additionally, activities were performed in specific circumstances (small room with poor ventilation in which players spend 45 min, intense physical activity, moist and warm atmosphere) at the very beginning of covid-19 epidemic. these factors could explain why no similar event has been reported until now. acquired data from our epidemiological investigation and scientific data that are available so far, raises possibility that transmission occurred indirectly through contaminated objects in the changing room or squash hall (doorknobs, clothes stands) or with virus aerosolisation in the squash hall. our findings advise caution in settings similar to observed. a familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster detection of sars-cov-2 in different types of clinical specimens evidence for gastrointestinal infection of sarscov-2 indirect virus transmission in cluster of covid-19 cases aerosol and surface stability of hcov-19 (sars-cov-2) compared to sars-cov-1 transmission of 2019-ncov infection from an asymptomatic contact in germany presumed asymptomatic carrier transmission of covid-19 estimating the asymptomatic proportion of coronavirus disease 2019 (covid-19) cases on board the diamond princess cruise ship sars-cov-2 transmission from presymptomatic meeting attendee airborne influenza a virus exposure in an elementary school recognition of aerosol transmission of infectious agents: a commentary infection risk in gyms during physical exercise turbulent gas clouds and respiratory pathogen emissions: potential implications for reducing transmission of covid-19 cluster of coronavirus disease associated with fitness dance classes, south korea. emerging infectious diseases 26 acknowledging the possibility of presented transmission is especially important when implementing less strict preventive measures and reopening gyms or other indoor sport facilities, wellness centres and night clubs.we emphasise the need for further research in modes of transmission in order to implement appropriate control measures to prevent nosocomial spread and superspreading events. the data that support the findings of this study are available on request from the corresponding author (z. simonović). the data are not publicly available due to restriction regarding participants' privacy. key: cord-290591-yi6yjjne authors: desai, angel n.; madoff, lawrence c. title: bending the epidemic curve: advancements and opportunities to reduce the threat of emerging pathogens date: 2019-04-03 journal: epidemiol infect doi: 10.1017/s095026881900058x sha: doc_id: 290591 cord_uid: yi6yjjne this invited editorial introduces a special issue of epidemiology & infection while also discussing advances in emerging infectious diseases. the world health organization (who) recently enumerated 10 threats to global health for 2019, notably emphasising ebola and other high-threat emerging pathogens as growing priorities [1] . although many of these diseases have the potential to cause public health emergencies, a lack of timely surveillance and effective interventions continue to hamper preparedness efforts [2] . moreover, the annualised financial impact of a global pandemic has been estimated to be as high as us$80 billion, severely burdening already constrained national budgets and healthcare systems [3] . outbreaks in resource-limited settings are often further complicated by conflict, fragile health systems, disruptions in healthcare delivery and socio-economic disparities [4] . these factors can make implementing appropriate outbreak prevention and control strategies difficult. to address these shortcomings, investments in surveillance to inform disease forecasting and ultimately effective prevention strategies are paramount to tackling the challenges posed by emerging pathogens. this edition of epidemiology & infection highlights new insights into many of the emerging infectious diseases mentioned in the who report, as well as re-emerging diseases that are gaining global prominence. they address a variety of diagnostic, therapeutic and epidemiological advances in outbreak preparedness. several papers review data on priority pathogens with a focus on resource-limited settings. for example, sikkema et al. present a systematic review on middle east respiratory syndrome coronavirus with the aim of characterising the distribution and spread of infection in dromedary camels [5] . aditi et al. review the recent nipah virus outbreaks in bangladesh and india, shedding light on transmission patterns of this emerging pathogen while also highlighting the importance of ongoing surveillance [6] . a review analysis of h5n1 and h9n2 by parvin et al. discusses genetic variations among avian influenza viruses circulating in bangladesh and the impact of accumulating mutations noted in poultry. lessons learned from the who response to the recent 2017 pneumonic plague outbreak in madagascar are presented by heitzinger et al., who highlight specifically the challenges of implementing rapid infection prevention and control measures in epidemic settings [7] . these studies also underscore the critical importance of the one health approach [8] . these, and indeed the great majority of emerging disease threats, are zoonotic and require us to consider other hosts and the environment in addressing them. outbreaks of re-emerging infectious diseases in high-income countries are also discussed, often implicating products imported across borders as well as trans-national spread due to as yet unknown causes. other papers discuss zoonotic and vector-borne disease epidemiology and present opportunities for predictive modelling. several diagnostic advances are also mentioned, elucidating changing epidemiological trends in recently recognised re-emerging pathogens. this issue of epidemiology & infection represents a diverse overview of current concerns surrounding emerging infectious diseases globally. all highlight the importance of supporting ongoing surveillance efforts as the cornerstone of disease prevention. early recognition of an outbreak allows control measures to be initiated in a timely way that can shift the epidemic curve, reducing its impact and possibly its geographic spread. enhanced surveillance measures with an emphasis on innovation, transparency and incorporation of the one health model are critical to epidemic preparedness measures in the future. it is also crucial to encourage research during outbreaks through rapid data sharing to facilitate rapid response efforts, as is promoted through organisations such as the international severe acute respiratory and emerging infection consortium (isaric) [9] . vaccination, if it can be implemented in time, can also bend the epidemic curve. promoting platforms for rapid vaccine development and deployment could provide a significant boost to outbreak control. the coalition for epidemic preparedness innovations (cepi), a public-private coalition that has been working to halt epidemics through the development of appropriate vaccines, is promoting both pathogenspecific and agnostic platform approaches [10] . these efforts deserve wide support and encouragement. in order for vaccines to be effectively employed, the growing threat of vaccine hesitancy worldwide must also be countered using methods grounded in the social sciences. governments, through their public health agencies and in coordination with efforts like the global health security agenda, must adopt preparedness plans and exercise them before outbreaks become major threats [11] . emerging and re-emerging infectious diseases will continue to present significant challenges in the coming years, and investing in novel methods for detection, prevention as well as therapeutics should remain priorities for the global public health community. angel n. desai, 0000-0001-8962-9427; lawrence madoff, 0000-0003-2589-7777 ten threats to global health in 2019 productive disruption: opportunities and challenges for innovation in infectious disease surveillance the inclusive cost of pandemic influenza risk refugees of the syrian civil war: impact on remerging infections, health services, and biosecurity in turkey global status of middle east respiratory syndrome coronavirus in dromedary camels: a systematic review nipah virus infection: a review using evidence to inform response to the 2017 plague outbreak in madagascar: a view from the who african regional office review analysis and impact of co-circulating h5n1 and h9n2 avian influenza viruses in bangladesh coalition for epidemic preparedness innovations international severe acute respiratory and emerging infection consortium acknowledgments. angel desai is supported in part by grant number t32 ai007433 from the national institute of allergy and infectious disease. the contents of this editorial are solely the responsibility of the authors and do not necessarily represent the official views of the nih. key: cord-277159-klhmed21 authors: bassal, r.; wax, m.; shirazi, r.; shohat, t.; cohen, d.; david, d.; abu-mouch, s.; abu-ghanem, y.; mendelson, e.; ben-ari, z.; mor, o. title: seroprevalence of hepatitis e virus in dromedary camels, bedouins, muslim arabs and jews in israel, 2009–2017 date: 2019-02-22 journal: epidemiol infect doi: 10.1017/s0950268819000062 sha: doc_id: 277159 cord_uid: klhmed21 hepatitis e virus (hev) is an emerging cause of viral hepatitis worldwide. recently, hev-7 has been shown to infect camels and humans. we studied hev seroprevalence in dromedary camels and among bedouins, arabs (muslims, none-bedouins) and jews and assessed factors associated with anti-hev seropositivity. serum samples from dromedary camels (n = 86) were used to determine camel anti-hev igg and hev rna positivity. human samples collected between 2009 and 2016 from >20 years old bedouins (n = 305), non-bedouin arabs (n = 320) and jews (n = 195), were randomly selected using an age-stratified sampling design. human hev igg levels were determined using wantai igg elisa assay. of the samples obtained from camels, 68.6% were anti-hev positive. among the human populations, bedouins and non-bedouin arabs had a significantly higher prevalence of hev antibodies (21.6% and 15.0%, respectively) compared with the jewish population (3.1%). seropositivity increased significantly with age in all human populations, reaching 47.6% and 34.8% among ⩾40 years old, in bedouins and non-bedouin arabs, respectively. the high seropositivity in camels and in ⩾40 years old bedouins and non-bedouin arabs suggests that hev is endemic in israel. the low hev seroprevalence in jews could be attributed to higher socio-economic status. hepatitis e virus (hev) is an emerging pathogen and one of the causes of viral hepatitis in the world [1] . the infection is mostly silent, but when symptoms do appear, illness is usually selflimiting [1] . hev infection can also become chronic in immunocompromised individuals, such as those receiving organ transplants or chemotherapy, as well as among individuals with hiv infection [2] . moreover, extra hepatic manifestations of hev infection, mostly neurological, are increasingly recognised [3] . currently, eight hev genotypes are known, all belonging to a single serotype [4] . of those, hev-7 and hev-8 have been identified in dromedary (1-humped) and bactrian (2-humped) camels, respectively [4, 5] . hev-7, identified in a liver transplant recipient had been linked to the consumption of camel products [6] . human hev infection, investigated using seroprevalence studies, was found to be more prevalent in older ages [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] , lower socio-economic status [21] , poorer residence areas [9, [14] [15] , among sheltered homeless adults [22] or uneducated people [14] , specific nationalities (for example, higher in mixed race donors and ethnic groups within china [12, 15] , or in immigrants from afghanistan [14] ), drinking water from wells or rivers [15] , consumption of meat products [7, 15, 17, 23] especially pork [24, 25] and following blood transfusions [1] . the majority of the population in israel comprises jews (74.8%) and arabs (20.8%), of which 84.7% are muslims [26] . while almost all israeli-arabs were born in israel, jews compose a mixed population of israeli-born and non-israeli-born individuals. previously, we have shown that the overall hev seroprevalence in israel was 10.6% and higher in arabs (22.5%) compared with jews (10.3%); with most of the anti-hev positive jews being immigrants not born in israel [21] . however, an assessment of anti-hev prevalence in specific subgroups of arabs has not been done. the bedouins in israel compose a unique nomadic muslim arab population with cultural, historical and social uniqueness. while non-bedouin muslim arabs live in villages spread around the country, mostly in the northern and central part of israel and are not physically related to camels, many of the bedouins who live in southern israel own dromedary camels and consume camel products [27] . this study aimed to assess hev seroprevalence in camels (dromedary), in bedouins living in the southern part of israel in the vicinity of camels, in non-bedouins arabs (muslims living in northern and central israel) and in israeli-born jews and to assess the factors associated with anti-hev seropositivity. camel serum samples (n = 86) from female dromedary camels aged 3 (n = 43), 7 (n = 25) and 10 and above years (n = 18) were collected between 2015 and 2017 from bedouin households residing in the southern district of israel (as part of the middle east respiratory syndrome (mers) national control programme) and were used for determining hev seroprevalence in the local 1-humped camels. human serum samples were collected from: bedouins living in the southern district, in tribes that own dromedary camels (overall 305; 296 tested for the first time in this study and nine added from our previous study [21] ), non-bedouin arabs (overall 320; 297 from the present study and 23 from our previous study [21] ) and from jews (195 samples presented in our previous study [21] ). all human samples were randomly selected using an age-stratified sampling design from the stored sera bank of the israeli centre for disease control. the samples were selected from those collected between 2009 and 2016, >20 years old who were born in israel. the selection was enabled by the availability of basic demographic information including age, gender, place of residence (city), birth country and population group (bedouins; non-bedouin arabs; jews [28] ) for each sample in this sera bank. the socio-economic status was allocated on the basis of the given address using the socio-economic residential classification published by the israeli central bureau of statistic [29] . this socio-economic status is based on 14 variables including demographic characteristics, education, lifestyle, etc. and was divided into low (1-5) vs. high (6-10). sera collection was approved by the legal department of the israeli ministry of health. total anti-hev antibodies were assessed in camel samples using hev-ab elisa kit (wantai, biologic pharmacy enterprise, beijing, republic of china) which detects total antibodies and is suitable for detection of anti hev antibodies in non-human sera. hev rna in camel sera was assessed with the realstar hev kit (altona diagnostics gmbh, hamburg, germany) which, according to the manufacturer, should detect all hev genotypes. human samples from the current study were tested with anti-igg hev elisa kit (wantai, biologic pharmacy enterprise, beijing, republic of china) which recognises human antibodies against all hev genotypes with 97.96% sensitivity and 99.99% specificity [30] . assays were performed according to the manufacturer's instructions. all serological equivocal results were considered as negative. in the previous study, we used the dsi-anti-hev-igg kit (diagnostic systems italy, saronno, italy) [21] . merging the current and previous study data were applicable after comparing the kits performances using 90 positive and negative samples, revealing 95.6% concordance between the kits. hev rna was not assessed in human sera due to lack of sufficient serum material. descriptive analysis was done for the study populations. prevalence rates of anti-hev antibodies in camels and in human sera were calculated by dividing the number of samples positive to anti-hev antibodies by the total number of samples tested in each group. for the human samples, we calculated the prevalence rates in each of the studied populations by age group, gender and socio-economic status. we used the cochran-armitage trend test to evaluate trends in binomial proportions and the χ 2 test to compare between population groups. logistic regression analyses were applied to assess the factors associated with anti-hev seropositivity. interaction was assessed for each variable associated with hev seropositivity. statistical significance was evaluated using 2-sided tests with an alpha level of 0.05. all analyses were performed using sas enterprise guide (version 7.12 hf5, sas institute inc., cary, nc, usa). of the samples obtained from camels, 68.6% (95% ci 57.7-78.2%) were anti-hev igg positive. the seroprevalence among camels 10 and above, 7 and 3 years old were 88.9% (16/18 15 .0% of the samples obtained from non-bedouin arabs (95% ci 11.3-19.4%) and 3.1% of the samples obtained from jews (95% ci 1.1-6.6%) were anti-hev igg positive. hev seropositivity was significantly higher in bedouins and in non-bedouin arabs compared to jews (p-value < 0.001). the difference in seropositivity rates between bedouins and non-bedouin arabs was statistically significant (p-value = 0.032). table 2 demonstrates the odds for being hev positive, by population group, demonstrating significantly higher odds among the older age groups of both arab populations. a single variable analysis demonstrated that seropositivity was significantly higher among arabs (bedouins and non-bedouin); older age groups (30-39 and ⩾40 years) and lower socioeconomic status (table 3) . no significant interaction was identified. these associations remained statistically significant in the multivariable analysis, except for the lower socio-economic status ( table 3) . in a sensitivity analysis performed for the classification of the equivocal samples as positive or negative, no significant difference was observed. hev prevalence among dromedary camels, has been studied in the past and following literature review we realise that our finding (68.6%) is higher than the 22.4% seroprevalence rates reported in ethiopia [31] , but similar to the 62.9% seropositivity among dromedary camels recently reported in egypt, a nearby country [32] . together, these results suggest that the 1-humped camels in our region are highly seropositive for hev. r. bassal et al. both bedouins (21.6%) and non-bedouin arabs (15.0%) were characterised by significantly higher hev seroprevalence rates compared with the overall low seropositivity rates (3.1%) observed in jews. together with the high seroprevalence of hev in camels, these results indicate endemicity of hev in israel. worldwide, the seroprevalence of hev documented in human populations varies between high (⩾20.0%), medium (10.0-19.9%) and low (<10.0%). high hev seroprevalence rates were reported in uganda (47.7%) [33] , poland (43.5% and 49.6%) [9, 18] , bolivia (34.8%) [13] , jordan (30.9%) [7] and south africa (25.3%) [12] . medium hev seroprevalence rates were reported in china (19.9%) [15] , belgium (18.3%) [34] , portugal (16.3%) [11] and the mediterranean region (11.9%) [35] . low hev seroprevalence rates were reported in new zealand (9.7%) [10] , scotland (6.1%) [36] and iceland (2.1%) [37] . accordingly, within israel, high, medium and low hev seroprevalence rates were observed in bedouins, non-bedouin arabs and jews. the higher seroprevalence rate in bedouins could be attributed to their low socio-economic status but could also result from exposure to camels, especially as the latter were highly seroprevalent (possibly a consequence of hev-7 infection [4] ). as camel samples were hev rna negative and fecal specimens were not available for analysis, this hypothesis could not be further explored. in jordan, a nearby country, an overall of 30.9% anti-hev prevalence rate was determined in the local population and owning camels was associated with increased odds of hev seropositivity [7] . future studies should aim to assess the hev rna prevalence in camel's feces and in their fresh blood samples, as well as in similar samples obtained from bedouins who own camels. with the recent report suggesting that hev originated in asia, most likely from a human ancestor that existed ∼4500-6800 years ago and that the split of camel-infecting genotypes occurred during camel domestication, hev circulation between camels and bedouins could be expected [38] . the high hev seropositivity rate observed in non-bedouin arabs, most of whom live in northern parts of israel, could be attributed to exposure to hev-3 which is possibly circulating in the country and was recently identified in local sewage facilities mainly in the north of israel [39] . as both judaism and islam forbid the consumption of pork, we find it unlikely that this is the main risk for the identified hev seroprevalence in the muslim population. the low hev seroprevalence observed in israeli-born jews may be associated with the overall higher socio-economic status characterising this population, living in better sanitary conditions than the bedouins and most of the non-bedouin arabs, rendering hev infection less conceivable. the variability observed between different population groups and in different studies, may be explained by exposure to potential risk factors associated with hev infection as lifestyle habits (dietary), environmental conditions, geographic location, occupation, religion (as the prohibition of consuming pork among jews and islam) but also to co-morbidities. differences may also be explained by the assays used for antibodies detection, demonstrating wide variation in the ability to detect hev antibodies (sensitivity and specificity) [40] . the association of hev seropositivity with age observed in our study has also been reported by others [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] and may be explained not only by the cumulative lifetime exposure to hev, but also by cohort effect, where certain population groups were more likely to be exposed to the virus during their life. the major limitation of our study is the lack of data on characteristics that might be important to exposure to hev, such as actual exposure to animals, consumption of camel meat and the quality of the local water source. additionally, the cross-sectional nature of the data cannot establish a temporal relationship between risk factors and outcome, thus limits the interpretation of the results. finally, hev rna from human samples could not be assessed due to limited serum volumes. however, for the first time, we have investigated samples from local camels and from specific human populations, with fair distribution by age groups and gender. based on this cohort, our results suggest that hev is endemic in israel and that specific population groups like bedouins and non-bedouin arabs are at higher risk of hev infection. further studies are needed to determine the hev genotypes circulating among dromedary camels and the specific arabs populations. author orcids. r. bassal, 0000-0002-0086-2968. hepatitis e: a disease of reemerging importance hepatitis e virus: advances and challenges easl clinical practice guidelines on hepatitis e virus infection hepatitis e virus genotypes and evolution: emergence of camel hepatitis e variants new hepatitis e virus genotype in camels, the middle east chronic infection with camelid hepatitis e virus in a liver transplant recipient who regularly consumes camel meat and milk seroprevalence and risk factors of hepatitis e infection in jordan's population: first report hepatitis e virus seroprevalence, seroincidence and seroreversion in the german adult population molecular and serological infection marker screening in blood donors indicates high endemicity of hepatitis e virus in poland prevalence of hepatitis e virus antibodies and infection in new zealand blood donors a nationwide serosurvey of hepatitis e virus antibodies in the general population of portugal racial differences in seroprevalence of hav and hev in blood donors in the western cape, south africa: a clue to the predominant hev genotype? seroprevalence of hepatitis a virus, hepatitis e virus, and helicobacter pylori in rural communities of the bolivian chaco prevalence, risk factors and molecular evaluation of hepatitis e virus infection among pregnant women resident in the northern shores of persian gulf seroprevalence and risk factors of hepatitis e virus infection among the korean, manchu, mongol, and han ethnic groups in eastern and northeastern china hev and hav seroprevalence in men that have sex with men (msm): an update from meat consumption is a major risk factor for hepatitis e virus infection hepatitis e virus igg seroprevalence in hiv patients and blood donors, west-central poland hepatitis e prevalence in a sexual high-risk population compared to the general population seroprevalence of hepatitis e virus in chronic hepatitis c in brazil prevalence of hepatitis e virus antibodies prevalence of hav ab, hev (igg) syphilis among sheltered homeless adults in tehran screening of ready-to-eat meat products for hepatitis e virus in switzerland hepatitis e virus infections in blood donors hepatitis e virus seroprevalence among adults, germany. emerging infectious diseases population by religion and population by population group human brucellosis outbreak acquired through camel milk ingestion in southern israel municipal authorities ranking by the socio-economic index wantai hepatitis e virus diagnostics. wantai hepatitis e virus diagnostics hev-igg elisa serological evidence of hepatitis e virus infection in dromedary camels in ethiopia hepatitis e virus infection in dromedaries hepatitis e virus seroprevalence and correlates of anti-hev igg antibodies in the rakai district clinical burden of hepatitis e virus infection in a tertiary care center in flanders hepatitis e virus seroprevalence rate among eastern mediterranean and middle eastern countries; a systematic review and pooled analysis hepatitis e virus (hev) in scotland: evidence of recent increase in viral circulation in humans low prevalence of hepatitis e in iceland: a seroepidemiological study origin and dispersal of hepatitis e virus hepatitis e virus genotype 3 in sewage and genotype 1 in acute hepatitis cases laboratory challenges in the diagnosis of hepatitis e virus acknowledgments. the authors would like to thank nadia pekurovski, israel center for disease control, for her assistance in data collection and the laboratory assistant. this research received no specific grant from any funding agency, commercial or not-for-profit sectors.conflict of interest. none. key: cord-285061-7vah0pjm authors: khosravi, a.; chaman, r.; rohani-rasaf, m.; zare, f.; mehravaran, s.; emamian, m. h. title: the basic reproduction number and prediction of the epidemic size of the novel coronavirus (covid-19) in shahroud, iran date: 2020-06-10 journal: epidemiol infect doi: 10.1017/s0950268820001247 sha: doc_id: 285061 cord_uid: 7vah0pjm the aim of this study was to estimate the basic reproduction number (r(0)) of covid-19 in the early stage of the epidemic and predict the expected number of new cases in shahroud in northeastern iran. the r(0) of covid-19 was estimated using the serial interval distribution and the number of incidence cases. the 30-day probable incidence and cumulative incidence were predicted using the assumption that daily incidence follows a poisson distribution determined by daily infectiousness. data analysis was done using ‘earlyr’ and ‘projections’ packages in r software. the maximum-likelihood value of r(0) was 2.7 (95% confidence interval (ci): 2.1−3.4) for the covid-19 epidemic in the early 14 days and decreased to 1.13 (95% ci 1.03–1.25) by the end of day 42. the expected average number of new cases in shahroud was 9.0 ± 3.8 cases/day, which means an estimated total of 271 (95% ci: 178–383) new cases for the period between 02 april to 03 may 2020. by day 67 (27 april), the effective reproduction number (r(t)), which had a descending trend and was around 1, reduced to 0.70. based on the r(t) for the last 21 days (days 46–67 of the epidemic), the prediction for 27 april to 26 may is a mean daily cases of 2.9 ± 2.0 with 87 (48–136) new cases. in order to maintain r below 1, we strongly recommend enforcing and continuing the current preventive measures, restricting travel and providing screening tests for a larger proportion of the population. according to the world health organization covid-19 dashboard [1, 2] , as of 29 april 2020, the novel coronavirus pandemic has spread to 213 countries, areas or territories with 2 959 929 confirmed cases and 202 733 confirmed deaths. some resources even report higher numbers of patients and deaths, and the numbers continue to escalate. [3] the islamic republic of iran was the first middle east country to report a case of death (19 february 2020) and is currently among countries with high prevalence of covid-19 [1, 3] . by 29 april, there were 92 584 confirmed cases in iran, 5877 of which had deceased [3] . given the rapid spread of the virus, the government immediately responded by establishing more than 40 laboratories within two weeks after the start of the epidemic, and consequently, there was a sudden spike in the reported number of positive cases. currently, there are 126 laboratories actively conducting pcr testing throughout the country. the timeline graph of the epidemic is presented in figure 1 . the first cases were immediately reported to the health department, and preventive protocols were developed and put in place to limit the further spread of the infection. these included cancelling in-person classes in schools and universities as of 25 february 2020, and switching to online platforms, as well as public awareness campaigns that encourage citizens to minimise face-to-face contact and promote social distancing. since the timeframe from 20 march to 02 april coincides with the persian new year or 'norouz' holidays in iran, there were even more stringent restrictions to limit the social activities, visiting family and friends, trips, shopping and festivals that are considerably more common around this time of the year. in epidemics, close monitoring and evaluation is necessary to investigate the efficiency of control measures, determine the potential community transmission patterns and predict the progression of the epidemic and the trajectory of the epidemic curve. one useful epidemic measure that helps investigate the transmissibility of infection is the reproduction number (r). the basic reproduction number (r 0 ) is the average expected number of new cases infected by a primary case and must be estimated early during an epidemic [4, 5] . r 0 can be affected by various factors such as the probability of transmission upon contact between an infected case and a susceptible person, the frequency of contact and the duration of infection in a person [6] . as the epidemic progresses, when the population begins to become immune and/or interventions are put in place, the effective reproduction number (r t ) can be estimated in real time [7] . the serial interval (si) of an infection is the mean duration between the symptom onset of two successive cases (the primary case and the secondary case). the force of infection (denoted λ), which describes the rate at which susceptible people acquire a given infection, is another useful parameter when implementing preventative measures [8] . according to the unpublished report of the ministry of health and medical education in iran, the incidence rate of covid-19 has been highest in semnan province (118 cases per 100 000 persons) and the highest incidence rate in iran by 01 april was seen in shahroud county. shahroud, in shahroud county and semnan province, is a city located in northeastern iran with a population of about 218 628 in 2016 [9] . the first confirmed case of covid-19 in iran was identified on 19 february in qom which is about 550 km from shahroud (fig. 2) . four days later a. khosravi et al. (23 february 2020), nasopharyngeal and throat swabs of five suspected cases in shahroud were submitted for viral nucleic acid testing, and two tested positive. one of these primary cases was a 74-year-old woman who had been hospitalised on 10 february, with chief complaints of fever and cough, and a travel history to qom. this indicates that the epidemic probably started almost one month before it was known to the public. given the high incidence rate in the region, the aim of this report is to estimate the r 0 of the covid-19 in the early stage of the epidemic (20 february to 04 march) and predict the trajectory of the epidemic and new cases in shahroud. the protocol of this study was reviewed and approved by the institutional review board of shahroud university of medical science (ir.shmu.rec.1398.160). the data for this study were all confirmed cases of covid-19 in shahroud county. the majority of cases were patients who were referred to imam hossein hospital (which is currently the only specialty hospital designated for the management of covid-19 patients in shahroud county); others were positive cases from public health centres. in the first month of the epidemic, all walk-in and referral cases were screened. of these, suspected cases were admitted and tested. after establishing new testing centres, the protocol was revised on 21 march 2020 so that all suspected outpatients would be tested at a public health care setting. for testing, two respiratory tract samples (throat and nasopharyngeal swabs) are collected and submitted for viral nucleic acid testing. all positive cases are systematically recorded in a designated registry which is used for follow-up and contact tracing. in this study, we used an informative prior distribution for the si, which was estimated as 7.5 ± 3.4 days for covid-19 in wuhan, china [10] , fit with a gamma distribution. we calculated the likelihood-based r 0 using a branching process with poisson likelihood. bootstrapping with 1000 times resampling was used for obtaining the distribution and confidence interval (ci) of r 0. we then used the estimates of r 0 , si and daily incidence to simulate the trajectories and project the future daily cumulative incidence where the main assumption was that the model follows a poisson distribution [5] . for each date t ≥ 2, the number of incident cases i t was drawn from a poisson distribution with mean r t t s=1 i t−s w s , where r t is the instantaneous reproduction number, w s is the discrete si distribution and i t−s is the incidence at time step ts. for a 30-day projection, we used a uniform distribution of 0.8 −1.3 for r 0 and bootstrapping with 1000 times resampling [5, 11] . for the projection in the next 30 days, we estimated a r 0 of 0.8 −1.3 based on the cultural background and the public health information level of the community, as well as the extent of case finding strategies. we also estimated r t for the first 67 days of the epidemic using the approach proposed by cori et al. [7] . data analysis was performed in microsoft excel and r (3.6.3) software using the 'incidence', 'earlyr', 'ggplot2' and 'projections' packages. r 0 estimation for 20 february to 01 april 2020 using the si distribution for the first two weeks (20 february to 04 march 2020), the maximum-likelihood value of r 0 was estimated at 2.7 (95% ci 2.1-3.4) for day 14 (04 march 2020), which is indicative of a propagated epidemic (fig. 3) . the maximum-likelihood value of r 0 decreased to 1.28 (95% ci 1.14-1.43) for day 30 (20 march 2020) and 1.13 (95% ci 1.03-1.25) for day 42 (01 april 2020). during the first 42 days of the epidemic (20 february to 01 april 2020), a total of 993 suspected samples were tested for covid-19 in shahroud, and 424 (42.7%) of them tested positive. among the positive cases, 47 people (11.1%) were public health cases or outpatients. during the timeframe between 02 april and 26 april, a total of 1208 new suspected cases were tested for covid-19 and 114 new confirmed cases (4.6 cases/day) were recorded during this period. the daily observed distribution of these confirmed cases by 26 april is illustrated in figure 4 . figures 5a-c. the overall predicted average number of new cases was 9.0 ± 3.8 cases per day. in figure 5b , the daily average of predicted incident cases was smoothed for the time span. the 30-day projected cumulative incidence in shahroud is shown in figure 5c ; approximately 271 (95% ci 178-383) new cases were estimated to be infected. based on the r t for the last 21 days (days 46-67 of the epidemic), it is predicted that the mean daily cases for 27 april to 26 may 2020 will be 2.9 ± 2.0 with 87 (48-136) new cases. the r t also showed a decreasing pattern to below 1 since day 28 of the epidemic which has remained near 1 to day 67 where it reaches 0.70 (fig. 6 ). the r 0 of an infection is commonly used to characterise its transmissibility during an epidemic. the trend of r 0 over time provides a measure of the effectiveness of control and prevention strategies in the community, and to control an epidemic, the goal is to reduce and keep the reproduction numbers below the value of 1 [7] . the basic reproduction number (r 0 ) is the average number of secondary infections transmitted by each case in a susceptible population. the biological potential of an agent affects r 0 , but it also depends on contact rate in the population and duration of infectiousness. it can be used for projection and modelling of infectious disease spreading in the population. in this study, the estimated r 0 of 2.7 (95% ci 2.1-3.4) during the early stage of the epidemic is in line with previous estimates [10, [12] [13] [14] . however, higher estimates of r 0 have been reported in larger populations [15, 16] . for precise estimation of r 0 , certain conditions must be met which include the precise detection of cases in the early stages of the epidemic, restricting the calculation to a small timeframe [11] , and choosing the appropriate estimation method [7, 17] . since the number and patterns of people's contacts vary by country for cultural and educational reasons, the r 0 can be different as well. also, r 0 can be different in subpopulations [18] . for precise detection of cases, all suspected cases (according to the screening protocol) and cases who have had close contact with confirmed cases should undergo viral nucleic acid testing. in shahroud, there were 993 suspected cases, and 427 of them tested positive during the first 42 days of epidemic. however, in the early stages of the epidemic in iran, there was limited capacity for testing, and the calculated r 0 may be an underestimation. during the study period of 67 days, the number of tests performed during the first 42 days and the second 25 days was 926 (43% tested positive) and 1208 (9.4% tested positive), respectively, which indicates a considerable expanding of testing to the community. while increased testing and case finding is expected to be associated with higher r 0 projections, we observed a steady decrease in r 0 and r t which can be attributed to the success of the interventions and not the higher number of immune persons. the mitigation and suppression strategies have managed to control the spread of the infection and reduce the reproduction rate through reduced contact and lower likelihood of transmission. maintaining these strategies is expected to control the size of the epidemic and keep the r below 1. the results of this study showed that r 0 has decreased temporally. this pattern, which is promising for controlling an epidemic, is due to interventions enforced by the health system starting from the early days of the epidemic. some of the most important measures were public education to promote social distancing and encouraging people to stay home. in addition, two other basic measures were taken in shahroud: (1) at the time of hospital discharge, all patients and their caregivers were provided with counselling and training on how to be isolated at home for 14 days, and families received information about how to care for patients; (2) active contact tracing with follow-up of patients' family members and friends, work colleagues and other possible contacts and referral of suspected cases to medical centres [8] . according to our 30-day projection for 27 april to 26 may, there should be a decrease in r 0 , and a total of 87 new cases are expected. this can be due to the spread of the virus by unidentified asymptomatic cases and increased testing on outpatients. so, we strongly recommend measures to identify these cases for total control of the epidemic. this study can inform health policymakers of the success of the preventive measures and interventions. it also emphasises the need for these measures to be continued along with stricter limitations in transportation until the transmission chain is broken and the epidemic is successfully controlled. the main strengths of this study include its careful design, taking throat and nasopharyngeal swabs for testing of all suspected cases, and systematic recording of positive cases. one limitation was that testing in the first period was only done for those admitted in the hospital. there were also potential limitations in the calculation of r 0 . given that r 0 is the average of r 0i in population subgroups; its value may be higher in some high-risk subgroups. as a result, the epidemic could be ongoing in these subgroups [18] . as the number of immune persons increases in the community, we determined r t which has decreased below r 0 . in conclusion, the r 0 of covid-19 in shahroud was considerably high at the onset of the epidemic, but with preventive measures and public education, it was reduced to 1.13 (95% ci 1.03-1.25) within 42 days. this reduction highlights the success of preventive measures in place, but mitigations and expansion of case finding should be continued. we strongly recommend mass screening and testing of suspected cases, travel restrictions especially during upcoming holidays and stringent monitoring of close contacts. these strategies will be challenged by the social and economic burden on the community, and sooner or later restrictions have to be relaxed. for example, low-risk business activities restarted in mid-april in most of the country, except tehran. our prescription for the community is to stop the traditional kiss, hug or hand-shake greeting gestures and place emphasis on face masks, hand hygiene, social distancing and strict monitoring of economic centres such as markets and public places. world health organization. coronavirus disease (covid-19) pandemic world health organization. coronavirus disease (covid-2019) situation reports, situation report 99 a simple approach to measure transmissibility and forecast incidence projections: project future case incidence modern infectious disease epidemiology, 3 edn a new framework and software to estimate timevarying reproduction numbers during epidemics seventy-five years of estimating the force of infection from current status data selected findings of 2016 national population and housing census early transmission dynamics in wuhan, china, of novel coronavirus-infected pneumonia estimation of the reproductive number of novel coronavirus (covid-19) and the probable outbreak size on the diamond princess cruise ship: a data-driven analysis nowcasting and forecasting the potential domestic and international spread of the 2019-ncov outbreak originating in wuhan, china: a modelling study preliminary estimation of the basic reproduction number of novel coronavirus (2019-ncov) in china, from 2019 to 2020: a data-driven analysis in the early phase of the outbreak china coronavirus: what do we know so far? modeling and forecasting trend of covid-19 epidemic in iran until may 13, 2020 extended sir prediction of the epidemics trend of covid-19 in italy and compared with hunan complexity of the basic reproduction number (r 0 ) epidemics, models and data using r acknowledgements. this work was supported by shahroud university of medical sciences (grant no. 98126). conflict of interest. the authors declare that there is no conflict of interest. a. khosravi et al. key: cord-276916-j53i5xfs authors: kraemer, m. u. g.; cummings, d. a. t.; funk, s.; reiner, r. c.; faria, n. r.; pybus, o. g.; cauchemez, s. title: reconstruction and prediction of viral disease epidemics date: 2018-11-05 journal: epidemiol infect doi: 10.1017/s0950268818002881 sha: doc_id: 276916 cord_uid: j53i5xfs a growing number of infectious pathogens are spreading among geographic regions. some pathogens that were previously not considered to pose a general threat to human health have emerged at regional and global scales, such as zika and ebola virus disease. other pathogens, such as yellow fever virus, were previously thought to be under control but have recently re-emerged, causing new challenges to public health organisations. a wide array of new modelling techniques, aided by increased computing capabilities, novel diagnostic tools, and the increased speed and availability of genomic sequencing allow researchers to identify new pathogens more rapidly, assess the likelihood of geographic spread, and quantify the speed of human-to-human transmission. despite some initial successes in predicting the spread of acute viral infections, the practicalities and sustainability of such approaches will need to be evaluated in the context of public health responses. infectious disease outbreaks pose a significant threat to human health. the frequency of such outbreaks is thought to have increased over the past decade. for example, quickly after an epidemic of ebola virus affected guinea, sierra leone, and liberia in 2013-2016 [1] , chikungunya virus (chikv) caused an extensive international epidemic in the americas and beyond, and was quickly followed by zika virus (zikv) emergence. to date, there have been more than 500 000 confirmed or probable cases of zikv but the true number of cases remains unknown [2] . yellow fever (yf), a vaccine-preventable disease, recently posed major public health problems. in 2015-2016, the largest yf outbreak since the 1980s was observed in angola and the democratic republic of the congo, causing 962 confirmed cases and 393 deaths [3] . yf also poses an ongoing public health risk to large, urban and under-vaccinated populations in the coastal areas of southern brazil, a country that successfully eradicated yf in the 1950s and 1960s [3] [4] [5] . examples of other emerging pathogens that have caused international health security concerns include the severe acute respiratory syndrome (sars) virus and the middle east respiratory syndrome coronavirus (mers-cov) [6] [7] [8] [9] . this list extends to other pathogens such as influenza, nipah and henipaviral diseases, and lassa fever [10] . these examples show the continued risks that infectious diseases pose and highlight the challenges of large international outbreaks to epidemic planning and response. during emerging infectious disease outbreaks, empirical information and mathematical modelling techniques are now commonly used to characterise and predict the spatio-temporal dynamics of the spread of pathogens. such analyses may help policymakers to evaluate the threat to public health, determine the resources required to reduce disease burden, and guide disease surveillance efforts and the deployment of interventions. in the last decade, our ability to perform such assessments has been improved by advances in a number of disciplines, including digital disease surveillance [11] , environmental modelling [12, 13] , genomics [14] and mathematical modelling [15] . for example, environmental variables such as rainfall and precipitation [13, [16] [17] [18] [19] [20] [21] [22] can be used to better understand the landscape within which the disease may be transmitted, and detailed transmission data from a small sampled population can be extrapolated to larger, un-surveyed areas [23] . attempts have been made to illustrate the spatial structure of epidemics mainly using human movement data [24] [25] [26] [27] , to provide mechanistic insights in how the disease may disperse locally [3, 28, 29] or how effective reactive vaccination campaigns may be [30] [31] [32] . there are continued efforts to reconstruct epidemic dynamics using information derived from pathogen genomic data, which contain unique information about the history of transmission [2, [33] [34] [35] . although each of these disciplines has an established relationship to disease prevention and control, the benefits of integrating them into a unified framework have yet to be fully achieved. here we describe the common applications and models used to predict acute viral diseases and discuss the current challenges and limitations. we then outline the advantages of integrating disparate data sources to advance our understanding of epidemic spread. we discuss how such research has been used in recent outbreaks and outline shortcomings that may be addressed in the future. phylogenetic and phylodynamic tools are increasingly being used to infer a range of outbreak properties [36] . common spatiotemporal analyses of pathogen genomes focus on mapping and predicting virus lineage exchange among locations, with the underlying aim of reconstructing the pathways of disease introduction and spread, albeit at a coarse spatial resolution, and often retrospectively [2, 8, 33, 35, 37, 38] . an additional feature that can be inferred from genomic data is the timing of individual founder introductions [39] . blue arrows in fig. 1 indicate the time when the first report was published inferring the likely geographic origin of four major international infectious disease outbreaks. phylogenetic tools can help to characterise the number of introductions that lead to disease transmission in a new location [41] , quantify the risk of cross-species transmission [42] , and infer ecological drivers of transmission [43, 44] . genome-derived estimates have been compared qualitatively to those from epidemiological data, but formal model-based integration of both data sources are rare [45, 46] . in principle, pairing genomic information with epidemiological inference should enable us to quantify the number of cases missed in each location and help to estimate parameters such as the basic reproductive number and doubling time of the epidemic, as done for zikv at the tail end of the epidemic (fig. 1a ) [46] [47] [48] . a common limitation when genetic data are used is the absence of a rigorous and formal sampling scheme. in many instances, genomic sampling is affected by convenience and expedience and may not reflect underlying incidence, although this can be improved post-hoc in large data sets via sub-sampling using, for example continuous phylogenetic inference [49] [50] [51] . strong sampling biases may affect estimates of the arrival time of a pathogen and its pathways of dissemination among locations [33] . static disease mapping is a powerful tool to visualise and defines the landscape within which transmission occurs, based on ecological drivers of transmission [17, 18, 22] . when combined with global data on human travel and mobility, it can be used to understand the global dynamic risk surface of infectious disease, especially when there are strong ecological determinants of transmission, as there are for the vector-borne diseases zika, dengue, chikungunya and yf [27, 52] . publication of reports that estimate geographic spread for the diseases in fig. 1 are indicated by green arrows. the global epidemic history of zika, for example, remains poorly understood. the challenge to accurately reconstruct the epidemic pathway of the virus is further complicated by its relatively unspecific clinical presentation. this may explain why the initial studies that aimed to understand the geographic origin of the zika epidemic in the americas were published relatively late into the epidemic (>1 year, fig. 1a ). for the other major outbreaks highlighted in fig. 1 , estimates of the geographic origin were documented between 6 and 8 months after the first reports of human cases ( fig. 1b-d ; table 1 ). however, given the underlying ecological determinants of transmission that restrict the reproduction of the virus in the mosquito vector species, large areas can be excluded from the risk of local virus transmission. when overlaying information on the reported presence of zika cases vs. the underlying ecological risk map, surveillance gaps may be identified [19, 27] . areas where there is a mismatch in the predicted presence and reported presence (i.e. cases detected) should be targeted for active surveillance. the spatial spread process of new pathogens, however, is not only determined by the underlying ecological determinants in each location but also by the dynamic nature of importation, often driven by human movements [61] . spatial models that take into account the patterns of human spread and mobility may, therefore, improve our ability to characterise and anticipate spatial expansion. different models have been proposed to predict the geographic spread of epidemics but rarely have they been used in real time during the course of an epidemic [3, 62] (fig. 1) . for example, during the yf outbreak in angola and the democratic republic of the congo, estimates of geographic spread to provinces outside luanda, the capital of angola, were published >6 months after the last cases were reported (fig. 1c) . such information could guide public health institutions to decide where and when to implement surveillance and control programs [27] . more work, however, is needed to dynamically map the spread of infectious diseases and to extract meaningful and interpretable quantities for public health practitioners. in parallel to these efforts to model the spread of pathogens at a meta-population level (e.g. between cities, regions, countries or continents), we also need to better understand transmission dynamics at a much more granular level and assess the characteristics of the inter-human transmission. while historically, the potential for inter-human transmission has often been summarised with a single statistic; the reproduction number r (i.e. the average number of secondary infections generated by a case). however, it has long been recognised that it is also essential to assess heterogeneities in individual r values, since the presence of super-spreading events may have a major impact on the risk of emergence and our ability to control outbreaks [63] . this was exemplified in a large mers-cov outbreak in south korea in 2015 in which only a small number of cases were responsible for the majority of infections [64, 65] . other factors that may drive the spatial differences in the reproductive number are ecological (population density, climatic factors, or others) and can now be readily incorporated in transmission models [66] . ideally, these assessments should be performed on detailed data documenting chains of transmission, as such data can provide precise quantification of the transmission potential and the impact of targeted interventions in different settings and over time, and allow testing specific hypotheses about the transmission process (e.g. what is the contribution of re-introductions to the overall dynamic?) [67] . however, such data are rarely available as it is difficult to identify the source of infection for most pathogens. as a result, sophisticated statistical techniques are often required to reconstruct chains of transmission and estimate transmission parameters from more limited data that may include: (i) in the context of zoonoses, the size of human clusters [68] [69] [70] or the proportion of surveillance cases that reported a contact with the natural reservoir [71] , (ii) the growth rate in the case count [72] [73] [74] [75] , (iii) partial data on chains of transmission [76] , or (iv) outbreak data where the timing of symptom onsets and location of cases are recorded in small communities such as households [77] [78] [79] [80] , schools [81] or villages [82] . in cases of high-density sampling, genomic data can help to reconstruct transmission chains [83] . mechanistic models of infectious disease dynamics can be used to make predictions about the future course of an outbreak within a given location [84] . increasingly, such models are being used in real time, such that predictions are updated every time a new data point becomes available [85, 86] . some other applications track pathogen evolution over time as data become available [87] . however, the perceived ability of such models to successfully or unsuccessfully make 'correct' predictions can generate considerable controversy [88, 89] . there are few studies that systematically investigate forecasting accuracy and its relationship to the length of time that is being predicted and to the quantity and quality of data available [90, 91] . other examples are forecasting challenges for ongoing epidemics such as chikv in the americas (https://www.darpa.mil/news-events/2014-08-15), evd in west africa [92] and seasonal influenza [85, 93] , designed and initiated by funding agencies and public health governments. this is an important area for future research. there are clear benefits to combining information from different data sources in order to better predict viral epidemic spread. previous work most commonly presents estimates from different sources side-by-side, for example, estimates of the epidemic reproductive number derived from genomic vs. epidemiological data [46] . such comparisons are important to assess the consistency of data sources and may help to derive new hypotheses. spurred by technological innovation such as portable sequencing using the minion device (oxford nanopore technologies, oxford, uk) [94] and by interdisciplinary collaborations during disease outbreaks, researchers have started to work to combine three types of transmission data: spatial, genomic and epidemiological which have now been published for three of the four major outbreaks we considered here (fig. 1, red arrows) [33, 38, 41] . for example, such interdisciplinary work helped to identify the introduction of zika into the americas [2] , investigated the main drivers of transmission of zikv through climatic suitability of its mosquito vectors [25] and tried to extrapolate how many people had been infected with the virus [23, 92, 93] . in the context of phylogenetic analyses, environmental and other spatial data may be helpful in reconstructing the drivers of transmission and spread using, for example, information on the reservoir or host movements [35, 95] . in turn, phylogenetic information may complement epidemiological analysis by providing more evidence on the transmission routes that are common in an outbreak [96] . this may be particularly useful for diseases that have a highly structured transmission dynamic, such as mers or sars, where a small number of people are responsible for the majority of secondary cases [63, 97] , transmission from the animal reservoir is frequent, or importation drives locally observed epidemics [33] . one common assumption in many epidemiological models is that it is equally likely for people to meet and infect others living in the same location and that population immunity is proportional to the demographic structure [98] . hence, observed cases are often assumed to arise from other cases that are reported locally as long as they are consistent with the generation time of the disease. however, a wellconnected location can, in principle, accrue a large number of incident cases through recurring introductions from elsewhere, rather than via local transmission [33] . these results can have large implications for surveillance and control, as different competing strategies (e.g. limiting importations or eradicating the disease locally) may be considered. while analytical approaches of various degrees of complexity have been proposed to probabilistically reconstruct transmission trees from incomplete outbreak data [73, 81, 97] , contact tracing, which can be very labour intensive [67] , remains a gold standard information source. this may allow us to is to determine the true distribution of cluster sizes (i.e. the number of subsequent cases resulting from each introduction) but is often only available for a small number of locations. however, using genomic data can help refine the understanding of heterogeneity in transmission but such framework does not yet allow to exactly quantify the fraction of observed cases that are attributable to local transmission versus introduction from elsewhere, or to determine how many importations are responsible for the local incidence, despite its crucial importance for eradication campaigns [42, 100, 101] . in the context of the zika outbreak in florida, combining genomic data from the outbreak with epidemiological analysis revealed that the outbreak was driven by a large number of introductions rather than by persistent local transmission. in the recent yellow fever outbreak in southern brazil, linking epidemiological, spatial and genomic data and techniques could provide insights into the transmission potential and risk of urban transmission [102] . one dataset and analysis alone would have not been strong enough to make such conclusions [33] . inferences about epidemic processes made using mathematical models rely on a number of assumptions. geographic modelling approaches, mostly informed by spatial ecology, attempt to fill gaps where no data has been observed, hence inferences may be uncertain, as the underlying ecological process may be poorly understood and dynamical aspects of the invasion process are ignored. these deficiencies can be ameliorated, in part, by adding virus genome data that contain information about past transmission and invasion patterns [103] . however, due to incomplete and poor sampling (as discussed above), genomic data alone may provide an incomplete picture of the timing of viral introduction and spread among locations. this, in turn, can be supported by the addition of epidemiological time series of reported cases and serological information about population immunity [104, 105] . despite this, building a joint inference framework that combines all available data sources and which characterises observation and sampling processes correctly is a daunting task. however, we are entering a period where the data for this task are becoming available in a timely fashion but need to ensure that results are communicated as soon as they are generated in order to avoid delays. initial successes have already led to important advancements in epidemic control and should progress to a tool-kit for guiding public health, hopefully available in real time for future epidemics. author orcids. m. u. g. kraemer http://orcid.org/0000-0001-8838-7147 ebola virus disease in west africathe first 9 months of the epidemic and forward projections zika virus in the americas: early epidemiological and genetic findings spread of yellow fever virus outbreak in angola and the democratic republic of the congo 2015-16: a modelling study the elimination of urban yellow fever in the americas through the eradication of aedes aegypti the hidden geometry of complex, network-driven contagion phenomena emergence and pandemic potential of swine-origin h1n1 influenza virus unifying viral genetics and human transportation data to predict the global transmission dynamics of human influenza h3n2 spread of a novel influenza a (h1n1) virus via global airline transportation who's blueprint list of priority diseases digital disease detection -harnessing the web for public health surveillance updates to the zoonotic niche map of ebola virus disease in africa mapping the zoonotic niche of marburg virus disease in africa towards a genomics-informed, real-time, global pathogen surveillance system modeling infectious disease dynamics in the complex landscape of global health the global distribution and burden of dengue progress and challenges in infectious disease cartography global distribution and environmental suitability for chikungunya virus mapping global environmental suitability for zika virus mapping the zoonotic niche of ebola virus disease in africa mapping the zoonotic niche of lassa fever in africa existing and potential infection risk zones of yellow fever worldwide: a modelling analysis model-based projections of zika virus infections in childbearing women in the americas assessing seasonal risks for the introduction and mosquito-borne spread of zika virus in europe potential for zika virus introduction and transmission in resource limited countries in africa and asia-pacific assessment of the potential for international dissemination of ebola virus via commercial air travel during the 2014 west african outbreak zika virus transmission in angola and the potential for further spread to other african settings nowcasting the spread of chikungunya virus in the americas local and regional spread of chikungunya fever in the americas yellow fever in africa: estimating the burden of disease and impact of mass vaccination from outbreak and serological data the revised global yellow fever risk map and recommendations for vaccination, 2010: consensus of the informal who working group on geographic risk for yellow fever global yellow fever vaccination coverage from 1970 to 2016: an adjusted retrospective analysis. the lancet infectious diseases genomic epidemiology reveals multiple introductions of zika virus into the united states evolution and spread of ebola virus in liberia virus genomes reveal factors that spread and sustained the ebola epidemic unifying the epidemiological and evolutionary dynamics of pathogens virus evolution and transmission in an ever more connected world epidemic establishment and cryptic transmission of zika virus in brazil and the americas bayesian phylogeography finds its roots situation report. yellow fever genomic and epidemiological characterisation of a dengue virus outbreak among blood donors in brazil mers-cov spillover at the camel-human interface explaining the geographic spread of emerging viruses: a new framework for comparing viral genetic information and environmental landscape data using viral gene sequences to compare and explain the heterogeneous spatial dynamics of virus epidemics middle east respiratory syndrome coronavirus: quantification of the extent of the epidemic, surveillance biases, and transmissibility pandemic potential of a strain of influenza a (h1n1): early findings birth-death skyline plot reveals temporal changes of epidemic spread in hiv and hepatitis c virus (hcv) insights into the early epidemic spread of ebola in sierra leone provided by viral sequence data bayesian molecular dating: opening up the black box phylogeography takes a relaxed random walk in continuous space and time phylodynamic assessment of intervention strategies for the west african ebola virus outbreak anticipating the international spread of zika virus from brazil genomic surveillance elucidates ebola virus origin and transmission during the 2014 outbreak assessing the international spreading risk associated with the 2014 west african ebola outbreak early epidemic dynamics of the west african 2014 ebola outbreak: estimates derived with a simple two-parameter model resurgence of yellow fever in angola modelling the large-scale yellow fever outbreak in luanda, angola, and the impact of vaccination transmission and evolution of the middle east respiratory syndrome coronavirus in saudi arabia: a descriptive genomic study potential for the international spread of middle east respiratory syndrome in association with mass gatherings in saudi arabia. plos current outbreaks interhuman transmissibility of middle east respiratory syndrome coronavirus: estimation of pandemic risk inferences about spatiotemporal variation in dengue virus transmission are sensitive to assumptions about intra-urban human mobility spatial spread of the west africa ebola epidemic superspreading and the effect of individual variation on disease emergence mers in south korea and china: a potential outbreak threat? mers-cov outbreak following a single patient exposure in an emergency room in south korea: an epidemiological. the lancet elsevier ltd big city, small world: density, contact rates, and transmission of dengue across pakistan chains of transmission and control of ebola virus disease in conakry strategies for containing an emerging influenza pandemic in southeast asia inference of r0 and transmission heterogeneity from the size distribution of stuttering chains detecting differential transmissibilities that affect the size of self-limited outbreaks using routine surveillance data to estimate the epidemic potential of emerging zoonoses: application to the emergence of us swine origin influenza a h3n2v virus how generation intervals shape the relationship between growth rates and reproductive numbers similar impacts of control measures practice of epidemiology a new framework and software to estimate time-varying reproduction numbers during epidemics real-time estimates in early detection of sars exposure patterns driving ebola transmission in west africa: a retrospective observational study a bayesian mcmc approach to study transmission of influenza: application to household longitudinal data household transmission of 2009 pandemic influenza a (h1n1) virus in the united states association between antibody titers and protection against influenza virus infection within households household transmission of influenza virus role of social networks in shaping disease transmission during a community outbreak of 2009 h1n1 pandemic influenza how social structures, space, and behaviors shape the spread of infectious diseases using chikungunya as a case study quantifying influenza virus diversity and transmission in humans mathematical modeling of the west africa ebola epidemic real-time assessment of health-care requirements during the zika virus epidemic in martinique temporal changes in ebola transmission in sierra leone and implications for control requirements: a real-time modelling study nextstrain: real-time tracking of pathogen evolution ebola: the power of behaviour change estimating the future number of cases in the ebola epidemic -liberia and sierra leone probabilistic forecasting in infectious disease epidemiology: the 13th armitage lecture evaluating probabilistic dengue risk forecasts from a prototype early warning system for brazil the rapidd ebola forecasting challenge special issue: preface results from the second year of a collaborative effort to forecast influenza seasons in the united states real-time, portable genome sequencing for ebola surveillance genomic epidemiology reconstructs the introduction and spread of zika virus in central america and mexico emerging concepts of data integration in pathogen phylodynamics unraveling the drivers of mers-cov transmission infectious diseases of humans: dynamics and control mapping residual transmission for malaria elimination spatial and temporal dynamics of superspreading events in the 2014-2015 west africa ebola epidemic measuring the path toward malaria elimination genomic and epidemiological monitoring of yellow fever virus transmission potential global spread of dengue virus types: mapping the 70 year history understanding herd immunity use of serological surveys to generate key insights into the changing global landscape of infectious disease key: cord-348039-kl1a0au3 authors: majowicz, s. e. title: what might the future bring? covid-19 planning considerations for faculty and universities date: 2020-04-29 journal: epidemiol infect doi: 10.1017/s0950268820000898 sha: doc_id: 348039 cord_uid: kl1a0au3 this paper applies a scenario planning approach, to outline some current uncertainties related to covid-19 and what they might mean for plausible futures for which we should prepare, and to identify factors that we as individual faculty members and university institutions should be considering now, when planning for the future under covid-19. although the contextual focus of this paper is canada, the content is likely applicable to other places where the covid-19 epidemic curve is in its initial rising stage, and where universities are predominantly publicly funded institutions. at the start of april 2020, the human health and disruptive effects of the covid-19 pandemic were being felt globally. academia has responded in multiple ways, from suspending in-person classes following social distancing directives, to mobilisation of those disciplines directly related to response efforts. for example, numerous modelling studies have been produced that illustrate the anticipated impacts of various interventions (e.g. social distancing) on the epidemic curve of covid-19 [1, 2] . while these models show epidemic curves that stretch into the fall and into 2021, there are sufficient uncertainties (e.g. how long will immunity last?) and complex dynamics (e.g. how will citizens react as social distancing measures remain in place?), meaning such results should not be used as the sole basis for planning for the future. in 2009, the public health agency of canada (phac) produced a planning document, to help prepare for canada's anticipated fall wave of pandemic h1n1 influenza [3] . this document contained: (a) descriptions of plausible future scenarios for how canada might experience the 2009 pandemic; and (b) planning considerations for phac including factors that could impact human and financial resources. the phac report was underpinned by two planning methodologies: scenario planning [4] , which aims to describe the range of plausible futures so that decisions and plans can be robust in the face of uncertainty; and a modified political, economic, social, technological (pest) analysis [5] , a framework for identifying macro-level factors in the wider environment that can impact organisations' abilities to function. its purpose was to prepare phac staff to think about the different ways the future could unfold, and to think about the different factors that could impact the way they did business, so that planning and decisions could be more robust and less likely to be thwarted by surprise. this paper applies a similar approach, in order to: (a) outline some current uncertainties related to covid-19, and what they might mean for plausible futures for which we should prepare; and (b) list factors that we as individual faculty members and university institutions should be considering now, when planning for the future under covid-19. although the contextual focus of this paper is canada, the content is likely applicable to other places where the covid-19 epidemic curve is in its initial rising stage, and where universities are predominantly publicly funded institutions. given that the pandemic is rapidly evolving, the applicability of this work to various contexts is also expected to change over time. there are several key 'axes of uncertainty' about how the pandemic will unfold in canada, that we should consider when planning for the future, for time horizons of the fall 2020 term (september to december 2020), the 2020/2021 academic year (september 2020 to august 2021) and beyond. these axes should be considered together, not singly. social distancing/activity restrictions: from 'even more lockdown', to 'returning to usual' will we still be social distancing, like we currently are, in the fall? will we have had to sustain the current level of social distancing from march to september? or will there be some sort of lifting (and reinstating?) of measures and restrictions? will measures be similar to how they are now, or less restrictive, or more restrictive? will measures be applied as broadly as they are now, or become more targeted to specific groups, characteristics or functions? these uncertainties mean that we could find ourselves in situations like 'things are similar to what they are now', or 'things are more restrictive, and we need to deliver upcoming terms even more remotely' or 'things are back to normal, but with uncertainties and the need to change quickly if the virus throws us a curveball or three'. impact on our people: from 'on our radar but rare', to 'big losses, heavy impacts' at the time of writing, most individuals within the university community (and ontario and canada as a whole) are mainly impacted by covid-19 via the social distancing measures currently in place, but it is reasonable to expect that the direct health impacts of the virus will become more widely felt. how many of our staff, faculty, and students will get sick? will we lose substantial workforce capacity (e.g. capacity for individual research projects, for university-wide operations, or to cover for absent instructors and administrators)? if we lose workforce capacity, will it be widespread or localised, ongoing or sporadic? will we face widespread grief, burnout and mental health impacts related to illness, intensive care, intubation, death or isolation? how will those populations currently marginalised by society or within the university community be disproportionately impacted? how will these impacts intersect with our cultures and customs? these uncertainties mean that we could find ourselves in situations like 'most people are not feeling substantial personal impacts', as we are now, to 'many people are substantially impacted' (e.g. they are sick or dealing with death and grief of family, friends and colleagues, there are deaths among the university community, there is a noticeable loss of workforce). universities as part of society: from 'doing what we usually do', to 'widespread mobilisation' at the time of writing, most of the university community (i.e. staff, faculty and students) are in the same roles and with essentially the same responsibilities as they were pre-pandemic. and although we are seeing academia voluntarily donating goods (e.g. personal protective equipment, reagents, swabs) and services s. e. majowicz (e.g. medical students conducting contact tracing) [6, 7] , the vast majority of university physical and human resources are still being used for the same purposes as they were pre-pandemic. a widespread and often subconscious assumption people seem to be making is that our roles and responsibilities as individual faculty or as universities, for example, are guaranteed to continue to be 'business as usual, albeit maybe remotely'. will we all stay in our roles as usual (even if working remotely)? will university resources be called upon in new ways to help the pandemic response, for example using laboratory or dormitory space for testing and community care? [8] will we 'lose' faculty, staff or students to front-line institutions so they can help aide response efforts, and if yes for how long? will public funding for universities be diverted? will immune individuals be mobilised (e.g. to fulfill essential public/ in-person functions, to donate plasma)? these uncertainties mean that we could find ourselves in situations like 'we are doing the same business we've always been doing', to 'ad hoc or individual volunteering of physical and/or human resources to support pandemic responses', to 'universities are obligated to repurpose resources to aide in the pandemic response', or perhapsin the most extremeeven to 'universities see their public funding diverted to pandemic response'. there are many dynamic factors influencing the pandemic and how it will unfold. considering and organising these factors by characteristics that predispose people to severe outcomes assess which segments of the university community are particularly vulnerable to severe outcomes, and create policies or accommodations to ensure adequate protection. because the duration of immunity is unclear, prepare for a possible future where a proportion of the university community is susceptible (and perhaps in isolation) at any given time. distribution of cases/death (e.g. by age, sex, risk factors) assess which segments of the university community are particularly vulnerable to illness/death; beyond policies/accommodations to protect them, create contingency plans to function in their absence (e.g. the older cohort of faculty and staff). social, economic factors create flexible options (e.g. for remote classes) that function in the face of changing socio-economic disadvantage (e.g. ability to afford internet connections) and broader social forces (e.g. caregiving responsibilities). groups that society/academia marginalises, or excludes from power, privilege create equitable institutional policies that adequately support groups that are typically marginalised or excluded. compliance consider situations where staff, faculty and students may not, or cannot, comply with public health directives or organisational policies, and create incentives for compliance (e.g. adjusting performance assessments so those whose research suffers when on-campus activities are suspended are not disproportionately disadvantaged). workforce availability consider how a proportion of the population ill or unable to work may drive the availability of external or temporary workers (e.g. sessional lecturers paid per course, casual staff), by reducing broader workforce availability while also increasing the demand for highly qualified individuals. political decisions make contingency plans in case political decisions (e.g. restrictions on supplies, protectionist policies) impact activities (e.g. ability to share data, research resources). social, political appetite for new ways of working consider how successfully conducting university activities under pandemic conditions (e.g. delivering remote classes) may lead to post-pandemic views on how universities can/should function (e.g. push for cost-saving, multi-institution online courses). equitable access to technology identify how issues like cost, geographic availability and connectivity to different technologies can (dis)advantage different groups (e.g. rural students with intermittent internet may attend fewer video classes). workforce capacity make contingency plans for classes, research projects and administrative tasks that account for some % of the workforce missing, some % at less than full capacity (overall, and at different times), and absent expertise, experience, authority and skills (e.g. identify instructors to cover different courses in the event of sudden illness/absenteeism). trust and reputation identify core business functions that rely on reputation and trust (e.g. universities' abilities to recruit students; researchers' abilities to build and sustain partnerships), and ensure all actions do not erode said trust/reputation (e.g. ways that students are treated during the pandemic will demonstrate how the institution values students). faculty expertise identify ways to reconsider workload, so faculty experts critical to the pandemic response (e.g. mathematical modelling, epidemiology), and key to the organisation's own planning (e.g. scenario planning, digital pedagogy, crisis communication, ethics) can devote adequate time to new activities. broad categories can challenge us to think broadly and plan for influences from 'unexpected' domains. the pest framework [5] uses political, economic, social and technological categories, which the phac h1n1 planning document [3] expanded to also include: the disease; population vulnerabilities; regulatory factors and the capacity to respond. other expansions of the pest framework include environmental, legal and ethical categories [9, 10] . a diagram showing the types of factors within these categories is given in figure 1 , and table 1 presents a selection of these factors, together with practical ways that faculty and universities might consider them in their planning. such planning can draw on existing contingency planning literature, including in the areas of outbreaks, disaster and emergency response and business operations, e.g. [11] [12] [13] . as well, such planning can and should be undertaken at all levels within academic institutions, in that individuals, departments and institutions can all evaluate their activities and abilities to deliver core business functions in light of the factors presented here. for example, individual researchers and research groups can conduct continuity and contingency planning for research projects, laboratory functions and graduate student theses. departments and institutions can, for example, create contingency plans for teaching commitments and administrative functions by identifying back-ups, or determining which can be temporarily suspended. additionally, departments and institutions can take a unit-or organisation-wide approach to allocating or redistributing common resources (e.g. online teaching supports) to best serve the needs of the whole. it is important to note that the factors and examples provided here do not form a complete list, and that different individuals, departments and institutions will have specific or unique issues with which they will have to deal (e.g. disruption of in-person data collection such as visit schedules for clinical trials). thus, comprehensive planning should include a full assessment of possible issues specific to each individual, department or institution, and should not be limited to the factors and examples presented here. given the range of uncertainties we currently face with covid-19, and the numerous broader forces that will influence how the pandemic will unfold, what concrete actions can we take? first, we can use the possible future situations to 'test' how well the decisions and plans we are currently making could hold up, under a range of different futures. for example, in preparing courses for fall offerings, we might choose to build in flexibility to allow pivoting between in person and remote delivery quickly, should the pandemic shift. second, we can use the possible future scenarios to ask 'what plans would i wish i had implemented now, if this future comes to pass?'. for example, research project supervisors can consider what alternate training to give graduate students and technical staff now, so they can cover for each other in case of illness and absence. third, we can use the broad list of factors to 'test' whether decisions and plans being made now might need to change if these factors change. for example, researchers and departments can consider how they would need to respond to future declines in graduate enrollment, for example to minimise impacts on research projects and teaching assistant capacities. or, as another example, staying abreast of information about predisposing factors (e.g. comorbidities) and immunity will allow institutions to remain flexible so our most vulnerable can continue to self-isolate at home even if social distancing measures lift. fourth, we can apply an equity lens throughout our planning and decision-making, to ensure thatat a bare minimumwe do not perpetuate or amplify existing barriers or disadvantages because of our individual and organisational decisions. covid-19 is manifesting in a world that is gendered [14] , ableist [15] and racialised [16, 17] , and it behooves each of us to ensure our individual and organisational responses work to oppose discrimination. specifically, we can work towards equitable policies, plans and decisions by taking three actions, by 'valuing all individuals and populations equally, recognising and rectifying historical injustices and providing resources according to need' [18] . finally, the axes of uncertainty and macro-environment factors given here can be expanded, both as the pandemic unfolds and new knowledge is generated, and by integrating perspectives from a wide range of backgrounds (e.g. economics, ethics, history, sociology, geography, planning) to identify additional key uncertainties about the future and key planning considerations. additionally, individuals and institutions in locations that are further along in the pandemic (e.g. asia) can delineate additional uncertainties and macro-environmental factors, for example those related to lifting social distancing measures, planning for subsequent potential waves, rebuilding resources or dealing with longer-term health and other impacts of both covid-19 and our responses to it. it will also be critical to hear how faculty and universities decide to deal with these issues, including in real time and via post-pandemic assessments (such as those aimed at improving pandemic plans). nevertheless, this paper can function as a starting point for individual and institutional planning, and to initiate conversations of how academia can plan for an uncertain future under covid-19. report 13 -estimating the number of infections and the impact of non-pharmaceutical interventions on covid-19 in 11 european countries mathematical modeling of covid-19 transmission and mitigation strategies in the population of ontario a tool for the potential fall 2009 wave of pandemic h1n1 to guide public health decision-making: an overview of the public health agency of canada's planning considerations living in the futures scanning the business environment canadian universities heed the call for help in the fight against covid-19 university of calgary medical students quadruple province's covid-19 contact-tracing capacity mayor: new haven asks for coronavirus housing help using the spelit analysis technique for organizational transitions healthcare public-private partnerships in italy: assessing risk sharing and governance issues with pestle and swot analysis contingency planning: preparation of contingency plans clinical review: mass casualty triage -pandemic influenza and critical care economic community of west african states disaster preparedness tabletop exercise: building regional capacity to enhance health security covid-19: the gendered impacts of the outbreak preventing discrimination against people with disabilities in covid-19 response racism and discrimination in covid-19 responses stop the coronavirus stigma now systems of power, axes of inequity: parallels, intersections, braiding the strands acknowledgements. i thank both referees for their constructive input during their review of this paper. in particular i thank referee 1 for noting some specific issues with which certain departments must deal (e.g. disruption of visit schedules for clinical trials), and referee 2 for the suggestion of the diagram in figure 1 .financial support. this research received no specific grant from any funding agency, commercial or not-for-profit sectors. s. e. majowicz reimbursed travel to attend face-to-face meetings. she has previously provided unpaid expertise as a member of the scientific advisory committee for cancer care ontario's infectious agents and cancer report, and the foodnet canada (formerly c-enternet) advisory committee (phac). key: cord-327867-1wkbjtji authors: da'ar, omar b.; ahmed, anwar e. title: underlying trend, seasonality, prediction, forecasting and the contribution of risk factors: an analysis of globally reported cases of middle east respiratory syndrome coronavirus date: 2018-06-11 journal: epidemiol infect doi: 10.1017/s0950268818001541 sha: doc_id: 327867 cord_uid: 1wkbjtji this study set out to identify and analyse trends and seasonal variations of monthly global reported cases of the middle east respiratory syndrome coronavirus (mers-cov). it also made a prediction based on the reported and extrapolated into the future by forecasting the trend. finally, the study assessed contributions of various risk factors in the reported cases. the motivation for this study is that mers-cov remains among the list of blueprint priority and potential pandemic diseases globally. yet, there is a paucity of empirical literature examining trends and seasonality as the available evidence is generally descriptive and anecdotal. the study is a time series analysis using monthly global reported cases of mers-cov by the world health organisation between january 2015 and january 2018. we decomposed the series into seasonal, irregular and trend components and identified patterns, smoothened series, generated predictions and employed forecasting techniques based on linear regression. we assessed contributions of various risk factors in mers-cov cases over time. successive months of the mers-cov cases suggest a significant decreasing trend (p = 0.026 for monthly series and p = 0.047 for quarterly series). the mers-cov cases are forecast to wane by end 2018. seasonality component of the cases oscillated below or above the baseline (the centred moving average), but no association with the series over time was noted. the results revealed contributions of risk factors such as camel contact, male, old age and being from saudi arabia and middle east regions to the overall reported cases of mers-cov. the trend component and several risk factors for global mers-cov cases, including camel contact, male, age and geography/region significantly affected the series. our statistical models appear to suggest significant predictive capacity and the findings may well inform healthcare practitioners and policymakers about the underlying dynamics that produced the globally reported mers-cov cases. this study set out to identify trends and seasonal variations; made a prediction based on the globally reported cases of the middle east respiratory syndrome coronavirus (mers-cov), extrapolated into the future by forecasting the trend and assessed contributions of various risk factors for the mers-cov cases. specifically, we consider the questions: (1) what is the underlying growth or trend of the globally reported mers-cov cases? (2) are there seasonal variations present in globally reported mers-cov cases over time and how do they affect the series? (3) what are the contributions of various risk factors for mers-cov cases? the motivation for this study is that to date, the world health organisation (who) places the mers-cov among the list of blueprint priority diseases. although a survey of the literature shows a rapid increase in research activities related to mers-cov [1] , yet there is still a general paucity of empirical literature examining trends and seasonality. to the best of our knowledge, there is a single study, which anecdotally examined seasonality and time series patterns of mers-cov to date [2] . given mers-cov remains a potential pandemic disease globally, it is important to understand the dynamics of the underlying growth or trend of the globally reported cases. the motivation for this study also comes from the aetiology of mers-cov, especially its causes, spread, the complexity of its diagnosis and mortality. mers-cov is a virus that causes severe viral pneumonia in humans, known to have a high mortality rate [3] [4] [5] [6] [7] and has clinical symptoms similar to severe acute respiratory syndrome coronavirus [8, 9] . it was first reported in saudi arabia [10] and after that, the virus exhibited outbreaks in several regions of the world, particularly saudi arabia and the republic of korea [10, 11] . additionally, the complexity of mers-cov and its diagnosis of infection have been acknowledged in the literature [12, 13] . according to who, 2160 laboratory-confirmed cases of mers-cov were reported at the end of january 2018, including 773 associated deaths (case-fatality rate: 35.8%) that were reported globally [14] . the majority of these cases were reported in saudi arabia. this is a time series analysis using publicly reported mers-cov monthly global cases. the who receives confirmed mers-cov cases from countries across the world. to date, new cases continue to be identified and reported to the who, specifically from the middle east region. these data are available at http:// www.who.int/csr/don/archive/disease/coronavirus_infections/en/. the latest report included one case from malaysia on 8 january 2018. we used time series of mers-cov cases reported between january 2015 and january 2018, where who began using standard case report. a research assistant retrieved data from who webpage and reviewed for quality by the second study author. the data retrieved include patient and clinical data such as age, gender, healthcare worker, comorbidity, the source of infection and geographical regions. the main outcome was the number of cases reported on a monthly basis from january 2015 to january 2018. the analysis was performed using stata 12 (stata corp., texas, usa) and microsoft excel 10. using the classical multiplicative time series model, we decomposed the original series into seasonal, irregular and trend components and examined their effects. additionally, we identified patterns, smoothened series, generated predictions and employed forecasting techniques based on linear regression. finally, we assessed contributions of various risk factors in mers-cov cases over time. p-values <0.05 were considered statistically significant. figure 1 shows that although cases of mers-cov are decreasing in the range period selected, the series exhibits peaks and spikes. figure 1 also reveals a negative trend of mers-cov series from january 2015 to january 2018. we investigated direction and significance of this trend, as well as stationarity of the series. while a unit root test for nonstationarity confirmed the mers-cov cases had a negative and statistically significant trend, the series was found to be stationary. the negative and statistically significant trend was also confirmed by subsequent regressions. we collapsed the monthly series into quarters and then smoothened out the series using a centred moving average in order to understand underlying growth component. we assumed the classical multiplicative time series model by decomposing original series into seasonal, irregular, and trend components ( table 1) our decomposition of the mers-cov cases shows that in 2016q2, the seasonality and irregularity components of the series were 8% below the baseline (the centred moving average). decomposing further, seasonality component was 11% above the baseline in 2016q2, while it was 41% below the baseline in 2016q4. our analysis also shows the de-seasonalised series of the original mers-cov by removing seasonality and irregularity components. using a linear regression, we then estimated the effect of time on the deseasonalised series to capture the underlying growth or trend component using a linear regression in order to make predictions. since the last available data was in january 2018, we also made a forecast of three more quarters (2018q2, q3, q4), an additional 9 months into the future. the forecast of the series revealed that mers-cov cases would approach zero by end of 2018 or beginning of 2019, making further extrapolation into the horizon infeasible. figure 2 shows decomposition of mers-cov series of the original series, centred moving average (smoothed series), forecast omar b. da'ar and anwar e. ahmed and linear forecast. together, after accounting for irregular and trend components of the series, seasonality was found to range between 41% below the baseline i.e. the centred moving average (of order 4) in some quarters and 11% above the baseline in other quarters. the average seasonality component was found to be 14% below the baseline. however, regression estimation revealed that, unlike the trend component, seasonality was not statistically significant, a fact also backed by our statistical test, which showed the monthly global mers-cov cases series were stationary. regressions table 2 shows results of the effect of trend and seasonal on mers-cov cases. we compared the seasonal dummy variables, interpreted by comparing them with quarter 4 (q4) (base season) while holding time constant. time, in this analysis, was represented by successive months and was interpreted as the effect of the linear trend on mers-cov cases over time, holding the effect of the seasons constant. the regression results revealed that after accounting for the trend, mers-cov cases quarter 1 (q1) each year averaged about 14 cases more than q4 cases, although the effect was not found to be statistically significant. similarly, after adjusting for the trend, cases in quarter 2 (q2) averaged around 16 cases more than q4 cases. quarter 3 (q3) cases averaged 19 cases more than q4 cases after accounting for the trend component time. it is important to note that the effects of seasonality were not statistically significant. however, what was revealed statistically significant is the negative trend. consistent with the negative trend shown in figure 1 , the regression results revealed that each additional quarter registered approximately an average decrease of one case, after adjusting for the season. in other words, the mers-cov cases decreased, on average, by four (4 × −0.8667627) per quarter year to year. the model as a whole appears to suggest statistically significant predictive capacity. we analysed fluctuations of reported mers-cov cases from period to another by graphing the residuals (generated from the regression of trend and seasonality on mers-cov cases) against time, as is the convention with time series analysis. the results indicated no clear patterns, suggesting that correlated errors are not a problem with this model. other diagnostic tests also revealed neither violation of the classical linear assumptions nor correlation between the reported mers-cov cases in each month with cases reported in earlier months. we further examined the effects of various risk factors of mers-cov cases such as camel contact, healthcare worker contact, exposure, gender and region. table 3 shows regression that adjusts for these factors. the results reveal camel contact, saudi table 3 ). the model in this estimation also appears to suggest statistically significant predictive capacity. using linear time series models and their application to the modelling and prediction of the globally reported mers-cov data, the present study identified trends, analysed seasonality, predicted and forecast evolution of mers-cov cases and assessed the contribution of various risk factors. the decomposition of the time series of mers-cov cases into trend and seasonality components and making predictions have not hitherto been studied in the context of mers-cov pandemic. in this study, we set out to understand the dynamics of its growth over time. the results of our time series analysis of globally reported mers-cov cases suggest a significant negative trend that is forecast to be eradicated in the near future unless something unexpected happens. our study showed that although seasonality oscillated below or above the baseline i.e. the centred moving average (of order 4) over time, the average seasonality component was found to be 14% below the baseline. even then, our analysis showed that, unlike the trend component, seasonality did not affect the series over time. many risk factors are associated with mers-cov cases, mortalities, or complications. our results indicate those aged under 30 years (reference category) are less likely to be a mers-cov case than those aged over 30, consistent with several studies that associated mers-cov with elderly patients [15, 16] . surprisingly, comorbidity did not show a statistically significant contribution to mers-cov cases. however, there are studies that showed mers-cov cases were associated with patients with comorbidities [17] [18] [19] . a recent systematic study, for example, suggests the prevalence of comorbidities of mers-cov cases, diabetes, hypertension and cardiac diseases [20] . while our analysis suggests males contribute to the global reported mers-cov cases, gender was reported to have a mixed effect on mers-cov mortalities in the literature. some studies showed men as high risk [7, 18] and mers-cov infects more males than females [5, [21] [22] [23] . other literature indicated that the frequency of deaths was less in men [24] . the literature showed that mers-cov can be spread via human-human [25] [26] [27] [28] [29] , or healthcare facilities [23, [30] [31] [32] . other studies revealed animal to human [33, 34] as the primary culprit of mers-cov virus transmission. specifically, the literature showed camels act as a direct source of human mers-cov infection [31, 35] , while healthcare workers were reported to be at higher risk [7, 16, 19, 24] . the results of our study in this regard were mixed. while our study indicated that the effect of camel cases on overall mers-cov reported cases are positive and significant, the contribution of healthcare workers was not. our analysis also showed evidence of geographical contributions to mers-cov cases such as saudi arabia and greater middle east compared with south korea. this can be seen as somewhat consistent with earlier studies that demonstrated a link between mortality associated with mers-cov and geography [15] . this finding is also intuitive in that saudi arabia and middle east, in general, remain the global epicentre of mers-cov, hence the name. the contribution of our study is that it adduces empirical evidence by making inferences and predictions based on the globally reported cases of mers-cov and extrapolated into the future by forecasting the trend. unlike previous studies that descriptively analysed seasonality patterns of mers-cov and influenza in the middle east [2] , our study presents statistically significant results of trends of global mers-cov cases, consistent with regularities underlying the empirical dynamics and classical time series analysis. however, there are limitations of this study. first, the data used for this study comprised 37 months (january 2015 to january 2018). while this was just enough for several years' worth of monthly observations to appropriately model seasonality, time series analysis can be sensitive to the number of observations. hence, sufficiently large number of observations might have provided a better fit and results. additionally, the analysis utilised who open source globally reported data, which may lack harmonisation from the various country sources. this study contributes to the time series analysis of mers-cov literature. in particular, our analysis of trends and seasonality components the series, the prediction based on the globally reported cases of mers-cov and extrapolation into the future by forecasting the trend is envisaged to help in understanding the dynamics of the reported cases over time. the study findings suggest a significant negative trend of the monthly and quarterly data from 2015 to 2018. however, a further extrapolation into the future reveals that the mers-cov cases are forecast to be zero by end 2018 or beginning of 2019 unless something unexpected happens. seasonality component of the series oscillated below or above the baseline, i.e. the centred moving average but did not affect the series over time. the results demonstrated that camel contact, exposure, gender, age and geography/region significantly contributed to the overall global reported mers-cov cases. the findings may well inform healthcare practitioners and policymakers about the underlying dynamics that produced the globally reported mers-cov cases. data. these dataset used and/or analysed are available at http://www.who.int/ csr/don/archive/disease/coronavirus_infections/en/. global research trends of middle east respiratory syndrome coronavirus: a bibliometric analysis differences in the seasonality of mers-cov and influenza in the middle east the predictors of 3-and 30-day mortality in 660 mers-cov patients development of a risk-prediction model for middle east respiratory syndrome coronavirus infection in dialysis patients epidemiological, demographic, and clinical characteristics of 47 cases of middle east respiratory syndrome coronavirus disease from saudi arabia: a descriptive study mers-cov infection in humans is associated with a pro-inflammatory th1 and th17 cytokine profile middle east respiratory syndrome coronavirus in al-madinah city, saudi arabia: demographic, clinical and survival data severe acute respiratory syndrome vs. the middle east respiratory syndrome viral shedding and antibody response in 37 patients with middle east respiratory syndrome coronavirus infection middle east respiratory syndrome coronavirus (mers-cov). fact sheet middle east respiratory syndrome corona virus alert verification in mers-cov diagnosis: an update comparative evaluation of three homogenization methods for isolating middle east respiratory syndrome coronavirus nucleic acids from sputum samples for real-time reverse transcription pcr epidemic and pandemic-prone diseases. mers situation update estimating survival rates in mers-cov patients 14 and 45 days after experiencing symptoms and determining the differences in survival rates by demographic data, disease characteristics and regions: a worldwide study building predictive models for mers-cov infections using data mining techniques clinical presentation and outcomes of middle east respiratory syndrome in the republic of korea risk factors for severity and mortality in patients with mers-cov: analysis of publicly available data from saudi arabia risks of death and severe disease in patients with middle east respiratory syndrome coronavirus prevalence of comorbidities in the middle east respiratory syndrome coronavirus (mers-cov): a systematic review and meta-analysis state of knowledge and data gaps of middle east respiratory syndrome coronavirus (mers-cov) in humans clinical course and outcomes of critically ill patients with middle east respiratory syndrome coronavirus infection hospital outbreak of middle east respiratory syndrome coronavirus treatment outcomes for patients with middle eastern respiratory syndrome coronavirus (mers cov) infection at a coronavirus referral center in the kingdom of saudi arabia evidence of person-to-person transmission within a family cluster of novel coronavirus infections the epidemiology of middle east respiratory syndrome coronavirus in the kingdom of saudi arabia clinical features and virological analysis of a case of middle east respiratory syndrome coronavirus infection an observational, laboratory-based study of outbreaks of middle east respiratory syndrome coronavirus in jeddah and riyadh, kingdom of saudi arabia middle east respiratory syndrome coronavirus infection: virus-host cell interactions and implications on pathogenesis hospital-associated middle east respiratory syndrome coronavirus infections hospital-associated middle east respiratory syndrome coronavirus infections mers-cov outbreak in jeddah-a link to health care facilities human-dromedary camel interactions and the risk of acquiring zoonotic middle east respiratory syndrome coronavirus infection early identification of pneumonia patients at increased risk of middle east respiratory syndrome coronavirus infection in saudi arabia human infection with mers coronavirus after exposure to infected camels, saudi arabia acknowledgements. the authors are grateful for the collegiality and research support at the college of public health and health informatics, king saud bin abdulaziz university for health sciences. the information and opinions contained in this work do not necessarily reflect the views or policy of these institutions. this research is supported by king abdullah international medical research centre (kaimrc), king saud bin abdulaziz university for health sciences, national guard health affairs, riyadh, saudi arabia.author contributions. obd analysed the data and wrote manuscript. aea retrieved the data from the who /registry website. aea reviewed analysis and manuscript. all authors approved final manuscript for submission. ethical standards. not applicable. key: cord-321260-oi37dfsp authors: ahmed, anwar e. title: estimating survival rates in mers-cov patients 14 and 45 days after experiencing symptoms and determining the differences in survival rates by demographic data, disease characteristics and regions: a worldwide study date: 2017-12-22 journal: epidemiol infect doi: 10.1017/s095026881700293x sha: doc_id: 321260 cord_uid: oi37dfsp although middle east respiratory syndrome coronavirus (mers-cov) has a recorded 5 years of circulation in 27 countries worldwide, there is no international study to assess whether there is variation in mortality by region. neither has there been a comprehensive study detailing how the disease characteristics of mers-cov influence mortality in patients presenting symptoms. this study aimed to assess how region, patient and disease characteristics influence 14and 45-day mortality in mers patients. the author utilised publically available data on mers-cov. the study included 883 mers patients reported between 5 january 2015 and 10 march 2017. data on patient and disease characteristics were collected. the mean age at mers-cov diagnosis was 54.3 years: 69.1% were male, and 86.7% of the cases were reported from saudi arabia. about 40% of mers patients studied were over the age of 60. the study estimated 14and 45-day survival rates after initial onset of symptoms: 83.67% and 65.9%, respectively. saudi arabian mers patients exhibited 4.1 and 5.0 times higher 14-day (adjusted hazard risk (ahr) = 4.1; 95% confidence interval (ci) 1.012–16.921) and 45-day (ahr = 5.0; 95% ci 1.856–13.581) mortality risk compared with mers patients in the republic of korea or other countries. similarly, middle eastern mers patients showed 5.3 and 4.1 times higher 14-day (ahr = 5.3; 95% ci 1.070–25.902) and 45-day (ahr = 4.1; 95% ci 1.288–113.076) mortality risk compared with mers patients in the republic of korea or other countries. the results demonstrated a link between mortality and geography, disease and patient factors such as regions, symptoms, source of infections, underlying medical conditions, modes of transmission, non-healthcare workers and those of older age. educational programmes, access to healthcare and early diagnosis could be implemented as modifiable factors to reduce the higher mortality rates in mers patients. middle east respiratory syndrome coronavirus (mers-cov) is an international public health challenge with considerable differences in methods of reporting death-related within and between countries [1] [2] [3] [4] [5] [6] [7] . the case fatality rate accounted for 30-63% in saudi arabia [1] [2] [3] [4] and 20-63% in the republic of korea [5] [6] [7] . recently, several epidemiological investigations have been published that assessed the factors associated with mortality in mers patients [8] [9] [10] [11] [12] [13] . however, the findings of most of these studies either represent a specific population group such as intensive care unit [8] , a single medical centre [9] or outbreak infection in the republic of korea [10, 11] . one study analysed publicly available data through the saudi ministry of health webpage to identify the factors associated with mortality in saudi arabian patients [12] . one limitation of this study was lack of multivariate risk modelling. rivers et al. [13] used a multivariate poisson regression model to identify factors associated with a high incidence of mortality, but their study excluded all cases (186) from the republic of korea. a recent study by chen et al. compared the risk factors and disease characteristics in south korean and saudi arabian populations, but did not assess the survival rate in these cases [14] . the authors of all previously published studies of mers-cov were not able to assess survival rate in mers cases from various countries that were reported to the world health organization (who). although the mentioned studies provided valuable information in terms of mortality risk in mers patients, a common limitation was noted in these previous studies: no global data were used, such as the associations between different geographical regions and mortality. the findings in these reports were made using data on specific populations or regions. per the author's knowledge, at this time there is no large cross-country comparison study of mers-cov-related mortality, and it is not yet clear whether the mortality rate of mers-cov can be identified by countries and regions. it is important to understand and compare mers-cov patients according to their clinical outcomes between multicountries. such a study may provide a useful multi-country work to reduce the mortality in the mers-cov population worldwide. the current study may provide multivariate risk models to identify patients at high risk of death in order to manage mers-cov patient clinical outcomes. this may improve public health plans by establishing effective programmes to prevent serious outcomes. these programmes can be addressed by the healthcare system or public health policies. the goal of this study was to estimate and compare the variations in the risk-standardised mortality rates by regions and by patients' characteristics among different countries across the world, specifically comparing regions that had the largest number of mers-cov cases. the author hypothesised that people of older age, with underlying medical conditions and from saudi arabia or other middle east countries are at high risk of death related to mers-cov. the study utilised a publicly available mers-cov database of case reports retrieved from the who: http://www.who.int/ csr/don/archive/disease/coronavirus_infections/en/. the who receives situational reports of confirmed mers-cov cases from all countries across the globe. the who provides a routine update through its website on new cases, deaths and current developments. prior to 26 january 2017, all reports were presented in a narrative format, describing case by case the disease characteristics of each patient. the author and research assistant accessed each of these narrative reports, and converted the patient information and disease characteristics into a line listing for analysis. beginning with the reporting period of 2-7 january, who enhanced their reporting practices, including an excel document with patient and disease characteristic data which can be easily analysed. the latest line listing update of mers cases was on april 2017 in a report from qatar. globally, the who report on 4 april 2017 indicated that the overall case fatality rate was 35.6% (690 deaths of 1936 cumulative laboratory-confirmed mers cases) in 27 countries (who, 2017). the author did not include all mers cases reported to the who because the early reports of mers used unstandardised case presentation that lacked important details. this study included mers cases reported to the who between 5 january 2015 and 10 march 2017. another reason for choosing this period to study is that in 2015, the who started reporting cases by the country where the confirmed case has occurred. the author performed quality control checks to detect invalid or missing data. the author excluded some cases from the republic of korea that were reported between 12 june and 21 july 2015, as no patient case report was used. they reported a summary of the updated cases (total cases and deaths) instead of case-by-case details. case #3 reported on 23 march 2016 has been excluded from saudi arabia due to the wrong date of symptom onset stated in his case report. the total cases included in the analysis were 883. the author obtained information on age, gender, date of notification, date of onset of mers symptoms, date of outcome or death, whether a patient or healthcare worker, symptomatic, underlying medical conditions, source of infection and the reported country. in order to assess the impact of older age on mortality among mers patients, patient age was classified into four groups: age <30, 30-59, 60-65 and >65 years. the data retrieved were from 14 countries. the author classified countries into three regions according to the geographical location and the number of cases: (1) saudi arabia 766 (86.7%), (2) middle east, but excluding saudi arabia 44 (5%), and (3) republic of korea or other countries 73 (8.3%). the republic of korea 65 (7.4) and other countries 8 (0.9%) were combined due to the small number of cases in the latter group. saudi arabia was analysed separately from the middle east countries because saudi arabia has a unique situation and recorded the largest number of mers cases. two primary end points were assessed: 14-and 45-day mortality related to mersafter developing symptoms. the study author estimated the survival rates using the time of symptom onset to outcome. mers-cov has an incubation period of 2-14 days, by which time symptoms usually occur. survival rates at the 14-day mark are medically significant as they demonstrate survival at the point at which patients most typically experience statistical analyses were performed with sas version 9.4 software. data from 883 mers-cov patients who reported to the who between 5 january 2015 and 10 march 2017 were retrieved and analysed. descriptive statistics were used to describe the study population (table 1) . unadjusted cox proportional hazards models (cphms) were used to estimate hazard risk (hr) and a 95% confidence interval (95% ci) for 14-and 45-day mortality (tables 2 and 3 ). multivariable cphms were used to estimate adjusted hazard risk (ahr) and 95% ci for 14-and 45-day mortality adjusting for gender, age, region, whether healthcare worker, having comorbidity, being asymptomatic and the source of infection (tables 2 and 3 ). the log-rank tests were used to compare survival curves for demographic and disease characteristics (figs 1-4) . the 14-and 45-day survival rates were estimated using the kaplan-meier estimator. a total of 883 patients with confirmed mers-cov infection were included in the analysis. the characteristics of the population can be found in table 2 illustrates unadjusted and adjusted 14-day hazard risks for all-cause mortality. the unadjusted 14-day analysis shows that age groups 60-65 and >65 years, saudi arabia, the study was conducted to estimate the survival rates in mers patients, specifically 14 and 45 days after experiencing mers-cov symptoms and to determine whether there is a significant epidemiology & infection difference in survival rates by demographic data, disease characteristics and regions using global mers data that were reported to who. the author used the cphms and the kaplan-meier estimator to calculate survival rates. according to the data, the estimated overall survival rate was 63.4% (95% ci 60.15-66.60%). the study estimated 14-and 45-day survival rates were 83.67% (95% ci 81.09-86.07%) and 65.9% (95% ci 62.68-69.04%), respectively. the author calculated region-specific 14-and 45-day survival rates (figs 1 and 3) . for the mers patients studied, the republic of korea or other countries (95.83% and 91.67%) had much higher 14-and 45-day survival rates than the middle east (84.09% and 75.00%) and saudi arabia (82.51% and 62.92%), respectively. the differences in the region-specific survival rates remain significant after accounting for patient and disease factors such as gender, age, comorbidity, symptoms, healthcare worker and source of infection. for instance, the hazard of death on 45-day post-symptoms was 4.1 and 5.0 times higher in the middle east and saudi arabia compared with the republic of korea or other countries, respectively. the disparities in survival rates between regions are probably explained by the delay in reporting mers symptoms, and consequently delay its diagnosis [15] . the median time from date of onset of symptoms to death was lower in the republic of korea or other countries than in the middle east and saudi arabia. more research studies are required to assess whether diagnostic delay [15] and disease characteristics can explain the differences in survival rate by geographical regions. those aged 60 years and older represent a large portion of the mers population (39.5%) and they predominantly contribute a high portion of deaths as well. the findings of this study showed that 14-and 45-day survival rates tend to rise with increasing age (figs 1 and 3) and were similar in women compared with men. the estimate of the 45-day survival rate in patients older than 65 years was 44.86%, in patients aged 60-65 years it was 60.38%, in patients aged 30-50 years it was 74.27% and in young patients aged 29 years or less it was 87.78%. these results support various epidemiological cohorts from saudi arabia and the republic of korea, which stated that older age is a risk factor for death in mers patients [8] [9] [10] [11] [12] [13] [14] . the high death rate among the older age group could be attributable to the greater number of patients with comorbidities in this group. the prevalence of comorbidities increases with age: 28.6% in the young age group <30 years, 62.7% in 30-59 years, 93% in 60-65 years and 97% in older than 65 years. thus, comorbidities may be associated with both age and mortality. several studies [10] [11] [12] [13] , including the current study, suggest that the presence of underlying conditions was associated with an increase in the hazard of death in mers patients. more details on underlying conditions are needed to estimate underlying condition-specific survival rates and would be useful to identify which underlying conditions were associated with the lower survival rates. moreover, prevention and disease management strategies should be assessed as interventions in mers patients with underlying conditions to reduce the mortality rate in this group. in concordance with other studies [9, 13] , being a nonhealthcare worker was associated with lower survival rates (figs 2 and 4) . the study included 116 (13.1%) who were healthcare workers, of which five died. the higher survival rates within the healthcare workers group could be attributed to educating healthcare workers on preparedness, access to healthcare or following proper infection control standards. an explanation for lower survival rates in the non-healthcare workers group is the large gaps in public awareness of the clinical symptoms of mers-cov [16, 17] . public health and health system interventions are needed to reduce the spread of the mers by raising public awareness in identifying the clinical symptoms and by early screening and diagnosis. this could reduce the high rate of mortality in nonhealthcare workers. patients who had acquired the infection from camels, hospitals and unknown sources of infections exhibited 2.5, 3.6 and 3 times higher mortality risk compared with patients who had acquired the infection from a household member, respectively (figs 2 and 4 ). the healthcare system may use this information to properly develop a support care plan to improve patients' outcomes. anwar e. ahmed this study has several limitations. the study used available data on the public source with no details on the underlying conditions. this information could be important to identify which underlying condition is associated with death. patient condition and clinical outcomes may not be final as data are updated routinely. another potential confounding factor was not available, such as access to healthcare, and it may be considered a modifiable factor to account for. finally, an inverse-probability kaplan-meier curve may also be appropriate to model this dataset, as it may provide perspective into the data and a visual way to show adjusted survival rates. despite these limitations, the study author was able to use the estimated 14-and 45-day survival rates, measuring the time from symptom onset to outcome. to date, no study has provided survival estimates and links to regions. the study provided information on region-specific 14and 45-day survival rates, which are found to account for the differences in mortality. the large sample size used was also the main strength of this study. future study could investigate diagnostic delay, which can be defined by the difference between date of symptom onset and the time of diagnosis. this may reduce poor outcomes. the study estimated 14-and 45-day survival rates after initial onset of symptoms: 83.67% and 65.9%, respectively. the results demonstrated a link between mortality and geography, disease and patient factors such as regions, symptoms, source of infections, underlying medical conditions, modes of transmission, non-healthcare workers and older age. educational programmes, access to healthcare and early diagnosis could be implemented as modifiable factors to reduce the higher mortality rates in mers patients. additional files. none. ethical approval and consent to participate. not applicable. consent for publication. the author read and approved the final manuscript. availability of supporting data. the data used for the analysis can be obtained from the study author. declaration of interest. none declared. association of higher mers-cov virus load with severe disease and death, saudi arabia treatment outcomes for patients with middle eastern respiratory syndrome coronavirus (mers cov) infection at a coronavirus referral center in the kingdom of saudi arabia the predictors of 3-and 30-day mortality in 660 mers-cov patients predictors of mers-cov infection: a large case control study of patients presenting with ili at a mers-cov referral hospital in saudi arabia estimating the risk of middle east respiratory syndrome (mers) death during the course of the outbreak in the republic of korea high fatality rates and associated factors in two hospital outbreaks of mers in daejeon, the republic of korea middle east respiratory syndrome coronavirus (mers-cov) outbreak in south korea, 2015: epidemiology, characteristics and public health implications presentation and outcome of middle east respiratory syndrome in saudi intensive care unit patients clinical aspects and outcomes of 70 patients with middle east respiratory syndrome coronavirus infection: a singlecenter experience in saudi arabia real-time characterization of risks of death associated with the middle east respiratory syndrome (mers) in the republic of korea mortality risk factors for middle east respiratory syndrome outbreak, south korea risk factors for severity and mortality in patients with mers-cov: analysis of publicly available data from saudi arabia risks of death and severe disease in patients with middle east respiratory syndrome coronavirus comparative epidemiology of middle east respiratory syndrome coronavirus (mers-cov) in saudi arabia and south korea diagnostic delays in 537 symptomatic cases of mers-cov infection in saudi arabia epidemiological, demographic, and clinical characteristics of 47 cases of middle east respiratory syndrome coronavirus disease from saudi arabia: a descriptive study middle east respiratory syndrome coronavirus (mers-cov): prevention in travelers acknowledgements. the author acknowledges the who for making mers-cov data publicly available. no funding was provided for this study.authors' information. college of public health and health informatics, king saud bin abdulaziz university for health sciences, riyadh, saudi arabia. key: cord-305264-0uhabgsr authors: weng, c-h.; saal, a.; butt, w. w-w.; bica, n.; fisher, j. q.; tao, j.; chan, p. a. title: bacillus calmette–guérin vaccination and clinical characteristics and outcomes of covid-19 in rhode island, united states: a cohort study date: 2020-07-09 journal: epidemiol infect doi: 10.1017/s0950268820001569 sha: doc_id: 305264 cord_uid: 0uhabgsr coronavirus disease 2019 (covid-19) has resulted in a global pandemic, and there is limited data on effective therapies. bacillus calmette–guérin (bcg) vaccine, a live-attenuated strain derived from an isolate of mycobacterium bovis and originally designed to prevent tuberculosis, has shown some efficacy against infection with unrelated pathogens. in this study, we reviewed 120 consecutive adult patients (≥18 years old) with covid-19 at a major federally qualified health centre in rhode island, united states from 19 march to 29 april 2020. median age was 39.5 years (interquartile range, 27.0–50.0), 30% were male and 87.5% were latino/hispanics. eighty-two (68.3%) patients had bcg vaccination. individuals with bcg vaccination were less likely to require hospital admission during the disease course (3.7% vs. 15.8%, p = 0.019). this association remained unchanged after adjusting for demographics and comorbidities (p = 0.017) using multivariate regression analysis. the finding from our study suggests the potential of bcg in preventing more severe covid-19. severe acute respiratory syndrome coronavirus 2 (sars-cov-2) is the cause of coronavirus disease 2019 (covid-19) and has resulted in a global pandemic. there is limited data on effective therapies. the bacillus calmette-guérin (bcg) vaccine, a live-attenuated strain derived from an isolate of mycobacterium bovis and originally designed to prevent tuberculosis, has shown some efficacy against infection with unrelated pathogens [1] . a recent study suggested deaths due to covid-19 were significantly lower in bcg-vaccinated countries when compared with bcg-non-vaccinated countries [2] . it is important for future prevention efforts to investigate this potential effect to see if bcg vaccine confers protection against more severe covid-19. to determine if bcg vaccination provided protection from covid-19, we reviewed a predominately latino/hispanic population receiving care at the major federally qualified health centre (fqhc) in providence, rhode island, united states. ninety per cent of households in this fqhc were under the 200% federal poverty level (fpl) and resided in providence. between 19 march and 29 april 2020, data on 120 (77.4%) out of 155 consecutive adult patients (≥18 years old) who were sars-cov-2 positive were available and patients were reviewed through 14 days. we characterised patients by demographics, immunisation status, symptoms during disease course, hospitalisation and comorbid disease. the above information was self-reported and through medical record review. bcg vaccination status was determined by review of clinical charts. all the patients with mild symptoms were advised to isolate at home. patients experiencing severe symptoms were referred to the hospitals in the same geographic areas by our triage team and clinicians using standard protocols. the clinicians in the emergency rooms were unaware of the patients' bcg status. patients were admitted if they showed significant hypoxia which may have required more aggressive oxygen support or if they presented with signs of haemodynamic instability. we report numbers (percentages) for binary/categorical variables and medians (interquartile ranges, iqr) for continuous variables. χ 2 tests and wilcoxon rank-sum tests were applied to compare the statistical significances. a multivariate regression model adjusting age, sex, ethnicity, cigarette smoking history and comorbidities was applied to examine the outcome. all analyses were run using stata 13.1 (statacorp, college station, tx, usa). the providence community health centers review committee approved the project. among the 120 patients, 82 (68.3%) had bcg vaccination. median age was 39.5 years (iqr, 27.0-50.0). the bcg-vaccinated population was on average 10 years older than the non-bcg-vaccinated population (median age 41.0 vs. 31.0 years, respectively, p = 0.390). thirty per cent were male and 87.5% were latino/hispanics (table 1 ). compared to those without bcg vaccination, patients with bcg vaccination were more likely to experience myalgia during the disease course (74.4% vs. 50.0%, p = 0.008). there were no significant differences between the two groups in experiencing cough (73.3%), shortness of breath (26.7%), nasal congestion/rhinorrhoea (51.7%), fever (61.7%), headache (60.0%), sore throat (38.3%), vomiting/diarrhoea (39.2%) or loss of smell/taste (60.8%). compared to a large case series from china [3] , our overall patient population experienced symptoms at a percentage similar to a recent study from washington state, united states [4] , with more patients experiencing myalgia, headache and loss of smell/taste. the difference could reflect geographic variation or differential reporting. covid-19 patients with bcg vaccination were less likely to be hospitalised during the disease course (3.7% vs. 15.8%, p = 0.019). this association remained unchanged after adjusting for demographics and comorbidities (p = 0.017) using multivariate regression analysis. one patient without bcg vaccination died. the comorbidities between the two groups showed no significant differences in chronic diseases including hypertension (24.2%), diabetes (12.5%), chronic kidney disease (3.3%) and being immunocompromised (0.8%). a higher percentage of patients without bcg had a history of chronic obstructive pulmonary disease (copd)/asthma, however, a recent study found the history of copd was not associated with the risk of hospitalisation among covid-19 patients [5] . among those who were hospitalised, none had a history of cigarette smoking and there was no significant difference between the two groups in copd/asthma (p = 0.257). comparing the comorbidities among the hospitalised patients between the non-bcg-and bcg-vaccinated patients, no statistical differences were found in hypertension (83.3% vs. 100% respectively, p = 0.453), diabetes (33.3% vs. 66.7%, p = 0.343), copd/asthma (33.3% vs. 0, p = 0.257), morbid obesity (33.3% vs. 33.3%, p = 1.000), chronic kidney disease (16.7% vs. 0, p = 0.453), none of the hospitalised patients had histories of liver cirrhosis or were immunocompromised. in this study, patients with bcg vaccination were more likely to experience myalgia and less likely to require hospital admission. myalgias may be related to the release of inflammatory mediators, such as interleukins (ils) [6] . bcg is known to elicit non-specific immune effects through the induction of the innate immune responses and the enhanced production of il-1β [1] . this may present as myalgias and help the body fight the infection. recent ecological studies comparing countries with and without universal bcg vaccination policies found that bcg vaccination appears to significantly reduce mortality associated with covid-19 [7] and mandatory bcg vaccination was associated with a flattening of the curve in the spread of covid-19 [8] . these studies suggest a long-lasting protection conferred by childhood bcg vaccination against covid-19. this duration of protection may persist for several years, as one study examining bcg vaccine protection against tuberculosis found a 50−60-year duration of protection [9] . a recent population-based study examining the cohort of israeli adults aged 35−41 years found that the bcg vaccine may not reduce the likelihood of acquiring sars-cov-2 (difference, 1.3%; 95% ci −0.3% to 2.9%; p = 0.09) [10] . however, the lower hospitalisation rate among bcg-vaccinated patients from our prospective cohort study suggests the potential of bcg in preventing more severe covid-19 among those who acquired sars-cov-2. limitations to this study included a small sample size, short study time frame, unknown bcg strain each patient received, unknown bcg booster status, a preponderance of female patients, and a predominately latino/hispanic population. future studies are needed to explore the efficacy of bcg vaccination in preventing covid-19 disease progression. concept and design: weng, saal, chan. acquisition, analysis, or interpretation of data: all authors. drafting of the manuscript: weng, chan. critical revision of the manuscript for important intellectual content: weng, butt, chan administrative, technical, or material support: weng, saal. supervision: weng, chan. financial support. this research received no specific grant from any funding agency non-specific effects of bcg vaccine on viral infections is bcg vaccination effecting the spread and severity of covid-19? allergy clinical characteristics of coronavirus disease 2019 in china symptom screening at illness onset of health care personnel with sars-cov-2 infection in king county factors associated with hospital admission and critical illness among 5279 people with coronavirus disease 2019 in new york city: prospective cohort study immune system involvement in specific pain conditions correlation between universal bcg vaccination policy and reduced morbidity and mortality for covid-19: an epidemiological study mandated bacillus calmette-guérin (bcg) vaccination predicts flattened curves for the spread of covid-19 long-term efficacy of bcg vaccine in american indians and alaska natives: a 60-year follow-up study sars-cov-2 rates in bcg-vaccinated and unvaccinated young adults acknowledgements. special thanks to ms. diane chaca for coordinating patient care.author contributions. ethical standards. the study was approved by the providence community health centers review committee. the authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the helsinki declaration of 1975, as revised in 2008.data availability statement. the data that support the findings of this study are available on request from the corresponding author, c-h w. the data are not publicly available due to their containing information that could compromise the privacy of research participants. key: cord-313415-5qrpucr4 authors: lai, rongtao; chen, erzhen; gao, weiyi; cheng, chengwei; xie, qing title: sentinel surveillance strategies for early detection of coronavirus disease in fever clinics: experience from china date: 2020-08-25 journal: epidemiol infect doi: 10.1017/s0950268820001892 sha: doc_id: 313415 cord_uid: 5qrpucr4 sentinel surveillance system plays a key role in screening and monitoring emerging and acute infectious diseases in order to identify the suspected cases in time. during sars period in 2003, fever clinics emerged in many cities in mainland china with the purpose to screen the suspected sars patients and to transfer the confirmed cases to designated hospitals for professional management. shanghai city has reserved the fever clinics and the designated hospitals since then. hence, clinicians in the front line are able to respond quickly to the emerging covid-19 outbreak with their accumulated knowledge and experiences from the past. one hundred seventeen fever clinics distributed in various district areas in shanghai have played a vital ‘sentinel’ role to fight against the covid-19 epidemic. most of suspected patients were identified in fever clinics and thereafter among these suspected patients the covid-19 cases were confirmed and were isolated quickly to avoid the spread. we would like to share the sentinel roadmap for screening and diagnosis of covid-19 to medical healthcare workers around the world, especially countries who are facing great challenges to cope with covid-19 and meanwhile with limited medical resources. these sentinel surveillance strategies will certainly provide insight into the early detection and timely isolation of suspected cases from the others. sentinel surveillance system plays a key role in screening and monitoring emerging and acute infectious diseases in order to identify the suspected cases in time. during sars period in 2003, fever clinics emerged in many cities in mainland china with the purpose to screen the suspected sars patients and to transfer the confirmed cases to designated hospitals for professional management. shanghai city has reserved the fever clinics and the designated hospitals since then. hence, clinicians in the front line are able to respond quickly to the emerging covid-19 outbreak with their accumulated knowledge and experiences from the past. one hundred seventeen fever clinics distributed in various district areas in shanghai have played a vital 'sentinel' role to fight against the covid-19 epidemic. most of suspected patients were identified in fever clinics and thereafter among these suspected patients the covid-19 cases were confirmed and were isolated quickly to avoid the spread. we would like to share the sentinel roadmap for screening and diagnosis of covid-19 to medical healthcare workers around the world, especially countries who are facing great challenges to cope with covid-19 and meanwhile with limited medical resources. these sentinel surveillance strategies will certainly provide insight into the early detection and timely isolation of suspected cases from the others. the world health organization (who) declared coronavirus disease-19 (covid-19) to be a global pandemic on 12 march 2020 [1] . some countries are struggling with a shortage of medical resources. in late january 2020, 117 fever clinics distributed in various district areas of shanghai have played a vital 'sentinel' role in the fight against the covid-19 epidemic. sentinel surveillance in fever clinics is different from routine surveillance. it is employed for discerning patients with suspected symptoms and signs, for timely isolation, for effectively blocking disease transmission during the early outbreak period before the pathogen has been identified, and for determining effective therapeutic methods; this strategy was used during the severe acute respiratory syndrome (sars) epidemic in 2003 [2] . information of patients with suspicious symptoms and imaging findings in the sentinel surveillance reporting system are reported to the centers for disease control (cdc) to provide an early warning of emergent acute infectious diseases. we aim to share our sentinel surveillance protocol for the screening and diagnosis of covid-19 patients with medical healthcare workers worldwide, especially with those in countries facing challenges in coping with the covid-19 outbreak due to limited medical resources. covid-19, which was first reported in wuhan, china, in december 2019, has spread rapidly [3] . the causative agent, severe acute respiratory syndrome coronavirus 2 (sars-cov-2), can cause a severe respiratory syndrome similar to sars and middle east respiratory syndrome (mers). the transmissibility and mortality of covid-19 are high [3, 4] . a sentinel surveillance strategy was employed in fever clinics to ensure early detection and timely isolation of suspected cases. in the early stage of the outbreak, testing kits, supplies and effective medicines were unavailable. to efficiently and effectively control the epidemic and to minimise its negative economic impact, rapid screening and immediate isolation of suspected covid-19 cases in fever clinics was important. in the early outbreak period, the use of the sentinel surveillance strategy in fever clinics can provide benefits in terms of identifying patients with suspected symptoms, effectively blocking disease transmission, and protecting vulnerable populations. under the sentinel protocol ( fig. 1) , the typical symptoms of covid-19 were fever, sore throat, dry cough and fatigue [5] . however, more than half of the patients are reported to present without fever on admission [6] . therefore, patients with a normal body temperature (oral temperature <37.3°c) but with respiratory symptoms and underlying disease; as well as those with an epidemiological history, including a history of travel to epidemic areas and contact with covid-19 patients are recommended to visit fever clinics. patients with an oral temperature >37.3°c are also advised to visit fever clinics, regardless of their contact history. in fever clinics, the following clinical assessments are performed: evaluation of vital signs, c-reactive protein measurements, complete blood count, testing for influenza a + b viral antigens and lowdose spiral chest computed tomography (ct). the screening results are usually obtained within 30 min. usually, a further 30 min are required for specialists from a multi-disciplinary team (mdt) to re-evaluate the results of lymphocytopaenia and/or suspicious chest ct findings as well as the epidemiological history; this mdt contains specialists from the departments of infectious disease, respiratory disease, emergency medicine and intensive care. if the patient confirms recent contact with a covid-19 case and has an abnormal blood count and/or abnormal chest radiologic findings, the patient is assigned an isolated room, and nasopharyngeal swabs and sputum and blood samples are sent to the shanghai cdc. upon obtaining two negative results of reverse transcription polymerase chain reaction (rt-pcr) for sars-cov-2 within an interval of ≥24 h, the patient is asked to home quarantine in a separate room for at least 14 days, and a rongtao lai et al. telephonic follow-up is arranged. if the symptoms persist and worsen, rapid detection of serum immunoglobulin m (igm) and igg antibodies against sars-cov-2 is necessary. covid-19 should also be considered in patients with community-acquired pneumonia showing no response to standard antibacterial treatment. if the rt-pcr test for sars-cov-2 shows positive results, the patient is transferred to a designated hospital for isolation and further treatment. patients presenting with respiratory symptoms during the 14-day quarantine undertaken due to obvious epidemic history are strongly recommended to visit a fever clinic and undergo rapid tests for the detection of serum igm and igg antibodies against sars-cov-2 or viral nucleic acids. this helps in controlling transmission and in monitoring disease progression and deterioration. during the sars epidemic in 2003, fever clinics emerged in many cities in mainland china for screening suspected sars patients and to transferring confirmed cases to designated hospitals for professional management [2] . shanghai city has learnt to respond to an outbreak, and to reserve the fever clinics and designated hospitals since then. hence, front-line clinicians in fever clinics are able to respond to the emerging covid-19 outbreak owing to their accumulated knowledge and past experiences. in daily life, fever clinics are used for screening influenza cases and seasonal epidemics. if patients tested positive for influenza a/b, they were prescribed oral antivirals and were required to home quarantine for at least 3 days after resolution of the fever. early symptoms of sars, mers, covid-19 and influenza are similar, fever, cough, sore throat and rhinorrhoea or nasal congestion. sars, mers and covid-19 have a longer incubation period than influenza [7] . the proportion of symptomatic patients requiring hospitalisation was higher in covid-19 patients than in influenza patients [8] . the differential diagnosis mainly relies on the detection of viral nucleic acids or serum antibodies. covid-19 concurrent with influenza has also been reported, especially in the epidemic period [9] . an increasing number of countries have been affected by this pandemic worldwide, and the mortality rate is increasing [1] . all affected countries are looking for a balance between protecting peoples' health and minimising the negative economic impact of this pandemic [1] . effective public health measures, such as contact tracing, social distancing and communal surveillance, have been implemented to prevent transmission at a community level [10] . in patients with symptoms like fever, dry cough, sore throat or fatigue, rapid screening, isolation and diagnosis in the sentinel departments, such as fever clinics, is crucial. lymphocytopaenia and bilateral/peripheral ground-glass opacities were considered as early indicators of covid-19 in the early outbreak period. the final diagnosis can be confirmed by real-time rt-pcr assays or by viral gene sequencing. however, resources for diagnosis are important in the early outbreak period, and unfortunately, these resources are still lacking in some countries. to date, there are no effective specific treatments for covid-19. fever clinics play a role in sentinel surveillance in the fight against covid-19. fever clinics are independent of emergency clinics. in the early outbreak period, the clinicians in fever clinics are responsible for rapid screening, identification and isolation. it is important to isolate suspected cases as soon as possible and confirm the diagnosis accordingly. however, this strategy has limitations. with this protocol, the suspected patients of having influenza were required to undergo low-dose chest ct to rule out atypical pulmonary infections, which was not necessary in the past. radiologic features on chest ct are useful in the early period to rule out bacterial aetiologies, although they are limited ability to distinguish viral infections. tests for antigen influenza a + b are also necessary. the short-term mortality rate is fairly high in critical cases. hence, this strategy not only helps in providing timely treatment to affected patients so as to decrease or delay the disease progression but also helps in efficiently isolating suspected cases in a short time period to protect public health. we hope that sentinel screening protocol will serve as an operating protocol for the first-line clinicians. covid-19: towards controlling of a pandemic evaluation of control measures implemented in the severe acute respiratory syndrome outbreak in beijing clinical features of patients infected with 2019 novel coronavirus in wuhan reconstruction of the full transmission dynamics of covid-19 in wuhan epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study clinical characteristics of coronavirus disease 2019 in china incubation period of 2019 novel coronavirus (2019-ncov) infections among travellers from wuhan, china comparing sars-cov-2 with sars-cov and influenza pandemics sars-cov-2 and influenza virus co-infection the l (2020) covid-19: too little, too late? financial support. this study was supported by the national natural science foundation of china (grant no. 81970514), the shanghai municipal key clinical specialty (shslczdzk01103). data availability. data sharing is not applicable to this article as no new data were created or analysed in this study. key: cord-275002-axp2ggbf authors: brandl, m.; selb, r.; seidl-pillmeier, s.; marosevic, d.; buchholz, u; rehmet, s. title: mass gathering events and undetected transmission of sars-cov-2 in vulnerable populations leading to an outbreak with high case fatality ratio in the district of tirschenreuth, germany date: 2020-10-13 journal: epidemiol infect doi: 10.1017/s0950268820002460 sha: doc_id: 275002 cord_uid: axp2ggbf in early march 2020, a covid-19-outbreak occurred in the district of tirschenreuth, germany. the outbreak was characterised by a rapid increase in case numbers and a comparatively high crude case fatality ratio (cfr; 11%). until the beginning of may 2020, 1122 cases were reported in the district. to investigate the outbreak, we analysed surveillance and other data available at the district health department, including data on cases living in care facilities and public health measures applied. furthermore, we compared the number of tests performed in tirschenreuth and in germany as a whole. we interviewed the first 110 cases in order to investigate potential exposures at the beginning of the outbreak. we found that returning ski-travellers from austria and italy and early undetected community transmission likely initiated the outbreak which was then accelerated by bavarian beer festivities. testing of mainly acute cases in the district of tirschenreuth resulted in a higher rate of positive tests compared to the whole of germany. despite adjustment for age, the cfr continued to exceed the german mean which was due to spread to vulnerable populations. strict public health measures likely contributed to control the outbreak by mid-april 2020. in december 2019, a novel betacoronavirus, sars-cov-2, was isolated from hospitalised patients in wuhan, china who were suffering from pneumonia with an unknown aetiology [1] . as of 9 july 2020, there have been more than 11.6 million cases of covid-19 (coronavirus disease 2019), the disease caused by sars-cov-2, reported worldwide [2] . on 10 march 2020, the first case of covid-19 was reported in the district of tirschenreuth (tir; 72 000 inhabitants). within 1 week, the number of cases had increased dramatically. on 18 march 2020, the local administration ordered a lockdown of mitterteich, the most affected town in the district. in the following weeks, a high crude case fatality ratio (cfr) was observed in the district (tir: 11%; germany: 4.3%). the potential role of local beer festivals as super-spreading events gained wide public attention. the local government of the tir district and the state of bavaria invited the robert koch institute (rki) for a joint investigation. together with state and local health departments, we aimed to identify exposures potentially responsible for the transmission during the outbreak and the underlying reasons for the high cfr. in order to investigate reasons for the rapid increase of cases and the magnitude of the outbreak, we used national and local surveillance data on laboratory-confirmed, reported covid-19 cases in tir from 17 february to 11 may 2020. we used the standard case definition of the rki to classify cases 1 . in our study, all persons with sars-cov-2 infection confirmed by rt-pcr were included. we used epidemic curves to analyse cases over time. to this end, we used the date of onset of symptoms, or date of reporting where the onset of symptoms was not available. we calculated the 7-day reproduction number r from notification data by using a moving 7-day average [3] . in order to detect events and exposures that may have led to the outbreak, we looked at the first 110 cases in tir (by date of onset of symptoms) in more detail. these cases reported symptoms of covid-19 before the second, large peak of cases occurred in the course of the outbreak. we gathered additional information on potential exposures of these cases (attendance at large public events, previous contact to known covid-19 cases, travel history, especially skiing holidays in areas with high risk for acquiring in the period 17 february to 8 march 2020. to this end, we used available data from the local health department and conducted semistandardised telephone interviews. the semi-standardised questionnaire was created by the authors in collaboration with the employees of the local health authority in tirschenreuth. local staff conducted the phone interview and data were collected in an excel spreadsheet. multiple answers per question were possible. in the course of the outbreak investigation, unfortunately there was not sufficient time to test the questionnaire in advance. the questionnaire used for the telephone interviews is provided in the supplementary material of this study (s1). to evaluate the measures taken in the course of the outbreak, we analysed cases in the whole tir district and those living in the town of mitterteich. we calculated the date of infection with sars-cov-2 by using the date of symptom onset minus the mean incubation period of 5 days [4] . we smoothed the generated curve of infections/day calculating the 3-day moving average. we used the day with the highest number of cases as the reference (100%). in order to investigate the high crude cfr in tir, we used surveillance data to compare the distribution of factors associated with infection between tir and the rest of germany. the factors included age, underlying medical conditions (cardiovascular diseases, chronic lung diseases, chronic liver diseases, diabetes, cancer and immunosuppression [5] ) and the proportion of vulnerable persons (cases living in care facilities). we used german surveillance data to categorise cases as living in a community facility according to §36 of the german infection protection act. to provide information on outbreaks in elderly care homes, we examined data documented at the local health authority. to determine if the crude cfr was distorted by the underlying age structure of the district, we calculated an age-standardised cfr. we used strata in 10-year-steps. to compare the age of different groups of cases, we used the wilcoxon rank-sum test. all statistical analyses were performed in stata se 15.1. p-values <0.05 were considered statistically significant. we contacted local laboratories to estimate the number of sars-cov-2 tests performed in the period of 7 march 2020 (first laboratory confirmation in tir) and 13 may 2020 and compared the data with those reported for germany [6] . between 10 march and 11 may 2020, 1122 covid-19 cases were notified to the local health authorities, 129 (11%) of whom had died by the time of this analysis (11 may 2020; table 1 ). the epidemic curve is shown in figure 1 . the first laboratoryconfirmed case in the region was notified to the local health authorities on 10 march 2020 with disease onset on 8 march 2020. investigations and interviews with cases revealed that there were sporadic cases with disease onset in february and early march, the first peak with 28 cases occurred on 10 march 2020. a second and even higher peak was observed on 16 march 2020 with 55 cases. the number of cases started to continuously decline at the end of march. at the beginning of may, only sporadic cases were notified. the 7-day reproduction number r declined steadily from 2.7 on 24 march to below 1.0 on 5 april. overall, 80% of cases were symptomatic, especially in the earlier weeks of the outbreak, while in the later stages the proportion of asymptomatic cases increased. due to the magnitude of the outbreak and limited resources during march 2020, mainly close contacts of known covid-19 cases with acute symptoms were tested for sars-cov-2. by 13 may 2020, the total number of laboratory tests in tir (8422 tests/100 000 inhabitants) had exceeded those performed in all of germany (3787 tests/100 000) by a factor of 2.2. nevertheless, 17% of all tests in tir were positive for sars-cov-2, while only 6.3% of all tests were positive in germany as a whole. in order to identify early drivers of transmission, we analysed the first 110 cases with the onset of symptoms between 18 february and 12 march 2020. the most frequently reported exposures included having been guests at the small local beer tradition between 3 march and 7 march 2020 (13%), skiing vacation in austria or italy in february/march (11%), and the big, 1-day beer event in mitterteich (9%) (fig. 2) . both traditional beer festivities attract large numbers of local and regional visitors. three cases had been skiing and guests at the smaller beer festival; one had been skiing and went to the big 1-day beer event. twenty-seven (25%) cases reported other possible exposures with large numbers such as attending birthday parties, funerals or religious services. for 38 (35%) cases, no known exposure could be determined. we looked at public health measures taken in tir, the bavarian state and germany and analysed the impact of those measures on the course of the outbreak. a series of measures were implemented between 10 march and 18 march (fig. 3 , highlighted area). on the national level, events with more than 1000 participants were banned on 10 march 2020. in tir, an extensive information campaign was initiated, including the establishment of a 'corona hotline' on 13 march 2020. kindergartens and schools were closed starting from 11 march 2020 in mitterteich, the most affected town in the district. the state of bavaria announced the intention to declare a state of emergency on 15 march 2020. during this period, when looking at the date of symptom onset, case numbers in tir were still increasing, however appeared to reach a plateau (tir) or even decreased (mitterteich) when looking at the date of infection (fig. 3) . on 18 march, a complete lockdown was implemented in mitterteich, followed by a partial lockdown for the entire state of bavaria on 21 march 2020. when looking at the date of symptom onset, the numbers of cases started to decrease on 22 march, and had already begun to decrease shortly before the lockdown of mitterteich. we compared age, sex and predisposing chronic diseases of cases in tir with cases from the rest of germany (table 1) . among cases in tir, 58% were female compared to 52% in the rest of germany. cases in tir were significantly older with a median age of 56 years (interquartile range (iqr): 41-56) compared to 50 years (iqr: 32-63) in the rest of germany. however, there was no age difference among deceased cases, with a median of 83 years (iqr: 77-88) in tir vs. 82 years (iqr: 76-88) in the rest of germany. the crude cfr for tir was 11% (95% ci 9.7-14) compared to 4.3% (95% ci 4.2-4.4) in the rest of germany, while the age-standardised cfr was 7.3% (95% ci 6.2-8.3) in tir and 4.3% (95% ci 4.2-4.4) in germany ( table 1) . the higher median age and the high percentage of cases with underlying medical conditions were in agreement with the high number of cases in tir living in retirement homes. in the rest of germany, 8.0% of all cases were reported to live in a community facility, compared to 14% of all cases in tir (table 1) . of note, community facilities also include other institutions such as prisons, asylum centres or homes for disabled persons. in tir however, 145/154 cases living in community facilities were residents of elderly care homes. looking only at cases reported in community facilities, the cfr in the district of tir was 36% (95% ci 29-44) (age-standardised: 31%; 95% ci 24-37) and 20% (age-standardised: 20%; for both 95% ci 19-20) in the rest of germany. on the other hand, looking only at cases outside of community facilities, cases were still significantly older in tir (median age: 54 years, iqr: 38-69; germany: 49 years, iqr: 32-60) and crude and age-standardised cfr were still elevated (tir: crude: 7.5% (95% ci 6.0-9.4), standardised: 5.0% (95% ci 4.0-6.1)) compared to the rest of germany (crude and standardised: 3.0% (95% ci 2.9-3.1)) ( table 1) . three elderly care homes in the district were affected early in the covid-19 outbreak, with attack rates of 21-59% among residents, and 29-31% among staff (table 2 ). care home 1 was the first affected facility, with the first cases among personnel. a member of staff tested positive for sars-cov-2 on 16 march 2020 but had been symptomatic from 8 march 2020. since the symptoms were only mild, the case worked in the facility until the date of the test. symptom onset for the first resident was on 16 march 2020. after the outbreaks in the three described elderly care homes, enhanced protective measures and a general screening strategy were applied for all 15 community homes in the district (14 elderly care homes and one home for the disabled) from 9 april onwards. this report describes one of the first, severe community outbreaks of covid-19 in germany and in the history of the pandemic in europe. sars-cov-2 may have been introduced to tir from high-risk areas in austria and italy, the latter being severely affected by the epidemic starting february/march 2020 [7] . from 25 february 2020 onwards, the rki declared several regions in northern italy and austria as covid-19 risk areas, including popular ski regions mentioned as exposure sites of german cases [8, 9] . several outbreaks in germany and europe have been linked to exposure during skiing holidays in austria or italy [10, 11] . at the time, measures such as cancelling of social events, wearing protective masks or reduced contact in the community in general were not yet established in germany and the virus was able to spread rapidly in the community. mass-gathering events such as the carnival in germany in february enhanced the spread of sars-cov-2 in communities [12] . in tir, possible accelerators of the pandemic might have been the local beer festivals at the beginning of march. however, the steep increase observed in the epidemiological curve is likely due to a variety of factors rather than by mass gatherings alone. it can be speculated that unnoticed transmission occurred at the end of february in tir. böhmer et al. and others showed that cases may be infectious even before symptom onset and the virus can be transmitted by cases with only mild symptoms, leading to unrecognised transmission [13] . our data show that the first case reported the onset of symptoms on 18 february 2020. retrospectively however, it is not clear if these early symptoms were due to covid-19 as the case was only tested in march after a second episode of respiratory symptoms. we used a calculated date of infection, assuming a mean incubation period of 5 days to evaluate public health measures implemented in tir, bavaria and germany, respectively. our data suggest that the extensive information campaigns at the local level, particularly in mitterteich, were valuable to avoid a further increase of cases in the whole district of tir. the lockdown of mitterteich and the partial lockdown in the bavarian state probably contributed to dropping case numbers and the combined measures most likely allowed control of the outbreak by mid-april. we calculated the reproduction number r from dates of reporting due to the completeness of these data. however, considering the delay of disease onset and reporting as well as our analysis of symptomatic cases, measures were most likely effective earlier. our data showed that cases in tir were older than in the rest of germany, but age alone cannot explain the higher cfr. no analysis of underlying medical conditions was possible in our study. however, a report of the wido scientific institute stated that 31% of the general population in tir suffers from at least one underlying medical condition aggravating the course of covid-19, while this is the case for 26% of the general german population [14] . the combination of vulnerable populations hit by the pandemic and applied testing strategies, primarily focusing on acute cases and thus missing asymptomatic and mildly symptomatic transmitters, has been shown to lead to high cfr in other regions [15] . starting from the first reported cases in tir, testing of all close contacts of cases regardless of symptoms was performed in accordance with the testing strategy of the bavarian state (valid until 19 march 2020). however, due to the large number of cases and limited resources, primarily acute cases were tested at the height of the tir outbreak as suggested by national and international guidelines [16, 17] . even though individual measures such as usage of antiviral disinfectant in elderly care homes had already been implemented at the end of february; these did not prove to be enough to prevent initial transmission into and within these facilities. extensive measures in elderly care homes, including a complete ban on visitors from mid-march, extensive use of personal protective equipment, a thorough screening strategy and isolation of positive cases were likely effective strategies to prevent transmission in further care homes in tir. in conclusion, the combination of exposures in high incidence areas and undiagnosed infections followed by intense transmission may have led to accelerated community transmission in tir. subsequent spreading of sars-cov-2 to vulnerable populations, including people in care homes for elderly, resulted in a high cfr. we could show that local and regional public health capacities are essential for the response. measures at the local level allowed a timely response and were particularly helpful to control the outbreak. we recommend strengthening of local public health capacities in germany, as these were crucial in implementing targeted measures in a timely manner. immediate support of local public health authorities from the regional and national level should be guaranteed in cases of emergency. public health measures implemented in the course of the covid-19 epidemic should be evaluated thoroughly to allow a targeted and effective response in similar situations in the future. a novel coronavirus from patients with pneumonia in china who. coronavirus disease 2019 (covid-19): situation report-170 schätzung der aktuellen entwicklung der sars-cov-2-epidemie in deutschland -nowcasting incubation period and other epidemiological characteristics of 2019 novel coronavirus infections with right truncation: a statistical analysis of publicly available case data covid-19), version 26 who. coronavirus disease 2019 (covid-19): situation report-71 spread of sars-cov-2 in the icelandic population der erste monat mit covid-19-fällen im land-kreis wittenberg investigation of a covid-19 outbreak in germany resulting from a single travel-associated primary case: a case series. the lancet infectious diseases monitor: gesundheitliche beeinträchtigungen -vorerkrankungen mit erhöhtem risiko für schwere verläufe von covid-19. verbreitung in der bevölkerung deutschlands und seinen regionen case-fatality rate and characteristics of patients dying in relation to covid-19 in italy who. laboratory testing strategy recommendations for covid-19. available at acknowledgements. we would like to thank all employees of the local health district in tirschenreuth involved in interviewing cases and in contributing to data collection. we thank simon packer (public health england), jan walter (rki), christina frank (rki) and viviane bremer (rki) for their helpful comments.financial support. this research received no specific grant from any funding agency, commercial or not-for-profit sectors. data availability. aggregated data from a limited version of the german notification system database can be retrieved via survstat@rki 2.0 https:// survstat.rki.de/. detailed data are confidential and protected by german law and are available from the corresponding author upon reasonable request. key: cord-291361-2vn1o7ag authors: li, jing; ding, jiguang; chen, li; hong, liang; yu, xiaoqi; ye, enling; sun, gangqiang; zhang, binbin; zhang, xinxin; sun, qingfeng title: epidemiological and clinical characteristics of three family clusters of covid-19 transmitted by latent patients in china date: 2020-07-06 journal: epidemiol infect doi: 10.1017/s0950268820001491 sha: doc_id: 291361 cord_uid: 2vn1o7ag from 21 january 2020 to 9 february 2020, three family clusters involving 31 patients with coronavirus disease 2019 were identified in wenzhou, china. the epidemiological and clinical characteristics of the family cluster patients were analysed and compared with those of 43 contemporaneous sporadic cases. the three index cases transmitted the infection to 28 family members 2–10 days before illness onset. overall, 28 of the 41 sporadic cases and three of 31 patients in the family clusters came back from wuhan (65.12 vs. 9.68%, p< 0.001). in terms of epidemiological characters and clinical symptoms, no significant differences were observed between the family cluster and sporadic cases. however, the lymphocyte counts of sporadic cases were significantly lower than those of family cluster cases ((1.32 ± 0.55) × 10(9)/l vs. (1.63 ± 0.70) × 10(9)/l, p = 0.037), and the proportion of hypoalbuminaemia was higher in sporadic cases (18/43, 41.86%) than in the family clusters (6/31, 19.35%) (p < 0.05). within the family cluster, the secondand third-generation cases had milder clinical manifestations, without severe conditions, compared with the index and first-generation cases, indicating that the virulence gradually decreased following passage through generations within the family clusters. close surveillance, timely recognition and isolation of the suspected or latent patient is crucial in preventing family cluster infection. since december 2019, an epidemic of coronavirus disease 2019 (covid19) , associated with severe acute respiratory syndrome coronavirus 2 (sars-cov-2), emerged in wuhan, hubei province, china [1, 2] . sars-cov-2 has been characterised by high infectivity through human-to-human transmission and relatively low mortality [3] . the mean r0 of sars-cov-2 is estimated to range from 2.24 to 3.58 [4, 5] . as a result, the epidemic of covid-19 has rapidly spread to the whole country and worldwide. as of 26 march 2020, 195 countries had been affected, and the cumulative number of confirmed cases had reached 81 968 and 389 199 in china and worldwide, with 3293 and 17 914 deaths, respectively. currently, the epidemiological characteristics of covid-19, especially transmission patterns, have not been well elucidated. the first family clustering study reported that five family members who travelled to wuhan from shenzhen were infected with sars-cov-2, and when they returned to shenzhen, the additional family members who did not travel to wuhan became infected with the virus [6] . the epidemiological and phylogenetic analysis indicates that sars-cov-2 can be transmitted person-to-person in hospital and family settings [6] . subsequently, several studies also reported the family cluster transmission resulting in the infection of 3-11 family members [7] [8] [9] [10] [11] , even during the incubation period. however, further investigation is required to understand the transmission patterns among family members. wenzhou is one of the regions with a high prevalence of infection outside hubei province, probably because of the close economic cooperation and convenient public transportation between two regions. in the early stage of the epidemic, over 7000 people came back to ruian, a county-level city under the administration of wenzhou. among these people, 74 were diagnosed with covid-19, and three family clusters of 31 members were identified. in this study, we aimed to investigate the epidemiological and clinical characteristics of these three family clusters of covid-19 cases by comparing them with sporadic cases, which would provide insights for epidemic control in the context of the current serious situation worldwide. from 21 january 2020 to 9 february 2020, 74 covid-19 patients, who were all positive for the nucleic-acid test of sars-cov-2 and received isolation and treatment in the designated ruian people's hospital, were enrolled in the present retrospective study. among the patients, 31 patients were identified to belong to three different families according to their family relationships and history of close contacts (figs 1 and 2) , while the other 43 patients were sporadic cases. the epidemiological history, including exposure history, travelling vehicle, contact tracing, family member relationship, date of illness onset and date of admission and isolation, were collected in detail for each patient by two attending physicians. the transmission chain was carefully evaluated to clarify the relationship within the cluster members, according to the close contact history and exposure time. a close contact was defined as an act of sharing a meal, party, vehicle or living room with a confirmed or latently infected patient within 14 days. an index case (g0) was defined as the original source of sars-cov-2 infection among the family. the patients who were infected by contact and exposure to the index case were defined as the first-generation cases (g1), and the patients who were infected by contact and exposure to the first-generation patients were defined as second-generation cases (g2) and so on. the patients without an infected family member were defined as sporadic cases. index case a, a 46-year-old female who came back from wuhan on 17 january 2020, had dinner with four family members on 18 january 2020. she developed symptoms of cough 10 days later and was confirmed with covid-19 on 1 february 2020. the four family members, cases a1-4 (g1), developed the disease 5−14 days after the dinner and were called firstgeneration cases of index case a (figs 1a and 2a) . index case b was a 57-year-old female who came back from wuhan on 15 january 2020. she developed a fever 4 days later and was diagnosed with covid-19 on 23 january 2020. during the next two weeks, four family members, cases b1, 2, 10 and 11 (g1), who had a close contact history with index case b were confirmed to be infected by sars-cov-2 sequentially. then, case b2 (g1), one of the four first-generation cases of index case b, transmitted the infection to five family members, cases b3-6 and 9 (g2), who were called second-generation cases. another first-generation case, b11 (g1), transmitted the infection to three family members, b12-14 (g2). moreover, one of the second-generation cases, b6 (g2) transmitted the infection to cases b7-8 (g3), the third-generation cases (figs 1b and 2b) . index case c (g0), a 43-year-old male, came back from wuhan and joined a party with four family members and two classmates on 18 january 2020. he developed illness with cough 2 days later and was diagnosed with covid-19 on 26 january 2020. the six persons, cases c1-6 (g1), who joined the party also developed an illness within the next 2 weeks. case c6 (g1) was responsible for the second-generation infection to three family members, cases c7-9 (g2), while case c5 (g1) was responsible for transmission to one family member, c10 (g2) (figs 1c and 2c) . all patients were diagnosed with covid-19 by real-time reverse transcriptase-polymerase chain reaction (rt-pcr) assay, according to the guideline for diagnosis and treatment for novel coronavirus pneumonia released by the national health commission of china (5 th edition) [12] . the rt-pcr tests for influenza a and b for all patients were negative. written informed consent, according to the declaration of helsinki, was obtained from each patient. this study was approved by the ethics committee of the ruian people's hospital (approval number: yj20200013). the clinical information of all enrolled patients was retrieved from the hospital history system, including the demographic data, laboratory test results, radiological results, treatment regimens, duration of treatment, duration of hospitalisation and treatment outcomes. the applications of intranasal oxygen inhalation and assisted mechanical ventilation along with comorbidities including hypertension, diabetes, chronic obstructive pulmonary disease, chronic kidney disease and malignant tumours were recorded. the patients were divided into different clinical types, according to the guidelines by the national health commission of china [12] . patients who presented with classic symptoms and positive sars-cov-2 rna but without pneumonia lesions on computed tomography (ct) scan were defined as mild cases, and those with classic symptoms, positive sars-cov-2 rna and pneumonia lesions on ct scan were defined as common cases. in addition, patients who met the following criteria were defined as severe cases: (1) respiratory distress, a respiratory rate (rr) ≥30 beats/min; (2) an oxygen saturation level less than 93% in resting state and (3) a partial pressure of oxygen (pao 2 )/oxygen concentration (fio 2 ) ≤300 mmhg (1 mmhg = 0.133 kpa). sars-cov-2 rna was detected by rt-pcr assay with the taqman probes targeting orf1ab, n and e genes, and expressed as the cycle threshold (ct) value (shanghai biogerm medical biotechnology co., ltd). the amplification products for genes with a ct value of less than 38 were considered as positive. sputum samples or throat swab samples were taken for analysis at baseline and then every 2-3 days until hospital discharge. continuous variables were expressed as mean and standard deviation (sd) or median and interquartile range (iqr), and categorical variables were expressed as a number (%). the values were compared by student's t-tests, one-way anova or mann−whitney test, or kruskal−wallis as appropriate. all data analysis was performed with r software (version 3.6.2) and empowerstates software (http:// www.empowerstats.com, x&y solutions, inc., boston, ma). a twosided p value of less than 0.05 was considered statistically significant. of all 74 patients, 35 were male, and 39 were female. the mean age was 44.26 years old. the most common symptoms were table 1) . of the 43 sporadic cases, 28 (65.12%) returned from wuhan while three (9.68%) of the family cluster cases returned from wuhan (p < 0.001). the family members were infected by the three index cases who were in the latent period 2-10 days before the onset of illness. interestingly, the time of illness onset of patient a1, who had no travel history to wuhan or contact with other patients, was earlier than that of index case a. patient a1 (g1) developed severe pneumonia subsequently. the incubation period of sporadic cases (4.00 (2.00-7.00) days) was similar to that of the family cluster (6.00 (4.00-7.00) days) (p = 0.192). the time from symptom onset to hospitalisation, the time from symptom onset to diagnosis and the duration of hospitalisation were not significantly different between sporadic and family cluster cases (all p > 0.05) ( table 1) . there was no significant difference in the frequency of common symptoms, including fever (86.05 vs. 80.65%), cough (79.07 vs. 77.42%) and sputum (51.16 vs. 54.84%), between sporadic and family cluster cases (all p > 0.05). also, the proportions of mild, common and severe types were similar between sporadic (18.60, 76.74, and 4.65%, respectively) and family cluster (22.58, 70.97 and 6.45%) cases (p = 0.848). however, the decrease of albumin was more frequent in sporadic cases (41.86%) than in the family cluster cases (19.35%) (p < 0.05). while the levels of alanine transaminase, aspartate aminotransferase and creatinine were not different between the two groups (all p > 0.05), the level of lymphocyte counts was significantly lower in sporadic cases ((1.32 ± 0.55) × 10 9 /l) than in the family cluster cases ((1.63 ± 0.70) × 10 9 /l) (p = 0.037). the viral load (ct value) was not different between the two groups ((30.55 ± 4.78) vs. (29.46 ± 4.37), p = 0.38) ( table 2 ). the imaging features of the pulmonary lesions on ct scan were not apparently different between the two groups (data not shown). jing li et al. among the family clusters, three index cases (g0) transmitted the infection to 14 first-generation cases (g1), who transmitted the infection to 12 second-generation cases (g2) and subsequently two third-generation cases (g3) (figs 1 and 2) . then, the epidemiological and clinical characteristics were compared between generations 0 and 1 (g0 + g1) and generations 2 and 3 (g2 + g3) cases. g0 + g1 cases were older than g2 + g3 cases (44.00 (42.00-55.00) vs. 37.00 (23.00-42.25) years, p = 0.039). the incubation time, the time from illness onset to hospitalisation, the time from illness onset to diagnosis and the duration of hospitalisation were not significantly different between the two groups (all p > 0.05). the proportion of mild cases in g2 + g3 cases (5/14, 35.71%) appeared to be larger than that in g0 + g1 cases (2/17, 11.76%) although the difference was not statistically significant (p = 0.153). the viral loads were not significantly different between g0 + g1 and g2 + g3 cases (29.31 ± 4.33) vs. (29.67 ± 4.62), p > 0.05. the level of lymphocyte counts tended to be lower in g0 + g1 cases ((1.48 ± 0.56) × 10 9 /l) than in g2 + g3 cases ((1.80 ± 0.83) × 10 9 /l), but the difference was not statistically significant (p = 0.213) ( table 2) . the differences of epidemiological and clinical characteristics between sporadic cases and generations of family clusters were determined. it was shown that the lymphocyte counts of sporadic cases were significantly lower than those of g2 + g3 cases (p = 0.033) but without a significant difference with g0 + g1 cases (p > 0.05). there were no differences in the incubation time, the time from illness onset to hospitalisation, the time from illness onset to diagnosis and the duration of hospitalisation among the three groups (p > 0.05) (tables 1 and 2 ). in this study, the epidemiological and clinical characteristics of three family clusters were investigated using sporadic patients as controls. such a specific population provides us an opportunity to analyse the relationship between transmission and the disease presentation in different settings where the sporadic patients were used as external controls to the family cluster patients. more patients in sporadic cases came back from wuhan than in the family cluster. all three index cases were latent patients without any symptoms at the time when they came back to ruian. this study revealed that sporadic cases had lower levels of albumin and lymphocyte counts than family cluster cases; otherwise, there were no significant differences in terms of other epidemiological characters and clinical features between the two groups. in addition, the lymphocyte counts in sporadic cases were lower than those in the cases of second and third generations family cluster cases although there was no difference in the lymphocyte counts among the different generations within family cluster. human coronavirus pneumonia is often associated with an elevated production of chemokines, which recruit massive inflammatory cell infiltration and release cytokines resulting in acute pulmonary injury [13] . the decrease of lymphocyte counts and elevation of cytokines/chemokines are the hallmark of coronavirus-associated pneumonia and are associated with the severity of the pneumonia. recent studies on covid-19 have demonstrated that the lymphocyte counts in the peripheral blood are remarkably decreased in patients who are admitted in the intensive care unit (icu), compared with non-icu patients [1, 14] . in addition, several studies on covid-19 or mers have shown that hypoalbuminemia is a frequent feature and associated with the severity of the pneumonia [15] [16] [17] , probably owing to increased energy consumption or altered pulmonary vessel permeability. the finding in this study that the decrease of lymphocyte counts and hypoalbuminaemia in sporadic cases, compared with family cluster cases, indicates that there is an increased immune activation or dysfunction and thus more severe pulmonary inflammation in sporadic cases than in family cluster cases. within the family clusters, virus transmission through different generations and the clinical presentations were investigated. the age of second-and third-generation patients was younger than that of the index and first-generation patients. it has been reported that when hosts of different sexes or ages were encountered, the pathogen may change optimal exploitative strategy, leading to considerable variation of pathogen transmission and virulence [18] . the trade-off between transmission and virulence would change in coordination with host immunity that is associated with age. sars-cov-2 seems prone to affect older males with comorbidities [19] . in this study, two patients, a1 (g1) and b1 (g1), developed severe pneumonia in first generation, while no severe cases were observed in the second or third generation. more mild patients were found in the second or third generation, which are also noted to be younger in age, may be attributed to age. in addition to, it could be inferred that following the passage through several generations within the family cluster, the virulence of sars-cov-2 decreased gradually. epidemiological evidence from covid-19 family clusters has suggested that most index cases are asymptomatic carriers, mild patients or even latent patients [7, 9, 10] , which is consistent with our observation. however, the characteristics of index cases with covid-19 are reportedly different from those in the mers family clusters, who have moderate or severe symptoms and are never asymptomatic carriers [20] [21] [22] . the higher virulence and mortality in mers may explain the different characteristics of the index cases. as the viral load of sars-cov-2 detected in the asymptomatic patients is similar to that in the symptomatic patients [23] , the asymptomatic index cases are capable of causing cluster infection in the family setting. a recent study with the largest sample so far in china showed that the median incubation period for covid-19 was 4 days (iqr, 2−7), with the longest incubation period up to 24 days, and only 43.8% patients presented with a fever at admission [24] , highlighting the importance of monitoring and isolating potential infected family members who have had an exposure history in the family setting. previous studies have demonstrated that the outbreak of sars-cov and mers-cov infections have resulted in large clusters of patients, most of which are associated with the nosocomial transmission, called super-spreader events [25] [26] [27] . since the virological and clinical characteristics are similar among sars-cov, mers-cov and sars-cov-2 [28] , it is worth noting the super-spreader events in covid-19. as of 7 march, the daily number of newly diagnosed patients had decreased to less than 150 cases in china. currently, there are no specific antiviral agents or vaccines available for sars-cov-2, which possesses a high infectivity, and thus advanced epidemiological surveillance and timely identification and isolation of suspected cases or individuals who had a close contact or exposure history remains a priority to prevent family cluster or superspread events. the chinese experience shows that intensive social interventions, including isolation are crucial in delaying and blocking the spread and subsequent outbreaks of the disease. in conclusion, family clusters of covid-19 can be caused by latent patients. the epidemiological and clinical symptoms are similar between the family cluster and sporadic patients, but the sporadic patients showed lower lymphocytes and hypoalbuminaemia. these findings indicate that close surveillance, timely identification and isolation of the suspected or latent cases is crucial in preventing family clusters or even super-spread events. clinical features of patients infected with 2019 novel coronavirus in wuhan a novel coronavirus from patients with pneumonia in china early transmission dynamics in wuhan, china, of novel coronavirus-infected pneumonia preliminary estimation of the basic reproduction number of novel coronavirus (2019-ncov) in china, from 2019 to 2020: a data-driven analysis in the early phase of the outbreak estimating the unreported number of novel coronavirus (2019-ncov) cases in china in the first half of january 2020: a datadriven modelling analysis of the early outbreak a familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster asymptomatic cases in a family cluster with sars-cov-2 infection potential presymptomatic transmission of sars-cov-2 a familial cluster of infection associated with the 2019 novel coronavirus indicating potential person-to-person transmission during the incubation period a family cluster of sars-cov-2 infection involving 11 patients in nanjing, china. the lancet infectious diseases transmission of covid-19 in the terminal stage of incubation period: a familial cluster national health commission of china (2020) guideline for diagnosis and treatment for novel coronavirus pneumonia pathogenic human coronavirus infections: causes and consequences of cytokine storm and immunopathology clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan, china predictive factors for pneumonia development and progression to respiratory failure in mers-cov infected patients analysis of factors associated with disease outcomes in hospitalized patients with 2019 novel coronavirus disease clinical and biochemical indexes from 2019-ncov infected patients linked to viral loads and lung injury explosive school-based measles outbreak: intense exposure may have resulted in high risk, even among revaccinees epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study family cluster of middle east respiratory syndrome coronavirus infections a family cluster of middle east respiratory syndrome coronavirus infections related to a likely unrecognized asymptomatic or mild case family cluster of middle east respiratory syndrome coronavirus infections sars-cov-2 viral load in upper respiratory specimens of infected patients clinical characteristics of coronavirus disease 2019 in china comparison of clinical course of patients with severe acute respiratory syndrome among the multiple generations of nosocomial transmission mers-cov outbreak in jeddah -a link to health care facilities the sars outbreak in a general hospital in tianjin, china -the case of super-spreader emerging threats from zoonotic coronaviruses − from sars and mers to 2019-ncov acknowledgements. we thank all participating patients and acknowledge the support of a grant from ruian science and technology bureau (ms2020023). this project was supported by the medjaden academy & research foundation for young scientists (grant no. covid-19-mja20200321). data availability statement. all data generated or analysed during this study are included in this paper. key: cord-000724-lzhobnch authors: zhang, j.; while, a. e.; norman, i. j. title: seasonal influenza vaccination knowledge, risk perception, health beliefs and vaccination behaviours of nurses date: 2011-11-18 journal: epidemiol infect doi: 10.1017/s0950268811002214 sha: doc_id: 724 cord_uid: lzhobnch the relationship between knowledge, risk perceptions, health belief towards seasonal influenza and vaccination and the vaccination behaviours of nurses was explored. qualified nurses attending continuing professional education courses at a large london university between 18 april and 18 october 2010 were surveyed (522/672; response rate 77·7%). of these, 82·6% worked in hospitals; 37·0% reported receiving seasonal influenza vaccination in the previous season and 44·9% reported never being vaccinated during the last 5 years. all respondents were categorized using two-step cluster analyses into never, occasionally, and continuously vaccinated groups. nurses vaccinated the season before had higher scores of knowledge and risk perception compared to the unvaccinated (p<0·001). nurses never vaccinated had the lowest scores of knowledge and risk perception compared to other groups (p<0·001). nurses' seasonal influenza vaccination behaviours are complex. knowledge and risk perception predict uptake of vaccination in nurses. annual epidemics of seasonal influenza result in about 3-5 million cases of severe illness and 250 000-500 000 deaths worldwide [1] . healthcare workers (hcws) can be a key source for influenza transmission in communities and hospitals as they are exposed to both infected patients and high-risk groups [2, 3] . vaccination is the most effective way to prevent infection and severe outcomes [1] and the principal measure to reduce the impact of epidemics, such as hospitalization, mortality and morbidity [2, [3] [4] [5] . moreover, studies suggest that the vaccination of hcws has substantial economic benefits as well as health-related benefits, including reduced absenteeism from work and the extra costs of sick leave and staff replacement [4, 6, 7] . for the above reasons, the world health organization (who), united kingdom department of health (doh) [8] , united states centers for disease control and prevention (cdc), other healthcare professional organizations and many countries' government agencies [1, 9, 10] strongly recommend the annual seasonal influenza vaccination of hcws. however, studies suggest that influenza vaccine uptake in hcws is often low worldwide [11] [12] [13] [14] . for example, the overall seasonal vaccination rate in england for hcws was 26 . 4% for the 2009/2010 season [15] . nurses, as the group having the most patient contact, are more reluctant to be vaccinated than other hcws [16] [17] [18] [19] [20] [21] [22] [23] . although predictors influencing nurses' vaccination practices have been identified to some extent regarding knowledge and risk perception [16] [17] [18] [19] [23] [24] [25] [26] [27] , further studies are needed to explore the influences on nurses' attitudes and practices regarding influenza vaccination and to identify the major influencing factors for their vaccination behaviours. this study aimed to examine the relationship between knowledge, risk perceptions, health beliefs towards seasonal influenza and vaccination and the vaccination behaviours of nurses. a cross-sectional survey was conducted of qualified nurses between 18 april and 18 october, 2010. qualified nurses attending continuing professional education courses at a large university in central london were invited to participate in the study. potential respondents were given a study information sheet and a questionnaire by the investigator. completed questionnaires were collected immediately by the investigator or returned by mail to the research team using freepost addressed envelopes. questionnaire completion was anonymous so that it was not possible to follow up non-response. ethical approval was obtained from the university ethics committee. the questionnaire collected the following data : (1) knowledge about seasonal influenza and vaccination (22 items requiring true, false or unsure responses) included five dimensions to assess general information, severity of influenza, influenza vaccination, high-risk groups and vaccination-recommended groups; (2) risk perception (12 items with a 4-point likert scale) towards influenza and pandemic with three dimensions (i.e. personal vulnerability to illness, negative consequences of contracting influenza and severity of influenza) ; (3) health locus of control including internal, chance and powerful others dimensions assessed by the multidimensional health locus of control (mhlc) scales [28] (18 items) ; (4) vaccination behaviours (nine items) including vaccination status (whether respondents had been vaccinated in the previous season), vaccination intent (whether respondents intended to be vaccinated next season) and vaccination history (how many times respondents had been vaccinated in the last 5 years) ; (5) reasons for accepting or refusing vaccination using two open questions; and (6) demographic characteristics (10 items) including gender, age group, highest educational qualification, place of work, clinical speciality, year of qualification as a nurse and whether or not respondents had direct patient contact. the cronbach's a-coefficients for the three newly developed scales (sections 1, 2, 4) ranged from 0 . 701 to 0 . 763 and principal components analysis produced a good fit and confirmed the internal design of the instrument. statistical analysis was performed using spss version 15.0 (spss inc., usa). the x 2 test or fisher's exact test was used to explore the statistical differences between categorical variables. the independentsamples t test was used to compare statistical difference between continuous variables in two groups. the one-way between-groups analysis of variance (anova) was used to explore the differences between more than two groups. logistic regression was performed to explore the impact of the variables on vaccination status. the two-step cluster analysis procedure was performed to explore the natural groupings (i.e. clusters) within the respondents. the clustering criterion was that the solution had smaller values of schwarz's bayesian information criterion (bic), a reasonably large ratio of bic changes and a large ratio of distance measures. a p value <0 . 05 was considered to denote statistical significance. in total, 672 questionnaires were distributed and 522 were returned representing a response rate of 77 . 7%. the characteristics of the respondents are summarized in table 1 . overall 188/508 respondents (37 . 0%) reported receiving a vaccination in the previous season with 44 . 9% never receiving a vaccination during the last 5 years. there was no difference in the demographic characteristics of the vaccinated or unvaccinated respondents in the previous season. the number of years qualified as a nurse for the two groups were 11 . 99¡9 . 085 years and 11 . 89¡8 . 624 years (p=0 . 898), respectively. comparison of knowledge and risk perception scores and sub-scores of mhlc are summarized in table 2 . there were significant differences in knowledge scores and risk perception between the vaccinated and unvaccinated nurses and between those with vaccination intent, no intent or unsure. there was no significant difference in the sub-scores of mhlc between the vaccinated and unvaccinated (data not shown in table) but there was a significant difference for the sub-score of powerful others between those groups with different vaccination intent. direct logistic regression was performed to assess the impact of a number of factors on the likelihood that respondents had been vaccinated in the previous season. the model contained five independent table 3 , only two of the independent variables made a unique statistically significant contribution to the model (knowledge score and risk perception score). the strongest predictor of vaccination status was the risk perception score, recording an odds ratio of 1 . 76, indicating that respondents who had higher risk perception scores were >1 . 76 times more likely to have been vaccinated in the last 12 months than those with lower scores, controlling for all other factors in the model. knowledge score with an odds ratio of 1 . 05 indicated that knowledgeable respondents were more likely to be vaccinated than the unknowledgeable, controlling for other factors in the model. the two-step cluster analysis procedure was used to explore the natural groupings within the respondents. first, the auto-clustering exploratory analysis was performed using the categorical variables of vaccination status, vaccination intent, vaccination history and the continuous variables of knowledge score and risk perception score. of the 522 respondents, 64 were automatically excluded from the analysis due to missing values on one or more of the variables. of the 458 respondents assigned to clusters, 195 (42 . 6%) were assigned to the first cluster, 143 (31 . subsequently the analysis was performed using the combined categorical variables of vaccination status in the previous season (=yes) and vaccination history and the continuous variables of knowledge and risk perception scores. the results were auto-clustered into four groups but not explainable. the procedure was repeated with the cluster number fixed to 2 due to the values of bic, ratio of bic changes and ratio of distance measures. of the total 188 vaccinated respondents, 12 were excluded due to missing values. of the remaining 176 respondents, 107 (60 . 8%) were assigned to cluster 1 and 69 (39 . 2 %) to cluster 2. vaccinated cluster 1 comprised those vaccinated only in the previous season, i.e. the newly vaccinated group and vaccinated cluster 2 contained those vaccinated in the previous season who had more than one previous vaccination, i.e. the continuously vaccinated group. then, the same analysis was repeated for the unvaccinated respondents and two clusters emerged, i.e. unvaccinated cluster 1 (never vaccinated) and unvaccinated cluster 2 (used to be vaccinated). the analysis had therefore separated the respondents into reasonable categories. a comparison of variables across all clusters revealed that the never vaccinated had the lowest knowledge score, risk perception score and powerful others sub-score of mhlc compared to the other clusters (p<0 . 001, p<0 . 001, p=0 . 020, respectively) and this difference was statistically significant. for the vaccinated, there were no significant differences across any variable for the newly vaccinated and continuously vaccinated clusters although there was a trend of higher average scores for knowledge and risk perception in the newly vaccinated cluster compared to those of the other clusters (p=0 . 652, p=0 . 288, respectively). for the unvaccinated, there were no statistically significant differences across the variables except for the mhlc 'powerful others ' sub-score (p=0 . 008). further comparisons were performed to explore whether there were differences across the different items of knowledge and risk perception in the clusters. in the clusters of never vaccinated, other vaccination history and vaccinated with intent, there were significant differences in knowledge related to general information, high-risk groups and vaccination of recommended groups with p values of <0 . 001, <0 . 003 and <0 . 006, respectively. on average those never vaccinated had the lowest score while those vaccinated with intent had the highest scores across all knowledge items. for only one item of risk perception, i.e. personal vulnerability to illness, was there a significant difference between the clusters of never vaccinated and other vaccination history and between never vaccinated and vaccinated with intent (p<0 . 000 respectively). those never vaccinated had the lowest average score. there was no statistically significant difference in the knowledge and risk perception item scores between the two vaccinated clusters. however, the newly vaccinated usually had higher scores than those of the continuously vaccinated except for one item, i.e. the vaccination of recommended groups. similarly, for the two unvaccinated clusters there was no difference for knowledge scores, but there was a significant difference in one risk perception item, i.e. personal vulnerability to illness (p=0 . 001). those never vaccinated had a lower score for this item than those who used to be vaccinated and they were also less knowledgeable compared to the other group. tables 4 and 5 . in this study, the seasonal influenza vaccination rate in nurses was 37 . 0 % which is higher than previous reports of vaccination coverage ranging from 14 . 3-26 . 4% in hcws in uk [12, 29, 30] and 16% in nurses reported by chalmers [27] and similar to o'reilly et al.'s reported vaccination coverage of nurses in elderly care units [19] . this higher vaccination rate might be explained to some extent by the uk media reports of the risk of seasonal influenza and h1n1 pandemics in 2009 which may have increased the sample nurses' risk perception towards influenza and consequently changed their vaccination decisions as noted in a previous study [31] . this study found that vaccination behaviours in nurses were more complex requiring an analysis of both vaccinated and unvaccinated nurses' behaviours. more levels of vaccination behaviours existed in the sample with the two-step cluster analysis revealing three whole population clusters, i.e. those never vaccinated, those vaccinated this season with intent next year, and those with other vaccination history. two clusters, the newly vaccinated and continuously vaccinated, were identified for the vaccinated group and another two clusters, never vaccinated and used to be vaccinated, were identified in the unvaccinated group. to improve the influenza vaccination rates in nurses, it may be helpful to develop different strategies which target the nurse groups of the never vaccinated and the occasionally vaccinated. we found that a lack of knowledge about influenza and vaccination was a strong predictor of nurses' vaccination behaviours, especially for those never vaccinated. this cluster had the lowest knowledge score, suggesting that increasing their knowledge might improve their vaccination behaviours. however, it seems there are 'persistent decliners ' who are in the 'habit ' of not having a vaccination. this suggests that future educational campaigns need to be persistent, durative, and intensive if their vaccination behaviours are to be modified. for those who had been vaccinated in the past but not in the current season, knowledge was also a predictor for their vaccination behaviours, which suggests that current vaccination campaigns have failed to address their misgivings about vaccination to maintain their compliance with the annual vaccination recommendation for hcws. between those occasionally vaccinated and continuously vaccinated, knowledge levels were not significantly different but the newly vaccinated in 2009 had on average higher knowledge scores than those continuously vaccinated. this may reflect an increase in their risk perceptions towards influenza due to widespread reporting of the risks in the media encouraging them to be vaccinated for the first time in their lives. this suggests that timing may be crucial to the success of vaccination campaigns making behaviour modification easier. future studies are required to explore the relationship between the content and timings of vaccination campaigns and nurses' first vaccination uptake. this study showed that the perception of personal vulnerability to illness was important in nurses making vaccination decisions. but perceptions of the negative consequences of contracting influenza and severity of influenza were not major factors, a finding which is consistent with findings of previous studies [16] . this suggests that future educational campaigns might be more effective if they focus on the negative personal consequences of contracting influenza and its sequelae rather than nurses' professional duty to protect patients or other vulnerable groups. additionally, the reasons which nurses gave for having vaccination focused upon their personal health motivation rather than a professional responsibility regardless of whether they were vaccinated or unvaccinated. concerns about the vaccine's side-effects and effectiveness or safety were the two most frequent reasons for not having a vaccination indicating continuing misconceptions about influenza vaccine in nurses. future educational campaigns may wish to consider providing targeted information to change these widespread myths in nurses. however, these concerns did not seem to influence vaccination decisions because both vaccinated as well as unvaccinated nurses noted these reasons against vaccination. it may be the case that 2 days of minor discomfort postvaccination is tolerable when set against a year's influenza protection. unvaccinated nurses reported 'no need ' as their reason not having a vaccination which is consistent with their low-risk perception of contracting influenza. the convenience of the vaccination programme was identified as an organizational reason highlighting the importance of easy access to vaccination to increase its coverage in nurses. our analysis of health locus of control data found that those never vaccinated had a lowest 'powerful others ' locus of control for their vaccination behaviours, indicating that they did not believe their health was something over which they had no control [32] . this pattern of health beliefs towards influenza vaccination is consistent with their low-risk perception of personal vulnerability to illness and 'no need ' as their reason refusing vaccination and may be an important factor for never vaccinated nurses. further studies are needed to explore what may influence this pattern of health locus of control in order to modify nurses' vaccination behaviours. some organizations have recently required mandatory seasonal influenza vaccination for hcws as a professional and ethical obligation to protect their patients' health [33, 34] . however, ethical issues have been raised with mandatory vaccination because, while promoting the interests of patients and employers, it challenges hcws' personal autonomy and freedom of choice [35, 36] . moreover, it has been suggested that vaccination is not the only avenue of influenza prevention and there are several other important measures that healthcare organizations may take to protect both patients and hcws [37] . further previous studies have also suggested that not all hcws support mandatory vaccination [38] . until mandatory influenza vaccination for hcws is accepted worldwide, continued efforts to improve nurses' vaccination behaviours will be required. this study has some limitations. first, there is possible selection bias of a convenience sample ; however, the broad range of qualified nurses together with a high response rate strengthen the results. the extent of bias is unknown especially regarding nurses not working in london or in different care settings. second, the survey relied on self-report vaccination data ; however, zimmerman et al. [39] found that selfreport data were reliable in comparison with medical records. third, the three factors explored relating to nurses' vaccination behaviours explained only 8 . 7-11 . 9% of the variance according to the logistic regression analysis (although it was statistically significant) and therefore our results cannot fully explain nurses' vaccination behaviours. additional predictors will need to be introduced into the model in future studies to fully explain nurses' vaccination behaviours. in conclusion, this study revealed that nurses' influenza vaccination behaviours are complex. knowledge and risk perception were identified as two predictors influencing nurses' vaccination decisions with the health belief pattern of 'less powerful others ' being an important predictor in the never vaccinated ; however, there are other influential factors which need to be identified in future studies. world health organization. world health organization (who) influenza (seasonal) factsheet n211 preventing nosacomial influenza by improving the vaccine acceptance rate of clinicians assessing the role of basic control measures, antivirals and vaccine in curtailing pandemic influenza : scenarios for the us, uk and the netherlands effects of influenza vaccination of health-care workers on mortality of elderly people in long-term care : a randomised controlled trial effectiveness of an influenza vaccine programme for care home staff to prevent death, morbidity, and health service use among residents : cluster randomised controlled trial organizational and environmental factors that affect worker health and safety and patient outcomes effectiveness of influenza vaccine in health care professionals : a randomized trial summary of flu immunisation policy centres for disease control and prevention. prevention and control of influenza. recommendations of the advisory committee on immunization practices (acip) prioritization strategies for pandemic influenza vaccine in 27 countries of the european union and the global health security action group : a review national seasonal influenza vaccination survey in europe influenza vaccination coverage rates in five european countries during season 2006/07 and trends over six consecutive seasons influenza vaccination among primary healthcare workers influenza vaccination acceptance among health-care workers : a nationwide survey influenza vaccination uptake monitoring on behalf of the department of health attitudes, knowledge and factors related to acceptance of influenza vaccine by pediatric healthcare workers correlation between healthcare workers' knowledge of influenza vaccine and vaccine receipt factors affecting nurses' decision to get the flu vaccine factors affecting influenza vaccine uptake among health care workers knowledge and attitudes about influenza vaccination amongst general practitioners, practice nurses, and people aged 65 and over influenza vaccination coverage among hospital personnel over three consecutive vaccination campaigns influenza vaccination rates and motivators among healthcare worker groups influenza vaccination in paediatric nurses : cross-sectional study of coverage, refusal, and factors in acceptance predictors of influenza vaccination amongst australian nurses impact of severe acute respiratory syndrome and the perceived avian influenza epidemic on the increased rate of influenza vaccination among nurses in hong kong influenza vaccination among registered nurses : information receipt, knowledge, and decision-making at an institution with a multifaceted educational program understanding healthcare worker uptake of influenza vaccination : a survey accessed trends in influenza vaccination coverage rates in the united kingdom over six seasons from 2001-2 to 2006-7 pandemic h1n1 (swine flu) and seasonal influenza vaccine uptake amongst frontline healthcare workers in england avian flu : the creation of expectations in the interplay between science and the media development of the multidimensional health locus of control (mhlc) scales policy statementrecommendation for mandatory influenza immunization of all health care personnel revised shea position paper : influenza vaccination of healthcare personnel mandatory vaccination of health care workers the ethics of mandatory vaccination against influenza for health care workers point counterpoint : mandatory flu vaccination for health care workers beliefs on mandatory influenza vaccination of health care workers in nursing homes : a questionnaire study from the netherlands sensitivity and specificity of patient self-report of influenza and pneumococcal polysaccharide vaccinations among elderly outpatients in diverse patient care strata we are grateful for the statistical advice of peter milligan. none. key: cord-354272-99vw735a authors: darling, n. d.; poss, d. e.; schoelen, m. p.; metcalf-kelly, m.; hill, s. e.; harris, s. title: retrospective, epidemiological cluster analysis of the middle east respiratory syndrome coronavirus (mers-cov) epidemic using open source data date: 2017-10-24 journal: epidemiol infect doi: 10.1017/s0950268817002345 sha: doc_id: 354272 cord_uid: 99vw735a the middle east respiratory syndrome coronavirus (mers-cov) is caused by a novel coronavirus discovered in 2012. since then, 1806 cases, including 564 deaths, have been reported by the kingdom of saudi arabia (ksa) and affected countries as of 1 june 2016. previous literature attributed increases in mers-cov transmission to camel breeding season as camels are likely the reservoir for the virus. however, this literature review and subsequent analysis indicate a lack of seasonality. a retrospective, epidemiological cluster analysis was conducted to investigate increases in mers-cov transmission and reports of household and nosocomial clusters. cases were verified and associations between cases were substantiated through an extensive literature review and the armed forces health surveillance branch's tiered source classification system. a total of 51 clusters were identified, primarily nosocomial (80·4%) and most occurred in ksa (45·1%). clusters corresponded temporally with the majority of periods of greatest incidence, suggesting a strong correlation between nosocomial transmission and notable increases in cases. middle east respiratory syndrome coronavirus (mers-cov) is a respiratory illness caused by a novel coronavirus originally discovered in 2012. mers-cov can cause severe acute respiratory symptoms, including fever, cough, and shortness of breath, and is fatal in approximately one-third of reported cases. presently, there is no vaccine to prevent infection and no specific antiviral treatment for those infected with the virus [1] . mers-cov is the sixth strain of human coronavirus identified. although mers-cov cases were first reported from the kingdom of saudi arabia (ksa) in september 2012, the first two known cases were retrospectively discovered in jordan from april 2012 [2] . since its discovery in 2012, mers-cov cases have predominately been reported from ksa; however, cases have also been reported from algeria, austria, bahrain, china, egypt, france, germany, greece, iran, italy, jordan, kuwait, lebanon, malaysia, the netherlands, oman, philippines, qatar, republic of korea (rok), thailand, tunisia, turkey, united arab emirates (uae), united kingdom (uk), united states (usa), and yemen [2] . according to the armed forces health surveillance branch (afhsb), which is an organization under the defense health agency that utilizes biosurveillance to protect and promote the health of the us armed forces, as of 1 june 2016, 1806 cases of mers-cov have been reported, including at least 564 deaths [3] . afhsb's death count (case fatality proportion -31%) includes only those deaths which have been publicly reported and verified. the dynamics of the transmission of mers-cov is essential to understanding the risk posed by the virus as well as instituting effective infection control and prevention practices in areas where humans are at a greater risk of exposure. despite beliefs that camels are the most likely reservoir of the virus, the limited camel-to-human transmission has been reported by cdc and who [4, 5] . many published findings suggest that camel calves play a potential role in mers-cov transmission [6] [7] [8] . they found the 'rate of virus isolation [was] significantly higher in calves', and calves are often more acutely infected with mers-cov than adult camels, suggesting increased infectivity among calves [6] [7] [8] . although camel breeding season has been a proposed contributor to mers-cov transmission among camels, there is insufficient literature to support that human infections are more common during this time of year. human-to-human transmission of mers-cov has been investigated as a potentially significant route of spread. researchers have found that close contact with infected individuals is required to transmit the virus from one human to another, supporting the role of limited human-to-human transmission in the mers-cov epidemic and, more specifically, its role in nosocomial and household clusters [9] . while it is known that the virus can be transmitted through respiratory secretions, the exact routes through which the virus spreads are not well understood [9] . in an effort to better understand the patterns of transmission, a retrospective analysis of epidemiological clusters identified throughout the ongoing mers-cov epidemic was conducted using open-source data. epidemiological literature, classified by the tiered-source classification system (tscs), addressing mers-cov cluster analyses was collected and reviewed. several key search terms were utilized to capture all cluster-related literature, including 'mers-cov', 'nosocomial', 'cluster', 'transmission', 'superspreader', 'contact tracing', and 'healthcare worker'. see supplemental 1 for a comprehensive list of key search terms. publications selected for inclusion in the literature review were classified using tiers. this system was developed by afhsb to categorize the sources used in the literature review by their credibility. literature published by official sources, including the who and cdc, was considered a tier 1 source. literature published by reputable sources other than the who and cdc (e.g. all peer-reviewed journals regardless of perceived impact factor, including the lancet or nature) was considered a tier 2 source. literature from a foreign source, such as the ksa ministry of health (moh) or a media source, was classified as a tier 3 source. inclusion of a source required two or more of the key terms in supplemental 1. the literature included in the analysis encompassed translatable studies, situation reports from public health agencies, and publications updated to include more recent cluster information. studies related to mers-cov that were identified as having one or more of the following characteristics were excluded from the analysis: an investigational period prior to 2012; published in a non-translatable language; molecular-based; a focus on viral reservoirs, genealogy, or genetics, preventive measures, or other coronaviruses. in total, 80 studies were selected for inclusion. of these, 20 were classified as tier 1 sources, 22 as tier 2, and 38 as tier 3 (table 1 ). these sources were used to identify clusters as well as verify and characterize associations between cases. a mers-cov cluster was defined as two or more persons with onset of symptoms within a 14-day incubation period who are associated with a specific setting [10] . clusters were further categorized as exported, nosocomial, and/or household clusters. an exported cluster was defined as any cluster that resulted from verified travel of an index case (from an area of known mers-cov transmission) within one incubation period (14 days) of symptom onset. if the index case was asymptomatic, verified travel from an area of known mers-cov transmission within 14 days prior to the date of the index case was laboratory confirmed for mers-cov. a nosocomial cluster was defined as a cluster associated with a healthcare or hospital setting. a household cluster was defined as a cluster associated with the same family and/or physical household. case identification and data collection were performed on an ongoing basis by epidemiologists at afhsb beginning with the emergence of the mers-cov outbreak in 2012. case demographics, including city and country of origin, age, gender, date of symptom onset (if any), asymptomatic status, mortality, comorbidities, healthcare worker (hcw) status, and date reported, were collected on a daily basis. each case was verified using tiers 1 and 2 sources. if a tier 1 or 2 source failed to verify a case reported by a tier 3 source, it was not included in the afhsb case line list. if an epidemiological link between cases was identified through a tier 3 source, tiers 1 and 2 sources were used to verify the link. if a tier 1 or 2 source was not available, supporting data from at least three separate tier 3 sources were used as verification. due to the lack of data available at the local level in the arabian peninsula, in the event that multiple ongoing nosocomial outbreaks were known to have been occurring in one area, all cases reported to have been associated with that area and possibly epidemiologically linked to one of the ongoing clusters were categorized under one cluster (e.g. riyadh, jeddah). for all exported mers-cov clusters, the city and country of origin were determined by the reported travel history of the index case. if travel history was only available in the country level, the capital city of the country of travel was used as the point of origin for the index case. for two identified clusters, the index case had to travel to multiple countries with a history of confirmed autochthonous mers-cov transmission. as a result, the country of most probable exposure, determined by the duration of stay in the country as well as active transmission reported in that country at the time of travel, was used as the point of origin. in order to visually display the overlap of clusters on the epidemiological curve, the start and end dates of each identified cluster were defined. the start date of each cluster was the date of symptom onset of the index case. in the absence of symptom-onset data, the 'report date', or the date a case was publically reported, was used instead. the end date of each cluster was determined by adding 14 days to the date of symptom onset or date of death of the last case identified in the cluster. the 14-day period is representative of the maximum incubation period of a mers-cov case, ensuring no additional cases could have been associated with a given cluster [10] . for asymptomatic cases, date of diagnosis was used in place of date of symptom onset. if the date of diagnosis was not publically available, date reported was used. april 2012 and june 2016, 817 (45·2%) cases were determined to have been associated with at least one of the 51 clusters identified in this analysis [2, 11] . a small portion of cases associated with one or more clusters were hcws (n = 159), and 106 clusterassociated cases were asymptomatic ( from another mers-cov case, including 41 nosocomial infections and 13 household infections. as the ksa moh did not specify which mers-cov case these were acquired from, it is unclear if any of these batched cases were associated with the 51 identified mers-cov clusters or were part of separate clusters. of the 51 identified clusters, 41 (80·4%) were classified as nosocomial clusters; 12 (23·5%) were household clusters; and eight (15·7%) were exported (table 3 ). ten clusters were classified as more than one type of cluster, including four exported nosocomial clusters, three exported household clusters, and three clusters with both nosocomial and household characteristics. three clusters displayed both nosocomial and household transmission characteristics, four clusters were classified as both exported and nosocomial, and three clusters were classified as both exported and household. the two countries reporting the greatest number of nosocomial clusters were ksa and rok, with 18 and 15 nosocomial clusters, respectively. of the eight exported clusters, two (rok1, tunisia1) had index cases with travel to multiple countries with a history of confirmed autochthonous mers-cov transmission (see table 4 , technical appendix). the average duration of each cluster was 44·4 days, with a range of 14-119 days. cluster size ranged from two cases to 182 cases, with an average of 16 individuals affected. over 90% of the cluster-associated cases were acquired in ksa (n = 558, 68·3%) and rok (n = 186, 22·8%), supporting the notion that nosocomial transmission, which accounted for 100% of the clusters identified in rok and 78% of the clusters identified in ksa, was a prominent driver of clusters in this epidemic. figure 1 depicts the epidemiological curve of the mers-cov outbreak using the estimated epidemiological week of illness onset for each case. symptom-onset date was available for 1267 cases (70·2%), including 623 (72·6%) of the clusterassociated cases. for asymptomatic cluster-associated cases, date reported was used for 98 of the cases, and date of diagnosis was used for eight cases. the cluster bands seen above the epidemiological curve in figure 1 illustrate the duration of transmission within each cluster and the corresponding overlap of identified clusters with peak mers-cov incidence over the epidemic period of this study. most temporal peaks correspond with at least two ongoing clusters (see fig. 1 ). the time periods of greatest incidence (>35 incident cases at the period's peak) in 2014 and 2015 correspond with at least four ongoing mers-cov clusters, of which at least two were classified nosocomial in each of these periods. see table 4 (technical appendix) for individual cluster data, such as cluster duration and a number of cases affected. the 51 clusters identified in this analysis spanned the duration of the epidemiological curve, demonstrating the consistency of cluster-driven transmission throughout the epidemic. clusters also corresponded with each period of increased mers-cov incidence and accounted for over 45% of the total confirmed cases reported, supporting the notion that human-tohuman transmission is a prominent driver of the mers-cov epidemic. the alignment of nosocomial clusters with the time periods of greatest incidence (>35 incident cases at the period's peak) suggests nosocomial transmission is a key driver of transmission in the mers-cov epidemic as opposed to seasonal events, such as the hajj or camel-breeding season that occur in the fall and spring, respectively. the absence of apparent seasonality in the identified clusters further supports the significant role humanto-human transmission in healthcare settings plays in the propagation of mers-cov. classification of each cluster into one or more of the three disease transmission categories (nosocomial, household, or exported) elucidated common characteristics among the clusters, including the overall concentration of case clusters in cities and healthcare institutions. with nosocomial transmission accounting for 41 of the identified clusters (80·4%), the role of healthcare facilities, transportation protocols during inter-and intra-hospital transfers, and the contribution of cultural norms such as 'doctor shopping' became apparent themes observed throughout this analysis. in rok, practices such as seeking healthcare at multiple facilities, or 'doctor shopping', and being cared for by relatives while hospitalized are believed to have contributed to the emergence of the largest outbreak outside of ksa during the mers-cov epidemic [12, 13] . additionally, four superspreaders, defined as a confirmed mers-cov case responsible for infecting 10 or more secondary cases, played a critical role in the perpetuation of transmission in rok. the mers-cov outbreak in rok included 15 clusters at separate healthcare institutions, each with an index case related to an ongoing mers-cov cluster at another facility. between may 2015 and july 2015, 186 laboratory-confirmed mers-cov cases were reported. the geographic distribution and rapid pace at which mers-cov spread in rok demonstrates the susceptibility of healthcare environments to human-to-human transmission of the virus and their catalytic role in mers-cov transmission. in ksa, clusters were primarily concentrated in cities, specifically active transmission at healthcare facilities within those cities, as illustrated by the largest identified clusters in the country (ksa8, ksa9, ksa17, ksa20, and ksa23; see table 4 , technical appendix). the most common theme in the perpetuation of transmission among nosocomial clusters was the role of more transient departments in the healthcare setting, such as dialysis units and emergency departments, which were implicated in a majority of the nosocomial clusters identified in this analysis. at least seven of the 23 clusters identified in ksa occurred in designated mers-cov treatment centers, which were presumably best equipped to handle these infectious cases. however, the nosocomial transmission that persisted in these clusters suggests inconsistent infection control practices across different departments of those designated hospitals [14] . hcws likely contributed to the continued transmission among patients and between wards within a hospital, as hcws represent 159 (19·5%) of the clusterassociated cases. of the 41 clusters classified as nosocomial, 30 clusters (73·2%) involved hcws. the use of tier 1 data collected as the mers-cov outbreak progressed is one of the greatest strengths of this comprehensive cluster analysis. collection of the data and emerging literature in real time allowed for an inclusive literature review from which cluster identification and evaluation of key components of the outbreak could be analyzed. the availability of opensource data and information provided the opportunity to precisely map each confirmed case temporally and geographically. utilization of the date of symptom onset for the epidemiological curve created an accurate representation of the progression of the epidemic and corresponding cluster durations (fig. 1) . the availability of open-source data from certain regions also served as a limitation in this analysis. due to limited demographic information, it was occasionally necessary to broaden the case criteria for a particular cluster, specifically clusters ksa8, ksa9, and ksa20 (see table 4 , technical appendix). if a case was reported from the city during the estimated time in which there was ongoing nosocomial transmission, had no travel or camel exposure in the 14 days prior to illness onset, and had no known household contact with a confirmed mers-cov case, the case was included in the case count for that particular nosocomial cluster. although the particular epidemiological details regarding the exact location of exposure were generally not provided for cases during the three aforementioned clusters, tier 1 sources often denoted if a case was under investigation for a possible link to a hospital with the known ongoing transmission of mers-cov, which reinforced the inclusion of these cases in their respective clusters. in addition to limitations on precise case inclusion in the larger clusters denoted above, availability of date of symptom onset may have created an artifact in the cluster date and duration, potentially altering its appearance on temporal scales. of the known symptomatic cases, illness onset data were missing for 334 cases, including 96 cluster-associated cases, and date reported was used as an approximation. on average, afhsb observed an 8-9-day lag from the date of symptom onset to the date a case was reported. considering our analysis aggregated these cases by estimated epidemiological week of illness onset, the impact of this possible artifact is likely to be relatively insignificant to the overall trends observed in this analysis. additionally, ksa retrospectively released information on 113 confirmed mers-cov cases on 3 june 2014 and on 16 cases on 19 september 2014 with minimal geographic and demographic data. all cases released on 3 june 2014 occurred between 5 may 2013 and 6 may 2014, with a majority of the cases (n = 84) occurring after 1 march 2014. the rest of the cases (n = 29) occurred between 5 may 2013 and 28 february 2014. as no other date was available, the date reported was used to represent the cases released in these two batches on the epidemiological curve. asymptomatic cases accounted for 11·4% (n = 205) of the total cases in this study population. over 50% (n = 106) of these asymptomatic cases were related to a cluster. this finding may support who's assessment that contact tracing efforts intensified as the epidemic progressed and are responsible for the detection of asymptomatic cases [15, 16] . a study performed by cdc analyzing case data between september 2012 and january 2016 found that there was likely an underrepresentation of asymptomatic cases reported from the countries of the arabian peninsula. estimations suggest that the total number of mers-cov cases from the region may be 2·3 times greater than the total number of cases recorded to date [17] . inconsistencies in reporting of asymptomatic cases from entities such as the ksa moh may have contributed to an underrepresentation of not only the total number of mers-cov cases in the outbreak but also the number of clusters, as well as the breadth and duration of the identified mers-cov clusters. the supplementary material for this article can be found at https://doi.org/10.1017/s0950268817002345. middle east respiratory syndrome (mers): prevention & treatment centers for disease control and prevention military-health-topics/ health-readiness/armed-forces-health-surveillance-branch/integrated-biosurveillance/surveillance-summaries) world health organization. a roadmap for research and product development against middle east respiratory syndrome-coronavirus (mers-cov) middle east respiratory syndrome coronavirus (mers-cov) origin and animal reservoir dromedary camels and middle east respiratory syndrome: mers coronavirus in the 'ship of the desert' co-circulation of three camel coronavirus species and recombination of mers-covs in saudi arabia acute middle east respiratory syndrome coronavirus infection in livestock dromedaries guidance for monitoring and movement of persons with potential middle east respiratory syndrome coronavirus (mers-cov) exposure surveillance for human infection with middle east respiratory syndrome coronavirus (mers-cov): interim guidance novel coronavirus infection update middle east respiratory syndrome (mers) in the republic of korea middle east respiratory syndrome coronavirus outbreak in the republic of korea infection prevention and control guidelines for the middle east respiratory syndrome coronavirus (mers-cov) infection moh's command and control center forms rapid response teams including 120 health specialists middle east respiratory syndrome coronavirus (mers-cov) global summary and risk assessment accessed 23 estimation of severe middle east respiratory syndrome cases in the middle east the authors acknowledge the support provided by afhsb's us government partners, including the centers for disease control and prevention and the united states forces korea. this study received no specific grant from any funding agency, commercial, or not-for-profit sectors. none. key: cord-288102-iom6lu7o authors: han, jing; shi, li-xia; xie, yi; zhang, yong-jin; huang, shu-ping; li, jian-guo; wang, he-rong; shao, shi-feng title: analysis of factors affecting the prognosis of covid-19 patients and viral shedding duration date: 2020-06-25 journal: epidemiol infect doi: 10.1017/s0950268820001399 sha: doc_id: 288102 cord_uid: iom6lu7o the clinical characteristics of patients with covid-19 were analysed to determine the factors influencing the prognosis and virus shedding time to facilitate early detection of disease progression. logistic regression analysis was used to explore the relationships among prognosis, clinical characteristics and laboratory indexes. the predictive value of this model was assessed with receiver operating characteristic curve analysis, calibration and internal validation. the viral shedding duration was calculated using the kaplan–meier method, and the prognostic factors were analysed by univariate log-rank analysis and the cox proportional hazards model. a retrospective study was carried out with patients with covid-19 in tianjin, china. a total of 185 patients were included, 27 (14.59%) of whom were severely ill at the time of discharge and three (1.6%) of whom died. our findings demonstrate that patients with an advanced age, diabetes, a low pao(2)/fio(2) value and delayed treatment should be carefully monitored for disease progression to reduce the incidence of severe disease. hypoproteinaemia and the fever duration warrant special attention. timely interventions in symptomatic patients and a time from symptom onset to treatment <4 days can shorten the duration of viral shedding. novel coronavirus disease (covid19) can cause inflammation in the lungs, and the disease develops rapidly. as of 31 march, the severity and mortality rates of patients with covid-19 in china were 23.25% and 4.06%, respectively [1] . a report from wuhan showed that the mortality rates were 61.5% among critically ill patients with covid-19 and 71% among those requiring mechanical ventilation [2] . during this pandemic, effective management of patients with severe covid-19 is crucial to reduce mortality among the infected population. viral shedding is one of the most important indicators of cure, and current reports have rarely analysed the duration of viral shedding. the clinical characteristics of 185 patients with covid-19 diagnosed in tianjin were analysed retrospectively to determine the factors affecting their prognoses and the duration of viral shedding with the aim of facilitating early treatment and improving patient prognosis. from 21 january to 8 may 2020, a retrospective study was conducted with all patients (⩾14 years) admitted to the hospital in tianjin who were discharged after receiving a confirmed diagnosis of covid-19. a total of 185 patients were finally included in this retrospective study. data were extracted from the hospital information system. standardised forms were used to collect patients' clinical characteristics, including age, sex, comorbidities, former/current smoking, current drinking, the time from symptom onset to treatment, clinical symptoms, body temperature at admission, the pao 2 /fio 2 ratio and complications. imaging and laboratory examination results within 24 h of admission were collected, including ct examination results, routine blood examination results, c-reactive protein (crp) levels, myoglobin (myo) levels and other laboratory examination results. according to china's novel coronavirus pneumonia diagnosis and treatment plan (seventh edition) [3] , the clinical classifications are mild, moderate, severe and critical. mild type: the clinical symptoms are mild, with no manifestations of pneumonia on imaging. moderate type: patients have fever, respiratory tract symptoms and other symptoms; imaging can show signs of pneumonia. severe type: in adults, one of the following conditions must be met: (1) shortness of breath and a respiratory rate ⩾30 breaths/min; (2) oxygen saturation levels measured with a finger pulse oximeter at rest ⩽93%; or (3) pao 2 /fio 2 ⩽300 mmhg. critical type: one of the following conditions must be met: (1) respiratory failure requiring mechanical ventilation; (2) shock; or (3) extrapulmonary organ failure requiring intensive care unit monitoring and treatment. patients were divided into a good prognosis group and a poor prognosis group according to their clinical classification at discharge. patients with mild and moderate covid-19 were included in the good prognosis group, and patients with severe and critical covid-19 and those who had died were included in the poor prognosis group. patients who met the following conditions were discharged: (1) body temperature had returned to normal (<37.3°c) and had remained normal for more than 3 days; (2) respiratory symptoms had improved significantly; (3) pulmonary imaging showed that acute exudative lesions had improved significantly; and (4) two consecutive nucleic acid tests on respiratory samples, such as sputum and nasopharyngeal swabs, were negative (a sampling interval of at least 24 h). the numerical variables with normal distributions are expressed as x + s, and comparisons between the two groups were performed with t tests. continuous variables with non-normal distributions are represented as the median (quartile, q), and non-parametric tests were used to compare the two groups. count data are represented as n (%) and were compared with the χ 2 test. variables (age, diabetes, pao 2 /fio 2 on admission, nlr and platelet count) with significant differences on univariate analysis and those with clinical credibility were included in a multiple factor regression [4] . multivariate logistic regression was used to evaluate the risk factors associated with a poor covid-19 prognosis. the results are presented as odds ratios (ors) and 95% confidence intervals (cis). to evaluate the discriminative performance of the logistic model, the area under the receiver operating characteristic (roc) curve was calculated, comparing the actual outcome to the outcome predicted by the model. in the univariate analysis, we used the log-rank test and kaplan-meier curve analysis for categorical variables and cox regression analysis for continuous variables. the variables with p < 0.05 in the univariate analysis and variables with practical significance were included in the cox proportional risk model, and the factors affecting the duration of viral shedding were identified through multivariate analysis. the baseline characteristics of all 185 patients are shown in table 1 . a total of 16.2% of the 185 patients had a poor prognosis. three non-survivors were included in the poor prognosis group. the mean age of the 185 patients was 44 ± 17.88 years, and 51.4% (95 patients) were male. the mean bmi of the patients was 24.61 ± 3.79 kg/m 2 . in this study, blood group b was the most common, accounting for 33.7% of the sample. former/current smokers accounted for 12.4% (23 patients) of the patients, and 23.2% (43 patients) of the patients currently consumed alcohol. a total of 35.7% (66 patients) of the patients had one or more comorbidities, the most common of which were hypertension, diabetes and coronary heart disease (chd). during hospitalisation, 9.2% (17 patients) of the patients had cardiac insufficiency, 31.9% (59 patients) of the patients had bacterial pneumonia and 12.4% (23 patients) of the patients had hypoproteinaemia. the median time from symptom onset to treatment was 4 (5) days. a total of 185 patients were divided into two groups according to prognosis: 155 patients were in the good prognosis group, and 30 patients were in the poor prognosis group. significant differences were identified between the two groups in terms of age (p < 0.001), comorbidities (p < 0.001), diabetes (p < 0.001), hypertension (p < 0.001), chd (p < 0.001), cardiac insufficiency (p < 0.001), hypoproteinaemia (p < 0.001), bacterial pneumonia (p < 0.001) and the time from symptom onset to treatment (p < 0.05). the physical signs and laboratory examinations of all 185 patients are shown in table 2 . among the patients included in this study, 74.6% (138 patients) had fever, and 51.4% (95 patients) had cough, which are the common symptoms of covid-19. the average temperature of the patients was 36.97 ± 0.81°c. no significant difference in temperature was found between the two groups (p > 0.05). in total, 8.1% (15 patients) of the patients were asymptomatic, and most of them were classified as having had mild covid-19 at discharge. the average pao 2 /fio 2 ratio was 427.01 ± 171.05 mmhg. the value in the poor prognosis group was lower than that in the good prognosis group. a significant difference was detected between the two groups (p < 0.001). the median number of lung lobes involved in ct was 3 (4). the number of lung lobes involved in the poor prognosis group was significantly higher than that in the good prognosis group (p < 0.001). significant differences were found between the two groups in terms of the lymphocyte count (p < 0.05), the nlr (p < 0.05), plt (p < 0.05), and the levels of crp (p < 0.05), myo (p < 0.001) and d-dimer (p < 0.05) within 24 h of admission. in the multivariate regression analysis, age (or = 1.089; 95% ci 1.046-1.133; p < 0.001), diabetes (or = 3.311; 95% ci 1.093-10.031; p < 0.05), the time from symptom onset to treatment (or = 1.185; 95% ci 1.042-1.347; p < 0.05) and pao 2 /fio 2 (or = 0.994; 95% ci 0.989-0.998; p < 0.05) were statistically significant (table 3) . we performed an roc curve analysis with the variables identified in the multivariate analysis to predict progression to severe disease in patients with covid-19. the model showed good discrimination (fig. 1) , with an area under the roc curve of 0.909 (95% ci 0.865-0.954), suggesting that these variables can be used to predict a poor prognosis of the disease. jing han et al. the median duration of viral shedding after covid-19 onset and hospitalisation was similar between the good and poor prognosis groups. the median duration of viral shedding was 17 (12) days from illness onset; the longest duration was 51 days, and the shortest duration was 4 days. the median duration of hospitalisation was 14 days. the median time to fever resolution was 3 days. the time to resolution of fever in the good prognosis group was 2 days, which was significantly shorter than that in the poor prognosis group (table 4 ). using survival curve analysis, the viral shedding durations were compared among patients with covid-19 with different clinical characteristics, and differences in the survival curves were analysed by the log-rank test. kaplan-meier analysis was used to evaluate the effects of age, comorbidities, hypoproteinaemia, bacterial pneumonia and other variables on the viral shedding duration. a cox regression model was used to analyse the levels of crp and ck. the results are shown in table 5 . a multivariate cox proportional hazard model was used to analyse whether age, hypoproteinaemia, the time from symptom onset to treatment, the time to fever resolution, the presence of symptoms, treatment with corticosteroids, crp levels and ck levels affected whether the covid-19 patients became negative for viral nucleic acid in the treatment period (0 = no, 1 = became negative). the results showed that hypoalbuminemia (hr = 0.514; 95% ci 0.31-0.852; p < 0.05), a time from symptom onset to treatment >4 days (hr = 0.68; 95% ci 0.5-0.925; p < 0.05), a time to fever resolution >3 days (hr = 0.537; 95% ci 0.392-0.734; p < 0.05) and symptomatic status (hr = 0.338; 95% ci 0.189-0.605; p < 0.05) were independent factors influencing the viral shedding duration, as shown in table 5 . the effects on the viral shedding duration of different groups of variables in patients with covid-19 are shown in figures 2-5. the world health organization emergency committee for international health regulations declared covid-19 a pandemic and a public health emergency of international concern. covid-19 is an acute infectious respiratory disease, and patients with covid-19 can worsen rapidly, progressing to acute respiratory distress syndrome (ards), septic shock, metabolic acidosis and coagulation dysfunction, which are difficult to treat. huang et al. showed that the time from the onset of covid-19 to the development of dyspnoea was 8 days, and that progression to ards occurred in 9 days [5] . our study also found that the time from symptom onset to treatment was an independent risk factor for severe disease. a time from symptom onset to treatment >4 days was an independent factor influencing the viral shedding duration. if patients are diagnosed and treated in a timely manner, the severity of the disease can be predicted, which has important clinical significance for medical staff who are diagnosing and treating patients. in this study, at the time of discharge, the number of patients with severe disease was 27, and the proportion of patients with severe disease had decreased to 14.59%, which was lower than the national average reported by the national health commission of the people's republic of china [1] and the value reported in the study by guan et al. [6] . based on an epidemiological study of 72 314 patients with covid-19, the mortality rate of patients with comorbidities was higher than that of patients without comorbidities, and the mortality rate of patients with diabetes was 7.3% [7] . the results showed that age and diabetes were independent risk factors for a poor prognosis. according to the current report, the proportion of covid-19 patients with diabetes mellitus is 10.1-20%, and the proportion of critical covid-19 patients with diabetes mellitus is 22.2% [5, 8] . the report by klekotka et al. [9] pointed out that diabetes increases the risk of respiratory tract infection and is an important risk factor for the aggravation of lower respiratory tract infection. patients with diabetes often have abnormal immune function, such as fewer immune cells and decreased nkt cell activity, rendering these patients a high-risk group for viral infections with an increased risk of severe disease [10, 11] . diabetes leads to a chronic inflammatory response, and long-term hyperglycaemia leads to vascular endothelial cell damage, thus reducing the patient's immune status. some studies have pointed out that the abnormal pro-inflammatory cytokine response in diabetic patients may result in severe covid-19 [12] [13] [14] . in mehta et al.'s [15] study, in patients with diabetes mellitus and covid-19, the levels of markers such as crp, fibrinogen and d-dimer were found to be elevated. this may be due to cytokine storms, which increase the risk of severe covid-19 and result in a poor prognosis. therefore, for patients with covid-19, early diagnosis and intervention are important to reduce the risk of death caused by chronic underlying diseases such as diabetes. age was also confirmed to be an important independent predictor of mortality in mers [16] and sars [17] . zhou et al. reported that in-hospital death was related to age at admission [18] . with increasing age, the prognosis deteriorates, especially among the elderly population, due to the decline in immune organ function and the combination of chronic diseases. under a certain degree of hypoxia, the heart, lung and kidney function of very elderly patients worsens, making them prone to multiple organ failure and increasing the risk of mortality [19] . the results showed that most patients with covid-19 exhibited symptoms of fever and cough, which is consistent with the clinical signs in sars and mers patients [20] . a few patients had diarrhoea, but the study found that the clinical symptoms at admission could not be used to predict the severity of the disease. covid-19 is a self-limiting disease, with most infected patients having mild cases [8] . however, in our study, more than 50% of the patients were found to have the involvement of five lung lobes on ct at admission. in a study by shi et al. [21] , they found lung abnormalities on ct scans in 15 patients with asymptomatic infections. therefore, patients with symptoms should visit a physician in a timely manner and undergo ct scans. according to the inspection results, the disease severity predicted the prognosis of the patients. we found that a lower pao 2 /fio 2 at the time of admission is a risk factor for a poor prognosis in patients with severe covid-19. according to yang et al. [2] , the substantial difference in the pao 2 /fio 2 ratio between survivors and non-survivors indicated that the pao 2 /fio 2 ratio is associated with the severity of illness and prognosis. therefore, we should pay attention to the pao 2 /fio 2 index and provide respiratory support and circulatory support in a timely manner. the results showed that the time from admission to a normal temperature was 7 days for patients with severe disease and 2 days for patients with mild disease. therefore, in the treatment of patients with covid-19, we should pay attention to the duration of fever. for patients with a long fever duration, we should intervene in a timely manner to improve their prognosis. the level and duration of infectious virus replication is an important factor in assessing the risk of transmission. viral shedding is one of the most important criteria for the treatment of patients with covid-19. unfortunately, the pathogenesis of covid-19 is not yet clear. sars-cov-2 is similar to sars-cov from 2003 and mers-cov from 2012, which mainly infect alveolar epithelial cells [22] . for patients with sars-cov infection, the positive rate of respiratory specimens peaked at 6-11 days after onset. more than 23 days later, respiratory specimens still showed positivity for the virus [23] . one-third of patients tested positive in respiratory specimens within 4 weeks [24] . the duration of positivity for mers-cov in respiratory tract samples lasted at least 3 weeks [25, 26] . in a recently published study, zhou et al. [18] reported that the duration of viral shedding was 20 days. our study showed a duration of viral shedding of 17 days, which is slightly shorter than the reported duration. it was found that the absence of hypoalbuminemia, time to fever resolution >3 days, time from symptom onset to treatment >4 days and symptomatic status were independent factors influencing the viral shedding duration. this finding is highly epidemiology and infection significant for the treatment of patients, the management of earlystage disease and the prevention and control of hospital infections. hu et al. [27] investigated the shedding duration of sars-cov-2 and found that age is an independent risk factor affecting the viral shedding duration. due to the low immune status of elderly individuals, it is more difficult for them to eradicate invasive pathogens. in our study, univariate analysis results showed that age was a risk factor for prolonged viral shedding duration, but no significant difference was found in the multivariate cox analysis. hypoproteinaemia can cause microcirculation disturbances and lead to insufficient perfusion of important organs and multiple organ dysfunction. moreover, a decrease in the level of serum albumin results in decreases in the levels of various enzymes related to antibody synthesis. this decrease in enzyme activity leads to a decrease in immunity and an increase in the likelihood of infection. at the same time, hypoproteinaemia can lead to respiratory muscle atrophy, thus reducing the body's resistance and increasing the length of hospital stay and the frequency of readmission. therefore, it takes longer for a lung infection to heal, affecting ventilation function, increasing the probability of multiple organ dysfunction and increasing mortality [28] [29] [30] . in a study in macaques, joseph [31] found that immunosuppressed macaques exhibited significantly higher levels of mers-cov replication in respiratory tissues and shed more virions. therefore, in patients with covid-19, the timely correction of hypoproteinaemia can shorten the duration of viral shedding. due to the limitations of this retrospective study, our study on hypoproteinaemia is lacking in detail. we will further investigate the relationship between hypoproteinaemia and viral shedding duration in future research. since viral load detection was not carried out in the early stages, we will further study the relationship between viral load and prognosis in the future. in a study on mers, memish [32] noted that the time of virus clearance in asymptomatic patients was earlier than that in symptomatic patients, which is concordant with the findings reported in this study. however, we were unable to demonstrate a correlation between the duration of viral shedding and prognosis in this work. in previous studies on the shedding duration of the h7n9 virus [33] , the time from symptom onset to treatment was also found to be an independent risk factor. currently, no specific antiviral drug is available for covid-19. we administer antiviral treatment with drugs recommended by national diagnosis and treatment guidelines [3] . patients with mild symptoms were treated with α-interferon aerosol inhalation and oral umifenovir (arbidol). umifenovir is a russian-made small indole-derivative molecule licensed in russia and china to prevent and treat influenza and other viral infections. for the treatment of patients with covid-19, umifenovir is the recommended drug in chinese diagnosis and treatment guidelines. severe patients were treated with lopinavir or ritonavir. approximately 90% of patients accepted treatment with a chinese medicine decoction. our study cannot analyse the impact factors of drugs on the viral jing han et al. shedding duration. however, early treatment is advantageous for improving the prognosis and shortening the viral shedding time. all covid-19 patients (⩾14 years) in tianjin were included in this study. this study represents the clinical characteristics of patients in a region. this study was a retrospective study with a relatively small sample size and was performed in a single centre; however, we anticipate that our study will be of significant interest given the importance of predicting patient prognoses for this disease and promoting clinical work. one of the limitations of this study is that we assessed patients for only a limited time when they were hospitalised; thus, a longer follow-up period might be needed to further assess the prognosis of and viral shedding in cured covid-19 patients. future studies with longer follow-up periods and larger sample sizes and studies conducted at multiple centres are needed. at present, no effective treatment is available for covid-19; therefore, we analysed clinical patient data to determine the factors affecting the prognosis of the disease. this information may be used to facilitate early intervention and treatment, thereby reducing the incidence of and mortality due to severe covid-19. this study showed that diabetes mellitus, age, the time from symptom onset to treatment and pao 2 /fio 2 can predict the prognosis of patients with covid-19. hypoproteinaemia and the fever duration warrant special attention. timely intervention in patients with symptoms and a time from symptom onset to treatment <4 days can shorten the duration of viral shedding. national health commission of the people's republic of china database kaplan-meier plot for fever resolution time clinical course and outcomes of critically ill patients with sars-cov-2 pneumonia in wuhan, china: a single-centered, retrospective, observational study national health commission of the people's republic of china database relaxing the rule of ten events per variable in logistic and cox regression clinical features of patients infected with 2019 novel coronavirus in wuhan clinical characteristics of coronavirus disease 2019 in china novel coronavirus pneumonia emergency response epidemiology team (2020) the epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (covid-19) in china epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study the etiology of lower respiratory tract infections in people with diabetes diabetes and the severity of pandemic influenza a (h1n1) infection understanding sars-cov-2-mediated inflammatory responses: from mechanisms to potential therapeutic tools covid-19, diabetes mellitus and ace2: the conundrum immune dysfunction in patients with diabetes mellitus (dm) covid-19 and diabetes mellitus: unveiling the interaction of two pandemics covid-19: consider cytokine storm syndromes and immunosuppression predictors of mortality in middle east respiratory syndrome (mers) outcomes and prognostic factors in 267 patients with severe acute respiratory syndrome in hong kong clinical course and risk factors for mortality of adult inpatients with covid-19 in wuhan, china: a retrospective cohort study weakened immunity in aged hosts with comorbidities as a risk factor for the emergence of influenza a h7n9 mutants mers, sars and other coronaviruses as causes of pneumonia radiological findings from 81 patients with covid-19 pneumonia in wuhan. china: a descriptive study what we learned from sars may provide important insights into understanding and management of the novel coronavirus disease 2019 viral shedding patterns of coronavirus in patients with probable severe acute respiratory syndrome persistent shedding of viable sars-cov in urine and stool of sars patients during the convalescent phase viral shedding and antibody response in 37 patients with middle east respiratory syndrome coronavirus infection viral load kinetics of mers coronavirus infection factors associated with negative conversion of viral rna in patients hospitalized with covid-19 impact of preceding respiratory viral infections on the clinical severity of patients with pneumococcal pneumonia. influenza & other respiratory viruses c-reactive protein and albumin kinetics after antibiotic therapy in community-acquired bloodstream infection human serum albumin homeostasis: a new look at the roles of synthesis, catabolism, renal and gastrointestinal excretion, and the clinical value of serum albumin measurements pathogenicity and viral shedding of mers-cov in immunocompromised rhesus macaques middle east respiratory syndrome coronavirus (mers-cov) viral shedding in the respiratory tract: an observational analysis with infection control implications factors associated with prolonged viral shedding in patients with avian influenza a(h7n9) virus infection acknowledgement. we thank all our colleagues who helped us with the current study. we are also grateful to many front-line medical staff for their dedication in the face of this outbreak, despite the potential threat to their own lives and the lives of their families. financial support. this research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.conflict of interest. the authors declare that they have no competing interests.data availability statement. the datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. key: cord-327701-1qgaxcqq authors: scott, e. m.; magaret, a.; kuypers, j.; tielsch, j. m.; katz, j.; khatry, s. k.; stewart, l.; shrestha, l.; leclerq, s. c.; englund, j. a.; chu, h. y. title: risk factors and patterns of household clusters of respiratory viruses in rural nepal date: 2019-10-14 journal: epidemiol infect doi: 10.1017/s0950268819001754 sha: doc_id: 327701 cord_uid: 1qgaxcqq viral pneumonia is an important cause of death and morbidity among infants worldwide. transmission of non-influenza respiratory viruses in households can inform preventative interventions and has not been well-characterised in south asia. from april 2011 to april 2012, household members of pregnant women enrolled in a randomised trial of influenza vaccine in rural nepal were surveyed weekly for respiratory illness until 180 days after birth. nasal swabs were tested by polymerase chain reaction for respiratory viruses in symptomatic individuals. a transmission event was defined as a secondary case of the same virus within 14 days of initial infection within a household. from 555 households, 825 initial viral illness episodes occurred, resulting in 79 transmission events. the overall incidence of transmission was 1.14 events per 100 person-weeks. risk of transmission incidence was associated with an index case age 1–4 years (incidence rate ratio (irr) 2.35; 95% confidence interval (ci) 1.40–3.96), coinfection as initial infection (irr 1.94; 95% ci 1.05–3.61) and no electricity in household (irr 2.70; 95% ci 1.41–5.00). preventive interventions targeting preschool-age children in households in resource-limited settings may decrease the risk of transmission to vulnerable household members, such as young infants. acute lower respiratory infection (alri) is the primary cause of child morbidity and mortality worldwide with the vast majority of childhood deaths related to alri occurring in resourcelimited settings [1] . respiratory viruses are increasingly recognised as a cause of severe alri in young children [2] . in many global regions where access to healthcare is limited, especially in rural areas, the true community-based burden of respiratory virus-associated alri remains poorly characterised [3] [4] [5] . in these settings, household surveillance studies can provide a more comprehensive evaluation of viral incidence, transmission and molecular epidemiology patterns in the community [4] [5] [6] [7] [8] . household surveillance can provide valuable information regarding the transmission networks within households. such knowledge may guide the development and implementation of preventative interventions to protect vulnerable groups from alri. for example, infants are at highest risk for severe alri from respiratory syncytial virus (rsv) [9] . major challenges to developing a safe and effective rsv vaccine in young infants have resulted in the development of alternative strategies including maternal rsv vaccination and delayed vaccine administration until >6 months of age [10] . targeting older groups for vaccination may protect vulnerable populations by interfering in transmission chains to young infants, the elderly and other high-risk groups. studies in rural kenya have identified school-age children as the primary introducers of rsv into households where an infant subsequently became infected [6] . these results were in agreement with a us study from the 1960s reporting older siblings aged 2-16 years as most likely to introduce rsv disease into families [11] . in contrast, modelling suggests that young children <5 years are more likely to transmit rsv and are the most efficient population to vaccinate in order to prevent disease in other groups [12] . few studies have analysed the transmission of other non-influenza respiratory viruses, such as human metapneumovirus (mpv) and human rhinovirus (hrv), and no studies have examined the household transmission dynamics of respiratory viruses in rural south asia, a region characterised by high rates of preterm birth and infant mortality [8, [13] [14] [15] . the aims of this analysis were to characterise the transmission of nine non-influenza respiratory viruses within households in rural nepal and to determine household characteristics associated with the transmission of respiratory viruses. we hypothesise that the presence of school-age children in a household will be associated with an increased risk of transmission. this prospective household surveillance study was nested within a randomised controlled trial designed to determine the effectiveness of influenza vaccine during pregnancy [16, 17] . the study site is in the low-lying region of rural southern nepal called the 'terai', with inhabitants broadly representative of the population of india, bangladesh and nepal [18] . all women of childbearing age in a part of one district were surveyed for pregnancy. pregnant women were enrolled in the primary trial as early as possible in pregnancy, generally during the second trimester, and followed until 6 months postpartum. at the time of randomisation, every third study mother and their families were selected to participate in a household surveillance substudy. as randomisation occurred at the time of vaccination, not initial enrolment, randomisation occurred after enrolment and the start of surveillance for some mothers. surveillance for the first participant enrolled in the household substudy began on 14 april 2011 and we included the households of substudy mothers enrolled prior to 1 may 2012. surveillance of the household ended 180 days after birth. in this area, many households consist of multiple families living in a single compound; households were defined as a group sharing a single cookstove. socio-demographic data were collected upon enrolment at the individual and household levels. birth assessments of study infants were performed shortly following birth. infants weighed within 72 h of the birth were considered to be low birthweight if the infant weighed <2.5 kg. trained field staff visited the home weekly and used a standardised form to inquire about respiratory symptoms and signs in mothers, infants and other household members for each day in the previous week. a mid-nasal turbinate swab was collected from mothers and other adult household members aged ⩾15 years with self-reported fever, plus one or more of the following symptoms within the previous 7 days: cough, sore throat, runny nose, nasal congestion or myalgias. swabs were collected from all children <15 years with at least one of the following symptoms: subjective fever, cough, draining ear, wheezing or difficulty breathing, in the previous 7 days. illness episodes were defined as symptoms that met described criteria and were separated by at least seven symptom-free days. only individuals with ⩾7 days of symptom diary recorded, with or without illness, were included in the analyses. households that did not have ⩾3 individuals with surveillance were excluded from the analysis as two-person households consisted of mother-infant pairs without surveillance of household members. respiratory swabs were collected, aliquoted and transported from the nepal field site to the university of washington in seattle, wa in a temperature-stable buffer (primestore; longhorn diagnostics, san antonio, usa). samples were tested by a real-time reverse transcriptase polymerase chain reaction (pcr) for 12 respiratory viruses, including rsv, mpv, influenza viruses a and b, parainfluenza virus 1-4 (piv 1-4), adenovirus (adv), human coronavirus (cov), hrv and bocavirus [19] [20] [21] . influenza transmission in household was the primary aim of the trial substudy and is being analysed separately. bocavirus was not included because of its prolonged shedding patterns. sequencing was performed for hrv-and rsv-positive samples from household illness clusters utilizing samples with pcr cycle threshold values <33 and 30 for hrv and rsv, respectively, based on previous difficulty sequencing low viral load samples [22, 23] . briefly, nucleic acid was extracted, and cdna was synthesised. a hemi-nested pcr protocol was used targeting the 5 ′ untranslated region and the second hypervariable region of the attachment (g) glycoprotein coding region for hrv and rsv, respectively [22, 23] . sequences were aligned using mafft v7.309, and maximum likelihood phylogenetic trees using the hky85 model with 100 bootstrap replicates were inferred using phyml 3.1 within geneious [24, 25] . sequences were considered to be the same virus type with ⩾98% identity. sequences >200 base pairs (bp) were submitted to genbank under accession numbers mh266546 to mh266612. for each examination of viral transmission, initial viral infections were those preceded by a 14-day period without infections of that virus type in a household. index or transmitting cases were established by identifying the individual(s) with the earliest reported respiratory symptoms prior to a virus-positive swab. we computed the incidence of secondary household illness cases following each initial viral infection. we defined transmission events as observed infections of the same virus in another member of the same household within the subsequent 14 days. the 14-day risk period began on the first day of criteria respiratory symptoms associated with virus-positive specimen collection in the week preceding the virus-positive swab in the index case. each initial infection and its corresponding risk period comprise a single data point in the regressions and include subsequent infections and time at risk from all household members reporting symptoms over that time period. to assess the risk factors for the incidence of secondary cases of these initial illnesses within households, we used generalised estimating equations, accounting for the potential similarity among repeated initial infections within households. the outcome was the number of secondary cases, with a fixed offset of the log time at risk; a log link was used to estimate the incidence rate ratios (irrs). multivariable regression was performed by first including all measures significant in the univariable analysis at p < 0.1 and then performing backward elimination to select a final model. measures were retained in final multivariable regression if significant at p < 0.05 or if had a substantial impact (⩾10% shift in estimate) on other significant covariates. a single index case type was included in the multivariable model as the index case was coded such that one index case type was compared to all other index cases. potential risk factors for transmission included maternal and household characteristics. some households included more than one enrolled mother-infant pair. among households with more than two mothers, household characteristics were compared with a sensitivity analysis using one mother's descriptors vs. the others. data were analysed using sas/stat 9.4 (sas institute inc.) and stata 15 (stata corp) statistical software. institutional review board approval for the randomised controlled trial was given by the johns hopkins university bloomberg a total of 752 households were enrolled with a median household size of 9 (range 2-31). five-hundred and fifty-five households contributed symptom reporting from at least three persons. within the 555 surveyed households, 3232 out of 5521 (59%) initially enrolled household members were surveyed for weekly respiratory illness. these 3232 individuals included in the transmission analysis consisted of 683 mothers, 665 infants, 1127 other adults ⩾15 years and 757 other children <15 years ( fig. 1) . characteristics of all households and individual characteristics of surveyed individuals are summarised in table 1 . within included households, 99% of study mothers and infants were surveyed, whereas only 39% of other adult household members and 58% of other children were surveyed. the proportion of other adults with surveillance was 32% vs. 53% among individuals <40 vs. ⩾40 years, and 37% vs. 43% in males compared to females. forty-nine per cent of other children aged 5-14 years were surveyed compared to 73% of other children aged <5 years. of children aged 5-14 years attending school, 65% were surveyed compared to 49% of non-school attending children. a total of 825 virus-positive initial illness episodes occurred within 362 households with a median of one (range 0-10) illness episode per household. in the 14 days following initial household illness, 110 subsequent illness episodes occurred, 88 (80%) of which were screened by pcr and 22 (20%) illnesses that did not have a swab collected despite meeting symptom criteria. eight per cent of illness episodes resulted in a pcr-confirmed secondary case within the household with a total of 79 transmission events in 68 household illness episodes. household illness clusters occurred in 58 households as some households experienced multiple illness clusters. the incidence of a pcr-confirmed transmission event of any virus occurring in the 14 days following initial fig. 3 ). hrv coinfection with another respiratory virus had more frequent transmissions (16.1% of hrv coinfections resulted in transmission) compared to monoinfection of hrv (5.8%), coronavirus (3.5%) and rsv (6.7%) (fig. 2 , see table 3 for statistical testing). we evaluated risk factors for transmission following initial respiratory viral illness ( fig. s2 ). of nine fully evaluated hrv transmission events, household sequences matched in six (18.1% of all hrv transmission events). in three events (9.1% of all transmission events), the individuals were infected with different hrv genotypes (fig. 4 ). all other episodes had insufficient data to confirm the transmission of specific viral genotypes. of 362 households contributing to the regression analysis of any virus transmission, 52 households included more than one mother-infant pair. within households with multiple mothers, there was 100% agreement in reporting of electricity within the home, 89% agreement in indoor cookstove use and 85% agreement of latrine within the home. the multivariable regression model was performed using the alternative mother's household information with similar results (supplementary table s1 ). a sensitivity analysis was also performed using a definition of transmission as a secondary case of the same virus within 28 days of initial infection within the household. using this definition, preschool child index case, coinfection as initial infection and a low birthweight infant in the household were associated with an increased incidence of household transmission (supplementary tables s2 and s3 ). in a prospective longitudinal study utilizing intensive weekly home-based active surveillance to evaluate the household transmission of nine respiratory viruses in rural south asia, initial infection in young children was associated with the greatest risk of symptomatic respiratory virus household transmission with spread to infants occurring in 45% of transmission events. southern nepal is a region where household crowding is common and there are high rates of infants born prematurely or low birthweight [26, 27] . our data demonstrate a significant burden of symptomatic respiratory viral illness in households; based on a multivariable model, young children and socio-cultural factors, such as socio-economic status, may predispose to the transmission of viruses in this region. in over 40% of transmission events of all viruses, preschool children (aged 1-4 years) served as an index case. a higher proportion of initial infection among this group resulted in secondary cases compared to other age groups, including school-age children and mothers, a finding confirmed in our multivariable model of transmission incidence. in rsv transmission, no index cases were older children and 15% of index cases were mothers. in contrast to our findings, a study of rsv transmission in the usa during the 1960s found that older siblings between 2 and 16 years most frequently introduced rsv into a household [11] . similarly, a kenyan household study examined sequencingconfirmed rsv transmission in 44 households with infants and identified school-age children as the most common index case resulting in infant infection [6] . our finding that preschool-age children, rather than schoolage children, are most likely to transmit non-influenza respiratory viruses is likely due to differences in study sample and design, as well as transmission patterns. households in our study experienced fewer respiratory viral illness episodes than reported in other household studies that included asymptomatic viral detections [5, 7, 15] . in a study of respiratory virus-positive influenzalike illness in households in vietnam, households experienced 1.6 illness episodes over a 1-year period (including influenza and bocavirus), whereas we found a mean of 1.4 illness episodes per household [8] . our surveillance sample includes a higher proportion of young children aged 0-4 years, including study infants, relative to the overall proportion in the sarlahi district of nepal (32.4% vs. 11.3%) and a lower proportion of children aged 5-14 years (11.6% vs. 27.9%) [18] . it is possible that the true transmitting cases were absent during the weekly household visit or asymptomatic according to our criteria, though this is less likely for younger children who have median viral shedding duration of longer than 1 week [31] . we likely did not capture the full a total for initial episodes, transmission episodes and transmission events are less than the sum of columns as infections with coinfections were counted a single initial infection in total. similarly, the total for index cases may be less than the sum of columns as coinfection transmission or transmission to multiple household members was only counted once in total. b median serial index was defined as the median number of days between symptom onset of index and secondary case. contribution of older children to transmission compared to the kenyan cohort. over half of rsv infections in children 5-15 years in that study were asymptomatic, with a smaller proportion of asymptomatic infections in infants under 1 year and children 1-4 years at 9% and 17%, respectively [28] . however, they also reported viral shedding in symptomatic rsv infections was 14 log 10 rna copies greater than in asymptomatic rsv cases suggesting that symptomatic episodes are more likely to transmit virus [29] . last, we collected weekly specimens and our findings may be biased if non-infant younger children had longer shedding duration compared to older children. however, this has not been demonstrated in studies of rsv and hrv shedding duration [28, 30] . our large cohort allowed us to use a multivariable analysis to identify the risk factors and protective characteristics associated with the incidence of transmission. while both infants and preschool children were frequently identified as the index case in a transmission event and can shed virus for prolonged periods, a preschool child index case was associated with a twofold increased risk of transmission and an infant index case was associated with a decreased risk of transmission [11, 31, 32] . whereas infants are more likely to transmit rsv via direct contact as compared with fomites, young children may transmit infection efficiently through both methods due to differences in mobility and behaviour [33] . coinfection as the initial infection was associated with an increased risk of transmission, including in our multivariable model. coinfections most commonly involved hrv and a greater proportion of coinfections resulted in a secondary case compared to monoinfection of most viruses. viral coinfection with rsv infection has been demonstrated to increase rsv viral load and shedding duration. however, this has not been consistently seen, including a study analysing seven respiratory viruses [29 31, 34] . finally, electricity in the household, a proxy for socio-economic status and housing conditions, was negatively associated with transmission. although an association between indoor air pollution and rsv infection has been reported in resource-limited regions, smoking and biofuel cookstove use were not associated with the risk for transmission in our model [4] . however, we had limited power to detect this association due to the use of indoor biofuel cookstoves in over 90% of households in our model and exposures were self-reported without actual measures of indoor air pollution. a study in peru demonstrated that age, occupation and household size can influence contact network size and pattern [35] . our findings, from a population consisting of crowded households, lower levels of maternal education and fewer children attending school compared to other household transmission studies, suggest that differences in socio-demographic, cultural and environmental contexts influence household transmission risk factors, including the source of household introduction. as we actively surveyed all women of childbearing age for pregnancy, our cohort is generalisable to households with young infants in southern nepal, a region representative of rural south asia [17] . in the sarlahi district during the study period, an estimated 25% of the population were below the poverty line and approximately 30% of infants were born low birthweight and 20% preterm [26, 27] . households in our study were crowded with over one-third containing >4 people per room and multiple family units frequently living in a single structure. the average population is young; the median age in the sarlahi district was 20 years [18] . eighty-four per cent of households used indoor biofuel cookstoves and half contained latrines. young demographics, crowded housing conditions and socio-economic factors may influence the patterns of respiratory (a and b) . each row represents an individual, each unfilled symbol represents 1 day of symptoms, black filled symbols represent positive specimen collection and varying symbols represent household member type. index cases are those whose symptoms first appear before the initial rsv-positive specimen. [36] . this socio-demographic, environmental and cultural context should be considered when implementing preventative strategies for the control of respiratory viral illness, such as vaccines, antivirals, hygienic measures and physical barriers. for example, there are multiple rsv vaccines targeting diverse populations from infants and children to pregnant women and other adults in various stages of clinical trials [10] . because the immune systems of neonates generally do not respond well to primary vaccination, immunizing mothers and other household members has been proposed as a method to protect vulnerable young infants from rsv [37] . while a model of kenya transmission data supports immunizing school-age children to diminish transmission of the virus to infants, our study suggests that in rural south asia, preschool-age children are more likely to transmit respiratory viruses to other household members [38] . this suggests that a 'one-size fits all' approach to rsv vaccine implementation, or other respiratory viral transmission prevention measures, may not be effective as transmission dynamics may differ across global settings. our study has several limitations. asymptomatic infections were not captured, affecting our ability to fully characterise the transmission chain. we expect that asymptomatic transmission may have impacted our ability to characterise hrv spread, particularly in transmission involving older children and adults, our ability to associate age of index case with transmission risk [30, 39] . moreover, we likely only captured a minority of adult illness as adults required subjective fever for specimen collection and fever occurs infrequently in adult rsv, mpv and hrv illness [7 40 ]. while we underestimated transmission, specifically spread involving individuals ⩾5 years, due to our symptom criteria, a previous study of rsv transmission demonstrated that the odds of transmission in symptomatic infection is five times that of asymptomatic infections [28] . this suggests that we likely captured index infections. we anticipate that some illness with shedding <7 days may have been missed due to our weekly surveillance. we expect that this represented a small minority of illness episodes as the estimated shedding duration of hrv and rsv in adults is 10 and 9 days, respectively [30, 31] . shedding for 1-2 weeks is common with paediatric respiratory infections [31] . moreover, while we performed sequencing on rsv and hrv samples involved in transmission chains, we were not able to phylogenetically verify transmission in the majority of episodes due to high cycle threshold values, and could not use sequencing data to define transmission. our sequencing results revealed a degree of misclassification with some hrv and rsv transmission events representing illness clusters with multiple virus types circulating in the household simultaneously. this was especially true for hrv, a finding in agreement with previous studies of hrv transmission, including in household and daycare settings [30, 39 41] . additionally, some households originally selected for the household substudy were not surveyed as intended. individuals within selected households who were not surveyed were primarily adults. among adults, males and those <40 years old were surveyed less frequently, a group not considered high risk for household transmission of respiratory viruses in previous studies. a significant proportion of the sarlahi population, especially men, are reported as absent from home, supporting the possibility that some household members were absent from the community during the study, although these data were not captured [36] . we also surveyed a higher proportion of preschool children compared to school-aged children. the differential exclusion of these subsets may have affected our results, including identification of index cases, especially if these persons were periodically present in the household. lastly, we did not collect data on the social mixing patterns of individuals in these households which would have provided valuable information regarding possible causal explanations for our findings. these results provide data that may help to optimise the implementation of preventative strategies with the aim of protecting vulnerable infants. south asia is an area of the world with a high incidence of low birthweight infants, household crowding and malnutrition, all risk factors for severe childhood alri. our study of non-influenza respiratory virus transmission within households in rural nepal highlights the importance of targeting preschool-age children to prevent the spread of respiratory viral illness. understanding the household transmission of respiratory viruses in rural resource-limited populations will help evaluate infection prevention strategies, such as immunisation of mothers and other household members in protecting infants, who are most vulnerable to respiratory viral infection. global and regional burden of hospital admissions for severe acute lower respiratory infections in young children in 2010: a systematic analysis viral pneumonia a prospective three-year cohort study of the epidemiology and virology of acute respiratory infections of children in rural india incidence and risk factors for respiratory syncytial virus and human metapneumovirus infections among children in the remote highlands of peru continuous invasion by respiratory viruses observed in rural households during a respiratory syncytial virus seasonal outbreak in coastal kenya the source of respiratory syncytial virus infection in infants: a household cohort study in rural kenya community surveillance of respiratory viruses among families in the utah better identification of germs-longitudinal viral epidemiology (big-love) study epidemiology and etiology of influenza-like-illness in households in vietnam; it's not all about the kids! the burden of respiratory syncytial virus infection in young children strategic priorities for respiratory syncytial virus (rsv) vaccine development respiratory syncytial virus infections within families vaccination strategies against respiratory syncytial virus factors associated with transmission of influenzalike illness in a cohort of households containing multiple children household transmission of respiratory virusesassessment of viral, individual and household characteristics in a population study of healthy australian adults rhinovirus transmission within families with children: incidence of symptomatic and asymptomatic infections designs of two randomized, community-based trials to assess the impact of influenza immunization during pregnancy on respiratory illness among pregnant women and their infants and reproductive outcomes in rural nepal year-round influenza immunisation during pregnancy in nepal: a phase 4, randomised, placebo-controlled trial national population and housing census evaluation of quantitative and type-specific real-time rt-pcr assays for detection of respiratory syncytial virus in respiratory specimens from children real-time reverse transcription-pcr assay for comprehensive detection of human rhinoviruses comparison of real-time pcr assays with fluorescent-antibody assays for diagnosis of respiratory virus infections in children molecular epidemiology of human rhinovirus infections in the pediatric emergency department molecular epidemiology of respiratory syncytial virus transmission in childcare new algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of phyml 3.0 recent developments in the mafft multiple sequence alignment program mortality risk in preterm and small-for-gestational-age infants in low-income and middle-income countries: a pooled country analysis nepal human development report frequent asymptomatic respiratory syncytial virus infections during an epidemic in a rural kenyan household cohort quantification and determinants of the amount of respiratory syncytial virus (rsv) shed using real time pcr data from a longitudinal household study virus shedding after human rhinovirus infection in children, adults and patients with hypogammaglobulinaemia influence of age, severity of infection, and co-infection on the duration of respiratory syncytial virus (rsv) shedding epidemiology of multiple respiratory viruses in childcare attendees modes of transmission of respiratory syncytial virus multiple versus single virus respiratory infections: viral load and clinical disease severity in hospitalized children a household-based study of contact networks relevant for the spread of infectious diseases in the highlands of peru nepal living standards survey 2011/11: statistical report maternal immunization evaluating vaccination strategies for reducing infant respiratory syncytial virus infection in low-income settings an intensive, active surveillance reveals continuous invasion and high diversity of rhinovirus in households the presence of fever in adults with influenza and other viral respiratory infections heterotypic infection and spread of rhinovirus a, b, and c among child care attendees acknowledgements. we thankfully acknowledge the original principal investigator of the main study, the late mark c. steinhoff. we acknowledge the contribution of the nepal nutrition intervention project study team in sarlahi and kathmandu. supplementary material. the supplementary material for this article can be found at https://doi.org/10.1017/s0950268819001754. key: cord-349298-8s69wprh authors: munywoki, p. k.; koech, d. c.; agoti, c. n.; kibirige, n.; kipkoech, j.; cane, p. a.; medley, g. f.; nokes, d. j. title: influence of age, severity of infection, and co-infection on the duration of respiratory syncytial virus (rsv) shedding date: 2014-06-05 journal: epidemiol infect doi: 10.1017/s0950268814001393 sha: doc_id: 349298 cord_uid: 8s69wprh rsv is the most important viral cause of pneumonia and bronchiolitis in children worldwide and has been associated with significant disease burden. with the renewed interest in rsv vaccines, we provide realistic estimates on duration, and influencing factors on rsv shedding which are required to better understand the impact of vaccination on the virus transmission dynamics. the data arise from a prospective study of 47 households (493 individuals) in rural kenya, followed through a 6-month period of an rsv seasonal outbreak. deep nasopharyngeal swabs were collected twice each week from all household members, irrespective of symptoms, and tested for rsv by multiplex pcr. the rsv g gene was sequenced. a total of 205 rsv infection episodes were detected in 179 individuals from 40 different households. the infection data were interval censored and assuming a random event time between observations, the average duration of virus shedding was 11·2 (95% confidence interval 10·1–12·3) days. the shedding durations were longer than previous estimates (3·9–7·4 days) based on immunofluorescence antigen detection or viral culture, and were shown to be strongly associated with age, severity of infection, and revealed potential interaction with other respiratory viruses. these findings are key to our understanding of the spread of this important virus and are relevant in the design of control programmes. respiratory syncytial virus (rsv) is a major viral cause of lower respiratory tract infection in children worldwide [1] with the key risk group being young infants [2] . no vaccine is currently available for this age group. development of alternative control strategies depends on the mechanisms of transmission, which are intrinsically related to viral shedding [3, 4] . detailed data on shedding in individuals in relation to age, infecting subtype (groups a or b), infection severity, and gender would help in identifying the source of infant infection. such data from the natural setting unaffected by sampling bias are limited and absent in resource-poor settings. additionally, this study assesses the impact of the presence of other respiratory viruses, prior to or concomitant with rsv, on rsv infection duration. any interaction might be mediated through direct interference or host immune and physiological responses. these possibilities have received very little attention in the literature. previous studies on rsv shedding have been mainly in the hospital setting limiting the generalizability of the results [3, 5, 6] . hospital studies are biased to young children with severe rsv disease and fail to precisely establish the start and, often, the end of shedding, particularly when symptoms do not coincide with virus shedding which is common for rsv [6] . community-based studies are likely to provide a more complete representation of the rsv shedding patterns. such studies require frequent nasopharyngeal swabbing regardless of symptoms and use of sensitive molecular techniques for viral testing in order to minimize the likelihood of missing infection episodes especially in older age groups. the current prospective study utilizes the above approach, with intensive sampling (every 3-4 days), molecular testing, and follow-up of individuals of all ages for one complete rsv season in a rural kenyan community [7] . this provides for more realistic estimates on duration of and influencing factors on rsv shedding which are required in designing rsv prevention strategies and to better understand the impact of vaccination on rsv transmission dynamics. the study was undertaken in rural coastal kenya within the kilifi health and demographic surveillance system (khdss) [8] . the study methods have been described elsewhere [7] . briefly, a prospective cohort study was undertaken with a recruitment target of 50 rsv-naive infants and their household members. a household was defined as a group of individuals living in the same compound and with common cooking arrangements. households were eligible if they contained a child born after the end of the 2008-2009 rsv epidemic, and one or more older siblings (aged <13 years). sampling visits were timed to begin and end coincident with the start and finish of the expected rsv season of 2009-2010 [7] . trained field assistants made twice weekly household visits, collecting deep nasopharyngeal swabs (nps) irrespective of symptoms and recording clinical illness data from all participants. nps were collected and tested for rsv (groups a and b) and other respiratory viruses [adenoviruses, rhinoviruses and human coronaviruses (nl63, 229e, oc43)] using real-time multiplex pcr as described previously [7] . in order to establish the genetic similarity of the rsv strains in suspected repeat infections, the ectodomains of the rsv attachment (g) protein gene were sequenced and analysed phylogenetically [9] . the authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the helsinki declaration of 1975, as revised in 2008. data were analysed using stata version 11.2 (statacorp, usa). the infection data were intervalcensored and three possible durations of viral shedding were estimated, i.e. minimum, midpoint and maximum, as described in the supplementary online material and shown in figure s1 . an rsv infection episode was defined as the period within which an individual provided specimens which were pcr positive for the same infecting rsv group with no more than 14 days separating any two positive samples. episodes where the first sample was positive for both rsv groups a and b counted as one infection episode. episodes where no samples were collected for >7 days before or after the infection episode were considered left-or right-censored, respectively. symptomatic infection was defined as the presence of one or more of the following symptoms: cough, nasal discharge/blocked nose, or difficulty in breathing at any time during the infection episode. co-infection was assigned when within the rsv episode any sample was pcr positive for another virus, i.e. coronavirus, rhinovirus, or adenovirus. presence of these viruses in the samples collected in the period of 14 days prior to start of rsv episodes was also defined. household outbreak was defined as a period within which more than one individual episodes occurred in members of the same household without an interval of 514 days in which a pcrpositive specimen was absent from the household. the proportion of the household members infected during household outbreaks measured the intensity of the outbreak. cox proportional hazards models were used to identify factors influencing the rate of loss of virus detection (hereafter referred as the recovery rate). the effect of left-censoring was accounted for in the multivariate model by including a dummy variable or by excluding the left-censored episodes. of the 493 individuals in the 47 households followed, 179 (36·3%) had at least one rsv infection from 40 (85·1%) different households. the median age (interquartile range; iqr) at the start of the first observed infection was 6·5 (iqr 2·4-14·5) years, and females numbered 96 (53·6%) ( table 1) . a total of 205 infection episodes were observed with 155 individuals experiencing one episode, 22 with two episodes and two individuals experiencing three episodes (fig. 1 ). rsv group a was associated with 88 infection episodes, rsv group b with 113 while seven episodes were co-infections. there were 177 (86·3%) fully observed episodes while 11 and 15 infection episodes were left-and right-censored, respectively, and two episodes were both left-and right-censored. of the 24 individuals with two or more episodes (suspected repeat infections), 17 (70·8%) were infected with the same rsv group and otherwise all except one group a infection followed group b. the mean age at the first infection for individuals with rsv group a and group b was 2·3 and 7·2 years, respectively. the duration between the episodes ranged from 17 to 54 days with median of 28 days. for the 17 homologous infections, sequencing of the rsv g gene was successful in 10 (59%) of the 18 possible pairs of samples (one individual had three suspected rsv episodes). the failure to sequence was mainly in samples with a pcr cycle threshold (c t ) value of > 28·0, an indicator of low viral load. only one of the successfully sequenced paired samples showed nucleotide differences: 13 nucleotide differences associated with three non-synonymous changes and a change in the stop codon position. the episodes in this individual (id no. 1803 in fig. 2 ) occurred 54 days apart. for the purposes of estimation of the shedding duration, all the episodes were considered distinct. from the 205 infection episodes, the mean duration of shedding based on minimum, midpoint and maximum estimates were 8·6 [95% confidence interval (ci) 7·5-9·7], 11·2 (95% ci 10·1-12·3) and 14·0 (95% ci 12·8-15·2) days, respectively, for all rsv episodes (fig. 3) . the corresponding mean durations of shedding for the 'fully observed' episodes were 8·2 (95% ci 7·1-9·4), 10·9 (95% ci 9·8-12·1) and 13·6 (95% ci 12·4-14·8) days, respectively ( table 2) . twenty-four individuals shed rsv for 521 days, and of these 10 (41·7%) were aged <1 year, six (25·0%) were aged 1-4 years, and eight (29·2%) aged 5-17 years. twenty-two (91·7%) of these infection episodes were symptomatic throughout or at some time point during the shedding. the prolonged shedders contributed 647·5 days of rsv shedding which was 29·7% of the cumulative shedding duration for all episodes based on the midpoint estimation. in 14 infected individuals, one or more samples were identified to contain both rsv groups a and b. the timing of these co-detections is shown in figure 4 . in most (12/14) episodes, rsv group a was shed for longer duration relative to group b. the hazard ratios (hrs) for the various factors from univariate cox regression were similar for minimum, midpoint and maximum estimates data (supplementary table s1 ). the midpoint data were taken forward for the multivariate analysis and the final model is reported in table 3 . the results were similar without and with inclusion of the left-censored rsv episodes (table 3 and supplementary table s2 , respectively). the proportionality assumption in the cox regression model was not violated based on the test of the schoenfeld residuals (supplementary table s3 ). the rate of recovery from rsv infection was age-dependent. the adjusted hr (ahr) were 1·98 (95% 1·30-3·02), 1·82 (95% ci 1·16-2·87), 2·10 (95% ci 1·20-3·66) and 1·31 (95% ci 0·36-4·81) in the 1-4, 5-14, 15-39 and 540 years age groups, respectively, relative to infants (<1 year). the rate of recovery was lower by 44% in symptomatic infections relative to asymptomatic infections (ahr 0·56, 95% ci 0·40-0·79). the presence of one or more additional viruses (rhinovirus, coronavirus, adenovirus) was detected in 86 rsv infection episodes. the rate of rsv recovery was lower (i.e. shedding duration increased) by 65% in episodes with co-infection compared to those without (ahr 0·35, 95% ci 0·23-0·51), with a similar result for each virus individually. detection of infection with any one or more of rhinovirus, adenovirus or coronavirus, in the 2 weeks preceding the start of rsv infection, but not during the rsv episode itself, was associated with a 56% increase in the rate of recovery (i.e. reduced shedding duration) from the rsv infection (ahr 1·56, 95% ci 1·02-2·39). in contrast, rsv episodes associated with detection of other viruses in the 2 weeks prior to and also during the rsv infection were associated with a 52% decrease in the rate of recovery relative to those with no other virus prior to and during the rsv episode (ahr 0·48, 95% ci 0·32-0·73). the rate of recovery of rsv episodes associated with spread in the household (outbreaks) was 42% lower than the single household episodes (ahr 0·58, 95% ci 0·43-0·78). a variable denoting the proportion of individuals infected during the household outbreak improved the model fit and was used in the multivariate analyses (likelihood ratio test p = 0·0229). the rate of recovery did not differ significantly by gender and infecting rsv group (ahr 0·97, 95% ci 0·77-1·23 and ahr 1·07, 95% ci 0·83-1·39, respectively). recovery rate was similar in suspected repeat infections compared to the first observed episodes (ahr 0·91, 95% ci 0·53-1·56). we observed 205 infections with rsv during one epidemic with a most realistic estimate of 11·2 (95% ci 10·1-12·3) days of shedding. the most conservative and least conservative estimates were 8·6 days and 14·0 days, respectively. the duration of shedding based on the most realistic estimate decreased with age: 18 days in infants and 9 days in adults (aged 515 years). symptomatic infections on average had longer virus shedding of 13·5 days compared to 7·8 days in asymptomatic episodes. the presented average durations of virus shedding are longer than published estimates of 6·7 days [4] , 3·4-7·4 days [10] , 3·9 days [11] , and 4·5 days [12] which could be attributed to differences in study methods. the present study was informed by critical review of the previous studies and it incorporated frequent sampling regardless of symptoms and screening by highly sensitive pcr methods. the specimen collection procedure was acceptable, recording a good compliance across all ages [7, 13] . the use of the sensitive viral detection method (pcr) results is likely to result in longer estimates of shedding. a community study nested within a birth cohort in coastal kenya targeting symptomatic rsv infections by okiro and colleagues reported a mean duration of shedding of 4·5 days [12] . our corresponding estimate in symptomatic cases was 13·5 days. in a subset of the children whose start of symptoms could be established from the clinic records, the okiro et al. study reported a longer duration of 7·7 days [12] . given that rsv shedding has been reported to start before illness [6] , the actual duration in the symptomatic children would have been an underestimate and our estimate of 13·5 days is likely to be more accurate. a rochester family study, in the usa, with similar design as the present study (collecting samples every 3-4 days regardless of symptoms) reported lower estimates of duration of shedding of 3·4-7·4 days [10] . the okiro et al. and rochester study used the immunofluorescent antibody test (ifat) and culture, respectively, which are less sensitive methods [14] . as a counter argument, it is not known to what degree pcr positivity equates with shedding of viable and infectious virus. thus, while the molecular methods might be more sensitive, the resultant increase in duration of shedding over more traditional methods such as culture (which directly measures viral infectivity) may not necessarily translate to increased period of infectivity. further work relating virus infectiousness and detectability is warranted. in the okiro et al. study, sampling started when participants were symptomatic and stopped at the first negative follow-up sample. the present study revealed instances where negative samples arose within rsv infection episodes. even though this observation raises, again, questions on the relationship between infectivity and shedding duration, accounting for periods of rsv-negative samples would still result to longer shedding duration compared to previous estimates. alternative estimation of the shedding patterns by calculating the area under the c t (viral load) curve would have some additional advantages and will be explored in future. prolonged shedders of > 3 weeks' duration have been reported [6, 10] . in the present study, 24 (13·4%) episodes in the 179 infected individuals involved shedding rsv for >3 weeks. most (22, 91·7%) of these episodes were symptomatic, and occurred in young children (median age 14·7 months). individuals with compromised immunity have been known to shed rsv for longer [15] but participants in the current study were not tested for hiv. the hiv prevalence in women and men aged 15-49 years in coastal kenya was 4·2% according to the kenyan demographic and health survey of 2008/2009 [16] . in settings where hiv prevalence is high, the effect of the poor viral clearance might influence the temporal epidemiology as was observed in south africa [15] . there was no obvious malnourished participant in the study cohort based on midupper arm circumference measurements. more than two rsv episodes were observed in 24 (13·4%) individuals. on average, the episodes were 4 weeks apart. most (70·8%) of the suspected repeat infections were with homologous rsv group. sequencing of the most variable region of the rsv genome, the ectodomain of the g gene, did not greatly assist in resolving the infection episodes as most (9/10) had identical sequences. it is, thus, not clear whether the two phases of rsv shedding were repeat infections with the same variant or were persistent infection with periods of low viral load that was undetectable by the methods used. using post-mortem lung tissue from infants, rsv rna has been detected even in children dying during inter-epidemic periods suggesting the persistence of rsv in the lungs of these infants [17] . similar observations have been made in experimental infection with rsv [18] and measles viruses [19] . the lack of variability in the virus identified in the two phases of infections suggests virus mutation might not be the primary mechanism for virus persistence or re-infection. regardless of whether it is re-infection within a short period or persistence, the observation represents an interesting phenomenon of rsv which has potential importance on our existing view of acute rsv infection, the development of immunity and effects on viral transmission. the age of the individual, infection severity, detection of other viruses before and during the rsv infection, and presence of concurrent rsv infections in the same household were all associated with virus shedding. the rate of recovery increased with increasing age, with individuals in 1-4 years, 5-15 years and 515 years age groups recovering 1·98, 1·82 and 1·97 times faster than the infants, respectively. the rochester family study reported similar findings where longer shedding was observed for children aged <2 years compared to those aged 2-16 years 14·8 -ci, confidence interval; rsv, respiratory syncytial virus. a all rsv infection episodes; b fully observed episodes; c left-censored episodes only; d right-censored episodes only; e episodes with both left and right censoring; f intervals between nasopharyngeal swab collection before, during, and after the rsv infection episode in days; g age (in years) at the start of the episode; h shedding duration (in days) based on the midpoint estimation. (9 vs. 4 days). the okiro et al. study did not find any association with age but found that children with previous rsv infection (using the assumption that those aged >3 years were by default experiencing a repeat infection) had 1·37 times faster rate of recovery compared to those without a history of infection [12] . in the present study, a subsequent rsv infection during the same rsv season was not significantly associated with reduced shedding duration, but such infections were few (n = 24). the current study is for one epidemic only, so age must act as a proxy of exposure to rsv in earlier epidemics. prolonged shedding enhances the possibility of personto-person transmission and makes young children a potential source of community spread of infection, both of which have important implications in the control and prevention of rsv infection. a study involving 23 hospitalized children (aged <2 years) with sampling extended beyond discharge reported an association of duration of shedding and symptom severity [4] . children with lower respiratory tract infection shed for longer than those with upper respiratory tract infection (8·4 vs. 1·4 days). duration of shedding may be related to severity of disease but evidence is controversial on the link between disease severity and viral load [5, [20] [21] [22] . an interaction between detection of other viruses (i.e. coronavirus, rhinovirus, adenovirus) in the nasopharynx and rate of recovery from rsv infection was observed. recovery from a viral infection just prior to rsv might have led to up-regulation of innate viral immunity or non-specific cross-reactivity that reduced subsequent rsv shedding. presence of co-infections might be a marker of low immunity associated with poor viral clearance. rsv episodes with concurrent spread in the household were associated with increased recovery rate. however, any conclusions other than association cannot be made since the extended duration of shedding increases the risk of both co-infection with other viruses and multiple rsv infections in the household. a different estimation framework is required to untangle these results. the problems of ascertainment and analysis (i.e. censoring and test sensitivity) are not completely eliminated by the careful study design, but did not seem to affect the association of the examined factors with virus shedding (supplementary table s1 ). short rsv infections occurring between specimen collections particularly in older individuals might have been missed but this is likely to be at random and bias hazard ratios towards null. in conclusion, this study defines rsv shedding patterns in the natural setting with significant potential for improved understanding of the spread of this important virus and with relevance the design of control programmes. for supplementary material accompanying this paper visit http://dx.doi.org/10.1017/s0950268814001393. global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic review and meta-analysis risk of primary infection and reinfection with respiratory syncytial virus shedding of infectious virus and virus antigen during acute infection with respiratory syncytial virus respiratory syncytial virus infections in infants: quantitation and duration of shedding quantitative shedding patterns of respiratory syncytial virus in infants patterns of shedding of myxoviruses and paramyxoviruses in children the source of respiratory syncytial virus infection in infants: a household cohort study in rural kenya profile: the kilifi health and demographic surveillance system (khdss) genetic relatedness of infecting and reinfecting respiratory syncytial virus strains identified in a birth cohort from rural kenya respiratory syncytial virus infections within families respiratory syncytial virus infections in previously healthy working adults duration of shedding of respiratory syncytial virus in a community study of kenyan children improved detection of respiratory viruses in pediatric outpatients with acute respiratory illness by real-time pcr using nasopharyngeal flocked swabs comparison of direct immunofluorescence, conventional cell culture and polymerase chain reaction techniques for detecting respiratory syncytial virus in nasopharyngeal aspirates from infants increased burden of respiratory viral associated severe lower respiratory tract infections in children infected with human immunodeficiency virus type-1 kenya demograpghic and health survey 2008-09 detection of respiratory syncytial virus nucleic acid in archival postmortem tissue from infants latency and persistence of respiratory syncytial virus despite t cell immunity prolonged persistence of measles virus rna is characteristic of primary infection dynamics respiratory syncytial virus load predicts disease severity in previously healthy infants evaluation of quantitative and type-specific real-time rt-pcr assays for detection of respiratory syncytial virus in respiratory specimens from children natural infection of infants with respiratory syncytial virus subgroups a and b: a study of frequency, disease severity, and viral load we thank the field and laboratory teams for their dedication in sample collection and testing, respectively. the article is published with the permission of the director of the kenya medical research institute. this work was supported by the wellcome trust (090853 and 084633). none. key: cord-345020-ai5tib7h authors: price, o. h.; sullivan, s. g.; sutterby, c.; druce, j.; carville, k. s. title: using routine testing data to understand circulation patterns of influenza a, respiratory syncytial virus and other respiratory viruses in victoria, australia date: 2019-06-17 journal: epidemiol infect doi: 10.1017/s0950268819001055 sha: doc_id: 345020 cord_uid: ai5tib7h several studies have reported evidence of interference between respiratory viruses: respiratory viruses rarely reach their epidemic peak concurrently and there appears to be a negative association between infection with one respiratory virus and co-infection with another. we used results spanning 16 years (2002–2017) of a routine diagnostic multiplex panel that tests for nine respiratory viruses to further investigate these interactions in victoria, australia. time series analyses were used to plot the proportion positive for each virus. the seasonality of all viruses included was compared with respiratory syncytial virus (rsv) and influenza a virus using cross-correlations. logistic regression was used to explore the likelihood of co-infection with one virus given infection with another. seasonal peaks were observed each year for influenza a and rsv and less frequently for influenza b, coronavirus and parainfluenza virus. rsv circulated an average of 6 weeks before influenza a. co-infection with another respiratory virus was less common with picornavirus, rsv or influenza a infection. our findings provide further evidence of a temporal relationship in the circulation of respiratory viruses. a greater understanding of the interaction between respiratory viruses may enable better prediction of the timing and magnitude of respiratory virus epidemics. influenza, respiratory syncytial virus (rsv) and other respiratory viruses are the cause of substantial morbidity and mortality, with children under 5 years of age and the elderly disproportionately burdened [1] . both influenza and rsv display distinct seasonality, however, the exact timing and magnitude of their annual epidemics remain difficult to predict [2] . a better understanding of the epidemiology of these pathogens is useful for the prevention and control of future epidemics and for optimising clinical management of patients [3] . moreover, this knowledge may inform prediction models used to estimate the timing and magnitude of influenza epidemics [2] . interference between respiratory viruses has been well documented. during peaks of influenza epidemics, the spread of other respiratory viruses, particularly rsv, appears to be limited [4] [5] [6] . delays in outbreaks of influenza during the 2009 pandemic in europe were linked to the annual rhinovirus epidemic associated with the beginning of the school year [7] [8] [9] . in turn, the influenza pandemic was observed to interfere with seasonal epidemics of rsv in france [10] and israel [11] , rsv and metapneumovirus in germany [12] , seasonal influenza in hong kong [13] and all respiratory viruses except rhinovirus in beijing [14] . studies investigating viral interference since the pandemic are sparser, though two studies reported that the timing and magnitude of respiratory virus epidemics were affected by the timing of the seasonal influenza a peak [15, 16] . collectively, these observations suggest interference may prevent respiratory viruses reaching their epidemic peaks concurrently, but also underscore the complexity of these interactions. the exact nature of interactions between different respiratory viruses remains unclear, although they are proposed to be driven by the innate immune system. once a viral infection is established, interferon production is believed to confer temporary immunity to neighbouring cells against infection by other respiratory viruses [17] . in vitro, infection with rsv is blocked by competitive infection of influenza a if the host is not infected with the two viruses simultaneously [18] . similarly, ferret models have shown that influenza a infection may prevent successive infection with rsv [19] and that coinfection with different influenza subtypes is dependent upon the order in which the viruses infect the host [20] . despite this apparent interference, viral co-infections do occur, albeit with insufficient frequency to maintain an epidemic-level spread of the co-infecting viruses. a recent study reported infrequent co-detection of rhinovirus with other viruses [21] , despite observations that rhinovirus continues to be shed for several weeks postresolution of symptoms [22] . negative associations have also been observed between the detection of influenza a, rsv, parainfluenza virus or coronavirus and co-detection of other respiratory viruses [8, 23, 24] , providing further evidence for a refractory period after initial infection during which the host is less likely to be infected by subsequent exposure to another respiratory virus. we used routine diagnostic testing data of specimens from both the community and hospitals at the victorian infectious diseases reference laboratory (vidrl) between 2002 and 2017 to describe relationships between respiratory viruses, with a focus on influenza a and rsv. from may 2002 to december 2017, 58 114 clinical specimens were collected from communities and hospitals and tested by polymerase chain reaction (pcr) for respiratory virus infection at vidrl. there were no inclusion criteria regarding symptoms, but it is assumed that testing was deemed clinically relevant. multiple specimen types were tested, but the majority were nose/throat swabs (64.2%) or nasopharyngeal aspirates (17.1%). the respiratory panel included nine viruses: adenovirus, influenza a, influenza b, parainfluenza virus, picornavirus (virus family includes rhinoviruses and enteroviruses), rsv, coronavirus (from 2010), human metapneumovirus (from 2012) and influenza c (from 2012). data were de-identified, but the date of birth, postcode of residence and sex of the patient were provided. data exclusions are shown in figure s1 . data from outbreaks, research and non-victorian residents were excluded (n = 10 325) as they followed different sampling methods. specimens collected from the same patient within 14 days were considered part of the same infection: where both specimens were positive for the same virus or both were negative, they were counted as one episode; where there were both positive and negative results, only the positive result was retained; and when two specimens were positive for different viruses, they were collapsed to represent one episode of co-detection. as a result, 8612 records were excluded. data from 2009 (n = 4232) were excluded as the influenza pandemic led to changes in referral and testing practices. data from 2016 to 2017 (n = 1293) were also excluded as the introduction of in-house testing at some referring hospitals led to a substantial decrease in samples tested by vidrl. demographic data of patients were compared using pearson's χ 2 test. weekly proportions positive for each virus were calculated to allow comparability and assess differences in virus epidemics between seasons. we compared our data to influenza notification rates in victoria obtained from the national notifiable diseases surveillance system (nndss) [25] to assess the representativeness of inter-seasonal peaks we observed. to assess timing and magnitude of epidemics, the proportion of positive specimens and the peak week of the epidemic were considered: those in the lowest quartile were considered early or small and those in the highest quartile were considered late or large. seasonality of viruses was assessed visually by time series analysis and for further investigation each virus was compared with influenza a and rsv using cross-correlations that estimated the association between peaks in epidemic curves at a lag or lead of up to 15 weeks. fisher's exact test was used to investigate any negative association between virus pairs among specimens with co-detections. multivariate logistic regression, adjusted for age category (<5, 5-19, 20-64 and ⩾65 years), sex and season, was used to produce odds ratios (or) and 95% confidence intervals for these associations and the chi-square test used to assess trend. adjustment for multiple comparisons was not performed [23, 26] . the significance level for all tests was set at p < 0.05. all data extraction, exclusion and analyses were performed in stata (version 14.2, statacorp, college station, texas). there were 33 652 pcr results from 2002 to 2015 (excluding 2009) included in this study. of these, 11 154 (33.1%) were positive for at least one of the nine viruses tested for (table 1) . picornavirus (rhinovirus) was detected most frequently (n = 5363, 33.1% of the positive specimens), followed by influenza a (n = 2259, 20.3%) and rsv (n = 1487, 13.3%). the proportion of tests positive for most viruses remained relatively stable (fig. 1) . however, there was a higher positivity rate of rsv pre-2009 (p < 0.001). the positivity rate of influenza a peaked and troughed; a year with a big epidemic was usually followed by a year with a smaller epidemic. of the influenza a-positive samples, 57.9% were a(h3n2), 23.7% were a(h1n1) and the remaining not subtyped. in most seasons, one subtype predominated, although in 2005 and 2014, the subtypes were observed to circulate with similar magnitude and timing and in 2013 they circulated as two distinct peaks of comparable magnitude. the rate of picornavirus detection increased from 2006 to 2010 and then decreased from 2011 to 2015 returning to a level similar to that observed at the beginning of the study period. more specimens tested were collected from males (53.9%) ( table 1) . as a proportion of total tests per sex, females were more likely to test positive for influenza a (p < 0.001) and metapneumovirus (p = 0.015), while males were more likely to test positive for picornavirus (p = 0.003) (table s1 ). there was no significant difference in sex distribution for the other viruses. patients residing in rural areas were significantly more likely to have a positive test than those in urban areas (p < 0.001). the same pattern was seen individually for rsv, parainfluenza virus and adenovirus. however, patients from urban areas were more likely to test positive for influenza a and metapneumovirus. associations between remaining viruses and area of residence were not significant. respiratory virus tests were most frequently requested in winter (june-august; n = 11 750, 34.9%) ( table 1 ) and were most likely to be positive in winter (p < 0.001). six of the nine viruses were most frequently detected in winter, but parainfluenza virus and metapneumovirus were most frequent in spring (september-november) and picornavirus was most frequent in autumn (march-may). tests positive for picornavirus were distributed relatively evenly across the seasons, so although the modal week was in autumn, a peak was less distinct compared to other viruses. the median age of positive tests was lower than that for all tests (36.9 (iqr: 2.4-61.5) and 45.3 (22.6-64.3) years, table 1 . demographic and temporal information for included specimens (table s1 ). notably, the median age of patients from whom specimens were collected increased fairly steadily from 27.6 years in 2002 to 61 years in 2015 (table s2) . time series analysis demonstrated annual seasonal peaks for influenza a and rsv (fig. 1) . peaks occurred most frequently in winter, with occasional peaks in late autumn (rsv) and early spring (influenza a). although influenza a virus circulation during summer in victoria is expected to be minimal, we observed increased influenza positivity rates in many summers during the study period, one of which was larger than its preceding winter peak (2013-2014). these inter-epidemic peaks were reflected in victorian notification rates (obtained from nndss) in summers from 2010 to 2011 onwards, visible as influenza activity not reaching zero as it had in previous summers (fig. 2) . seasonal peaks were also observed among the other viruses, except picornavirus, although they did not occur every year. picornavirus remained endemic throughout the year for the duration of the study period. cross-correlations were performed to ascertain whether the timing and magnitude of other viruses may differ relative to influenza a activity. results revealed a moderate to strong correlation between epidemic curves of influenza and rsv in 9 of 13 years. on average, where correlated, a seasonal epidemic of rsv occurred 6 weeks earlier than that of influenza a (table 2) , although there were 3 years where the epidemics occurred at similar times (2002, 2005 and 2006) . as a sensitivity analysis, we performed further cross-correlations to assess whether influenza subtype affected these interactions (table s3 ). the lag calculated for influenza a overall was consistently similar to that of the predominant influenza a subtype in a given season. in some years the lag calculated suggested influenza a(h1n1) circulated prior to rsv, however in these years the number of samples positive for influenza a(h1n1) was <10. no consistent pattern emerged when considering timing and magnitude of influenza a and rsv seasons ( table 2) : an early epidemic of one virus sometimes resulted in a later than usual epidemic of the other, but this was not always the case. likewise, a season with a high magnitude of infections with one virus did not necessarily result in a season with a low magnitude of the other. generally, influenza b epidemics occurred at a similar time to influenza a and in years that influenza a circulated early (2002, 2005, 2011, 2012) , influenza b activity was low (data not shown). co-detections of respiratory viruses occurred in 6.4% (n = 823) of positive samples. exploratory data analysis using univariate logistic regression suggested co-detections were more likely to occur in children under 5 years, males and during winter. odds of co-detection decreased as age increased. using the <5 year age group as the reference category, the ors (adjusted for sex and season) and corresponding 95% confidence intervals for co-infection co-detections occurred most frequently with adenovirus (40%), influenza c (39%) and coronavirus (20%), though the number of influenza c infections was small (n = 18) (fig. 3) . co-detections were rarest with picornavirus (10%) and influenza a and b (6%) infections. analysis of co-infections using fisher's exact test found a pattern of a negative association between detection of influenza a, rsv or picornavirus and co-detection of another virus (table 3) . these three viruses were involved in the highest number of significant negative associations (n = 5, table 3 ). no positive associations between viruses were considered statistically significant. multivariate logistic regression (adjusted for age group, sex and season) was used to further investigate the probability of co-detection given infection with influenza a, rsv and picornavirus (table 4 ). significant negative associations were observed for co-detection with all viruses where influenza a was detected and all but one virus for rsv and picornavirus detections (influenza b and human metapneumovirus, respectively). we used multi-year routine pcr testing data to establish patterns of respiratory virus circulation in victoria, australia. picornavirus (rhinovirus) was most frequently detected. children aged <5 years and those living in rural areas experienced a high burden of infection. time series analyses indicated the annual occurrence of epidemics for influenza a and rsv and less recurrent epidemics for influenza b, coronavirus and parainfluenza virus. picornavirus was observed to be endemic throughout the period of analysis. rsv epidemics generally began in autumn and peaked early winter, while influenza a began mid-winter and peaked late winter. the higher incidence of rsv observed pre-2009 may be a result for timing and magnitude of epidemic curves, proportion of specimens positive and the peak week of the epidemic were considered: those in the highest quartile were considered late or large; those in the lowest quartile were considered early or small. of the higher proportion of children under 5 years in our sample pre-2009, as rsv is considered the most important respiratory illness-causing pathogen in infants [27] . we observed summer peaks of influenza a in some years which was somewhat unexpected in a temperate climate but was only reflected in state-wide notification data after 2009. it is possible that inter-epidemic peaks we observed are a result of denominator data, while the increase in notifications resulted from a rise in testing after the 2009 pandemic [28] . in years that epidemics occurred, influenza b, coronavirus and parainfluenza virus peaked in winter and metapneumovirus in spring. like previous studies [15, 16] , time series analyses and crosscorrelations established distinct circulation patterns of rsv and influenza a. the two viruses rarely reached their epidemic peak concurrently, with rsv peaking an average of 6 weeks before influenza a. influenza a subtype did not affect cross-correlations: in seasons where significant correlation was observed, the lag calculated for influenza a overall was similar to that of the predominant subtype. in some seasons, influenza a(h1n1) appeared to circulate prior to rsv. however, in these seasons the number of a(h1n1) positive samples was <10, so the results should be interpreted with caution. an investigation into seasonal relationships between epidemic curves of other viruses was limited by the small proportion of positive tests. the endemic nature of picornavirus appeared to be unaffected by the circulation of other respiratory viruses, which supports previous observations of rhinovirus activity (most common species of picornavirus) [15, 16] . this may be a result of increased stability of the non-enveloped picornavirus during warmer months compared to other viruses, like influenza, which are restricted by temperature and humidity [29] . we also investigated the distribution and incidence of respiratory virus co-detection. improved availability and sensitivity of diagnostic tests has resulted in more regular detection of co-infections [30] , though the impact of viral co-infection on clinical severity remains unclear [31, 32] . a prospective household transmission study during the 2009 influenza pandemic reported that the infection wave caused by influenza a(h1n1)pdm09 was interrupted by a wave of non-influenza respiratory virus infections [33] . it found individuals infected with influenza a (h1n1)pdm09 were less likely to be infected by non-influenza respiratory viruses than non-infected individuals (rr: 0.32). further, there was a significant decrease in the duration of viral shedding in co-infections (of any respiratory viruses) compared to single infections. these observations suggest that such interactions may modulate influenza attack rate during outbreaks, thus shaping the epidemic and highlighting the importance of better understanding co-infections in the context of viral interference. we found co-detections of respiratory viruses in 6.4% of positive specimens, which falls in the 5.0-62.0% range of previous studies [31] . like other studies, we found co-infection was less likely with increasing age [34, 35] , which may be a consequence of pre-existing immunity or decreased viral shedding with increasing age [30] . we found adenovirus and coronavirus most likely to be part of a co-infection and influenza a and b least likely, corroborating results of a previous study [36] . while our data did not include patients' symptoms, immunological [37] and clinical [24, 38] data suggest that the effect of co-infection on clinical severity depends on the specific pathogens co-infecting the patient. infection by rhinovirus may result in temporary immunity of the host to infection by other respiratory viruses due to the production of cytokines [17] , thus resulting in a negative association between rhinovirus infection and co-infection with another virus [23] . moreover, it is believed to be the driver of epidemiological interaction between respiratory viruses at the population level, which is visible when two viruses may not reach their epidemic peak during the same period. while rhinovirus has been the focus of other studies investigating co-detection, we found that influenza a, rsv and picornavirus all had significant negative associations with co-detection of other viruses. results of further investigation by logistic regression adjusted for covariates that are predictors of codetection (sex, age and season) were compatible with influenza a, rsv and picornavirus conferring temporary immunity against infection by another respiratory virus. however, we cannot make causal inferences from the design used and therefore, cannot eliminate the role of other environmental factors. our study has some limitations. all inferences we made and indeed the majority of inferences made in other studies investigating respiratory virus interference are based on ecological data. with such data, we cannot determine whether events observed are the result of a biological mechanism, nor can we infer the direction of the postulated interaction, i.e. which virus impacts which. to make any form of causal inference, a prospective study that serially samples participants over multiple respiratory virus epidemics would be required. furthermore, we did not adjust for other potential drivers of viral interference, such as environmental (e.g. temperature, humidity), social and behavioural factors. also, we cannot rule out the possibility that our observations were the result of surveillance artefacts, that is, changes in testing patterns that are not a result of genuine fluctuations in viral circulation. while we excluded specimens isolated from outbreaks or for surveillance and samples from 2009 and 2016-2017 when testing patterns were obviously altered, we cannot be certain we controlled completely for this unknown. type/ subtype data for other viruses may have also improved the resolution of our findings, as other studies have noted variances in the timing of epidemics caused by different types of parainfluenza virus [4, 6] and rhinovirus [21] and there is scant information available for rsv. additionally, our study sample was drawn from patients ill enough to seek healthcare. as some viruses (such as picornavirus) are more likely to result in asymptomatic associations considered significant are bolded. the top cell represents the p-value for each measure of association and the bottom cell the or (and corresponding 95% ci) for infection. a no co-detections with these two viruses occurred infection than others, the distribution of viruses in our sample may differ from that in the population. finally, the referral base of paediatric samples for vidrl is limited as most victorian paediatric samples are forwarded to a children's hospital. given there is a high burden of respiratory virus infection in children, this may have limited our analyses. a strength of our data is that it spanned 16 years. accurately monitoring seasonal variation in respiratory virus epidemics has the potential to improve our understanding of interaction and interference between different respiratory viruses, although this remains challenging as surveillance systems for non-influenza respiratory viruses are limited in both scope and funding. our study confirms the existence of temporal relationships in the circulation of some respiratory viruses in victoria and provides further evidence to support the postulated effects of viral interference on magnitude and timing of respiratory virus epidemics. battle against respiratory viruses (brave) initiative temporal association between the influenza virus and respiratory syncytial virus (rsv):rsv as a predictor of seasonal influenza spatiotemporal characteristics of pandemic influenza epidemiology of acute lower respiratory disease in children interference between outbreaks of respiratory syncytial virus and influenza virus infection surveillance of respiratory viral infections by rapid immunofluorescence diagnosis, with emphasis on virus interference interference between outbreaks of respiratory viruses rhinoviruses delayed the circulation of the pandemic influenza a (h1n1) 2009 virus in france does viral interference affect spread of influenza? euro surveillance impact of the 2009 influenza a(h1n1) pandemic wave on the pattern of hibernal respiratory virus epidemics epidemiological changes of respiratory syncytial virus (rsv) infections in israel the 2009 pandemic influenza a (h1n1) coincides with changes in the epidemiology of other viral pathogens causing acute respiratory tract infections in children the impact of pandemic influenza a (h1n1) 2009 on the circulation of respiratory viruses influenza a/h1n1 2009 pandemic and respiratory virus infections early occurrence of influenza a epidemics coincided with changes in occurrence of other respiratory virus infections possible interference between seasonal epidemics of influenza and other respiratory viruses in hong kong the interferon response circuit: induction and suppression by pathogenic viruses in vitro growth profiles of respiratory syncytial virus in the presence of influenza virus investigating viral interference between influenza a virus and human respiratory syncytial virus in a ferret model of infection interval between infections and viral hierarchy are determinants of viral interference following influenza virus infection in a ferret model genotypic diversity, circulation patterns, and co-detections among rhinoviruses in queensland frequent detection of respiratory viruses without symptoms: toward defining clinically relevant cutoff values do rhinoviruses reduce the probability of viral co-detection during acute respiratory tract infections rate and influence of respiratory virus co-infection on pandemic (h1n1) influenza disease communicable diseases australia (cda) -national notifiable diseases surveillance system (nndss) no adjustments are needed for multiple comparisons respiratory syncytial virus infection the significance of increased influenza notifications during spring and summer of 2010-11 in australia influenza virus transmission is dependent on relative humidity and temperature co-infections with influenza and other respiratory viruses single and multiple respiratory virus infections and severity of respiratory disease: a systematic review clinical disease severity of respiratory viral co-infection versus single viral infection: a systematic review and meta-analysis intense co-circulation of non-influenza respiratory viruses during the first wave of pandemic influenza ph1n1/2009: a cohort study in reunion island community-acquired respiratory viruses and co-infection among patients of ontario sentinel practices epidemiology of viral respiratory infections in australian working-age adults (20-64 years multiple versus single virus respiratory infections: viral load and clinical disease severity in hospitalized children single versus dual respiratory virus infections in hospitalized infants: impact on clinical course of disease and interferongamma response viral etiology and the impact of codetection in young children presenting with influenza-like illness acknowledgements. we thank the laboratory staff members from victorian infectious diseases reference laboratory who undertook respiratory virus testing.financial support. the who collaborating centre for reference and research on influenza is supported by the australian government department of health.conflict of interest. none. key: cord-263245-2qub96mz authors: singh, d.; joshi, k.; samuel, a.; patra, j.; mahindroo, n. title: alcohol-based hand sanitisers as first line of defence against sars-cov-2: a review of biology, chemistry and formulations date: 2020-09-29 journal: epidemiol infect doi: 10.1017/s0950268820002319 sha: doc_id: 263245 cord_uid: 2qub96mz the pandemic due to severe acute respiratory syndrome coronavirus 2 (sars-cov-2) has emerged as a serious global public health issue. since the start of the outbreak, the importance of hand-hygiene and respiratory protection to prevent the spread of the virus has been the prime focus for infection control. health regulatory organisations have produced guidelines for the formulation of hand sanitisers to the manufacturing industries. this review summarises the studies on alcohol-based hand sanitisers and their disinfectant activity against sars-cov-2 and related viruses. the literature shows that the type and concentration of alcohol, formulation and nature of product, presence of excipients, applied volume, contact time and viral contamination load are critical factors that determine the effectiveness of hand sanitisers. the outbreak of respiratory infection with severe acute respiratory syndrome coronavirus 2 (sars-cov-2) virus has emerged as a serious global public health threat [1] . it is the third time in the last two decades that an animal coronavirus has emerged to cause epidemic infection in humans. the disease was first reported in wuhan province of china at the end of 2019 but rapidly spread to infect more than 23 million people as of 25 august 2020, and has been associated with > 800 000 deaths [2] . the world health organization (who) declared a pandemic on 11 march 2020 and the infection has spread across almost all countries and regions of the world. most infections appear to be asymptomatic or with mild flu-like symptoms but severe and life-threatening presentations including pneumonia, fever, nausea and gastrointestinal upset have been associated with individuals with predisposing factors, particularly age, respiratory insufficiency, diabetes and obesity, among others [3] . the who, and national disease control agencies, have continuously emphasised the importance of hand hygiene to reduce spread of the virus. who guidelines recommend maintaining hand hygiene, by frequent washing using soap and water for at least 20 s especially after going to the bathroom, before eating and after coughing, sneezing or blowing one's nose. when soap and water are not available, the food and drug administration (fda) recommends sanitising of non-visibly soiled hands with an alcoholbased agent containing 80% v/v ethanol or 75% v/v isopropanol [4] . enveloped viruses such as coronavirus and influenza a h1n1 are able to survive on inanimate surfaces for long periods [5] . it has been reported that some covid-19 patients discharged the virus in their stool for up to 73 days after symptom onset [6] , and as diarrhoea is a common symptom, faecal to oral cross-transmission is likely [7] , and hence maintaining effective hand hygiene is paramount. alcohol-based hand sanitisation is widely considered to be effective to reduce or eliminate bacterial/viral load, but with variable compliance rates [8] . the alcohols, ethanol, isopropanol and n-propanol as used for disinfection are commonly applied in the form of hand rub rinses, gels and foams. owing to the increasing demand for hand sanitisation to control the spread of sars-cov-2, some manufacturers have resorted to their own formulations, which are not validated and licensed for use. to combat this, the fda, who, the united states pharmacopeia (usp) and the central drugs standard control organization (cdsco), india, have produced guidelines for the formulation and manufacture of such preparations [4, 9, 10] . this review assesses available information on the composition, formulation and effectiveness of alcohol-based hand disinfection products with specific reference to their activity against sars-cov-2. sars-cov-2 is a new member of the family coronaviridae, order nidovirales, and comprise of two sub-families, coronavirinae and torovirinae [11] ; it is the seventh coronavirus known to infect humans [12] . sars-cov-2 is relatively large in size (0.12 μm) and characterised by the presence of highly glycosylated spikes on the protein membrane in a crown-like arrangement, hence the name, corona (fig. 1 ). it has a single-stranded positive-sense rna genome of 29 891 nucleotides. the glycosylated spike protein binds to the host angiotensin converting enzyme-2 (ace-2) protein which serves as a functional receptor for entry into host respiratory cells. this receptor also binds the earlier sars-cov but with 10-20 times less affinity than for sars-cov-2 spike protein [13, 14] . several antimicrobial compounds have been utilised for hand disinfection and include, among others, alcohols, chlorhexidine, chloroxylenol, hexachlorophene, benzalkonium chloride, cetrimide, triclosan and povidone-iodine [15] . the alcohols, namely ethanol and isopropanol, are most commonly used for skin disinfection due to their broad activity against bacteria, viruses and fungi [16] ; their mode of action against enveloped viruses is shown in figure 2 . lipid membrane dissolution and protein denaturation are key mechanisms of the antimicrobial action of ethanol, leading to the disruption of membrane and inhibition of metabolism [17, 18] . alcohols are amphiphilic compounds, as they possess both hydrophilic and lipophilic (hydrophobic) properties that facilitate their entry through the viral envelope. the outermost membrane of sars-cov-2 comprises lipids bound together by an alkane chain of hydrophobic fatty acids. contact of the virus with an alcohol leads to alteration in its membrane fluidity [19] . the presence of polar oxygen atoms weaken the lipophilic interactions between the non-polar residues, and increase the internal affinity of the membrane for water, thus destabilising and denaturing the protein structure [17] . the antimicrobial mechanism of alcohol against enveloped viruses is similar to that for bacteria as both have a lipid-rich outer membrane. non-enveloped viruses are relatively more resistant to this mechanism due to the lack of a lipid membrane. the family coronaviridae is comprised of four groups (table 1) . sars-cov-2 is considered to be taxonomically related to group 2 coronaviruses [20, 21] . virus and bovine viral diarrhoea virus (bvdv) are used for testing the effectiveness of chemical disinfectants and antiseptics against enveloped viruses according to dvv/robert koch institute (rki) guidelines [22] . the modified vaccinia ankara (mva) virus can also be used as a surrogate model for this purpose as it exhibits high stability against alcohol-based inactivation. the latter virus does not replicate in humans, thus eliminating the risk of disease through unintentional inoculation [23, 24] . bovine coronavirus (bcv) has been used as a surrogate virus for sars-cov [25] , and owing to its high (80%) relatedness to sars-cov-2, consequently may have potential value as a surrogate test agent for the latter. the two most widely used guidelines for testing and regulation of hand disinfectants are the european committee for standardization (cen) and the food and drug administration (fda), according to standards set by the american society for testing and materials (astm). en 1499 and en 1500 are the standard methods related to hygienic hand wash and hygienic hand disinfection respectively [26, 27] . in en 1499, agents are tested against a reference nonmedicated soap and in en 1500 against 60% v/v isopropanol, both applied for 1 min. in the latter standard, the test hand rub formulation should not be significantly inferior, in terms of log reduction of the challenge microbe, compared with the reference alcohol-based product. en 14476 is the standard method for evaluating the virucidal activity of disinfectants [28] and is based on an in-vitro quantitative suspension test in which agents should exhibit a minimum of 4-log reduction in viability of the microbe. poliovirus, adenovirus and murine norovirus serve as the basis for efficacy evaluation of surface disinfectants. pren 16777 is also a quantitative virucidal test method and is recommended for nonporous surfaces (in-vivo carrier test); a 4-log reduction is specified and ready-to-use surface disinfectants should be tested undiluted using adenovirus and murine norovirus as test pathogens. this test method simulates practical conditions and together with en 14476 forms the basis for biocidal product registration in europe [29] . a finger pad test method designed to compare the virus-eliminating effectiveness of hand washing and hand rubbing sanitisers using at least three healthy participants. exposure time should be 10-20 s for hand washing and 20-30 s for a hand sanitation. the recommended test viruses include adenovirus 5, feline calicivirus, rotavirus, rhinovirus and murine norovirus at a minimum of 10 4 infectious units with or without a soil load. a 4-log reduction in virus load must be demonstrated by the test product in the presence and absence of 5% foetal bovine serum [30] . this method evaluates the virucidal activity of hand wash and hand rub agents against viruses and is claimed to better reflect actual working conditions as it incorporates mechanical friction during whole-hand decontamination. at least three healthy participants are required and following application of virus suspension, the specified product exposure times are 10-20 s for hand washing and 20-30 s for a sanitiser. test viruses include adenovirus type 2 or 5, feline calicivirus, rotavirus, rhinovirus and murine norovirus in the presence and absence of 5% foetal bovine serum as an interfering substance to simulate dirty conditions [31, 32] . this method determines the efficacy of test disinfectants to inactivate viruses on disk carriers of brushed stainless steel, which act as a surrogate material for hard, non-porous environmental surfaces and medical devices [ alcohol type and concentration most alcohols exhibit a broad spectrum of germicidal activity against vegetative bacteria, viruses and fungi. in general, isopropanol is considered to have better activity against bacteria, while ethanol is more potent against viruses. however, the degree of effect depends on the percentage concentrations of the alcohol and the physical properties of the target microorganism. isopropanol is more lipophilic than ethanol and is consequently less active against hydrophilic viruses such as polioviruses. being a lipophilic enveloped virus, sars-cov-2 exhibits greater susceptibility to isopropanol than ethanol [20, 37, 38] . the optimum bactericidal concentrations of alcohols range from 60% to 90% v/v solutions in water but are generally ineffective against most microorganisms below 50% v/v [39] . the effect of different concentrations of alcohol against enveloped viruses is shown in table 2 [25, 37, [40] [41] [42] [43] [44] [45] [46] . a recent study has shown that >30% concentrations of ethanol or isopropanol were effective in inactivating sars-cov-2 within 30 s [47] . propanol has a marginally higher boiling point than ethanol, hence, the drying time of isopropanol is slightly longer compared to ethanol [48] . the who has recommended two alcohol-based hand sanitiser formulations which differ only in their alcohol constituent, and is widely followed throughout the world. formulation isopropyl alcohol 75% v/v, glycerol 1.45% v/v, hydrogen peroxide 0.125% v/v [49] . due to the inherent variability of raw materials and the volatility of alcohol, and in response to the covid-19 pandemic, the united states pharmacopeia has issued a revision of who formulation 2 by increasing the concentration of isopropanol to 91% v/v [10] . an n-propanol-based formulation has not been proposed owing to the lack of safety data on human use [49] . in march 2020, the fda recommended the industry to use either of the two who formulas but emphasised that ethanol should not be used at a concentration of <94.9% by volume. in a separate fda guideline addressing the preparation and distribution of alcohol for incorporation in hand disinfectants, mention was made of the search for other active constituents including the use of denaturants such as acetone [50] . there was also comment that the recommended amount of glycerol in the who formulation might negatively impact the effectiveness of isopropanol [50] . nevertheless, both who formulations have been shown to be effective against sars-cov-2 [47] . indeed, with regards to the latter, cdc recommends the use of alcohol-based sanitisers containing >60% ethanol or 70% isopropanol for personnel working in healthcare settings [51] . this is supported by the finding that the who formulation containing isopropanol had higher activity against enveloped viruses [52] . the virucidal efficacy of hand sanitisers depends on several factors. as illustrated by the ishikawa diagram (fig. 3) showing the key factors which determine the efficacy of alcohol against sars-cov-2. the most commonly used formulations for hand sanitisers are rinse, foam, gel, wipes and spray. the 70% ethanol-based liquid products have proved highly effective against the non-enveloped viruses, poliovirus and adenovirus following exposure for 30 s [53] . alcohol-based hand rubs in the form of foam, rinse and gel did not differ significantly in trials of antimicrobial activity but the application volume and drying time had a profound effect on their efficacy [54] . another study, however, found that alcoholbased hand wipes were comparable in activity to foam and gel products against enveloped influenza (h1n1) virus. this was ascribed to better mechanical friction achieved with wipes, resulting in additional physical removal of virus that might survive the antimicrobial treatment [40] . indeed, another comparative study concluded that hand gels are less effective for hand hygiene because of a shorter application time (<30 s) and therefore should not replace alcohol-based liquid hand disinfectants, or used as first choice agents [55] despite the benefit of reducing skin irritation and dryness associated with liquid alcohol agents preparations. however, gel preparations containing 62% ethanol have been reported to be superior to 70% ethanol for the inactivation of surrogate coronaviruses mhv and tgev on hard surfaces [41] . foams have an advantage of better compliance by users due to ease of handling, non-spilling and non-stickiness. bis-peg12-dimethicone is commonly used as the foaming agent. it is recommended that an amount equivalent in size to a golf ball should be applied to hands [56] ; they also have the added benefit of the shortest drying times compared with rinses and gels [57] . the approximate drying times of different alcoholbased formulations are given in table 3 . an increase in the volume of alcohol and contact time results in increased efficacy of alcohol-based hand sanitisers. one pump dispenser push releases approximately 1.5 ml of gel containing 70% alcohol has been found to be insufficient for complete coverage of both hands and hence, do not comply with astm efficacy standards [57] . the use of 3 ml volume for foam, rinse and gel sanitisers containing 70%, 80% and 90% alcohol, respectively, is necessary to meet en 1500 efficacy requirements, but the drying times of all preparations exceeded 30 s [54] . the amount of sanitiser used also depends on the size of the subject's hands; females are relatively smaller (mean of eight volunteers 148.39 cm 2 , rsd = 5.17), and a lower volume of the agent could be sufficient when compared with men's hands (mean of eight volunteers 183.63 cm 2 , rsd = 7.5) [58] . it is generally acknowledged that the ideal application volume is unknown, but us national guidelines suggest that a drying time of <15 s is insufficient [59] , while the who recommends use of a 'palmful' of product and that the hand-hygiene process should take at least 20 s [60] . rotter et al. found that 3 ml of the en 1500 reference product (isopropanol) takes more than 49 s to dry, despite a specified rub-time of 30 s [61] . similarly, a trial on disinfection of volunteer hands artificially contaminated with escherichia coli k12 showed that who formulations containing either ethanol or isopropanol did not comply with the en 1500 requirement as 60 s were taken to achieve the required log reduction. this led to the proposal that the ethanol concentration should be changed from 80% v/v to 80% w/w (equivalent to 85% v/v), and for isopropanol from 75% v/v to 75% w/w (equivalent to 80% v/v) [62] . the contact time of the agent is also relevant as a survey showed that the majority of nursing staff took only 6-24 s for hand cleansing [63] . it has also been suggested that better compliance might be achievable in the hospital setting through listening to background music during the process [64] . glycerin is added in hand sanitisers as a humectant to reduce loss of skin moisture. who-recommended formulations contain glycerin but other nontoxic or allergenic emollients miscible in water and alcohol are not permitted for skin care [49] . studies have shown that glycerol can reduce the efficacy of isopropanol-based sanitiser through agglomerates of flaking skin cells forming in the sticky glycerol [65] . a mixture of ethylhexylglycerin, dexpanthenol and a fatty alcohol serves as a suitable alternative with no effect on hand rub efficacy [66] . indeed, the removal of glycerol from a formulation markedly increased the bactericidal activity of an isopropanol-based sanitiser [67] . this negative impact of glycerol has been noted in fda guidelines regarding temporary compounding of alcoholbased hand sanitisers by industry during the covid-19 pandemic [4] . similarly, reducing the glycerol content from 1.45%, as per the who formulation, to 0.5% provided a better balance between antimicrobial efficacy and skin tolerance [68] . an extract of the aloe vera plant has also been used as an emollient [69] . ph human and canine corona viruses are reported to be more stable at a slightly acidic than alkaline ph [70, 71] but mild alkaline (ph 8) conditions are sufficient to induce conformational changes in the spike protein of coronavirus mouse hepatitis virus [72] . both high and low ph cause inactivation of sars-cov [73] . the virucidal activity of ethanol against poliovirus and ms2 phage is significantly increased on the addition of sodium hydroxide [74] due to protein denaturation [75] . sodium hydroxide has also been shown to have cidal activity against surface dried lipid enveloped human immunodeficiency virus (hiv), bovine diarrhoeal virus and pseudorabies virus [76] . other anti-viral agents include acetic acid and calcium hydroxide against influenza virus on hard and non-porous surfaces [75] . moreover, citric acid and urea (2%) have been reported to increase the effectiveness of alcohol-based sanitisers [37] ; citric and malic acid, in combination with 70% alcohol have also been suggested to enhance killing of rhinovirus on hands [77] . it is quite likely that the effect of hand sanitisers is reduced in the presence of dirt or soil on hands. a number of interfering substances have been used to simulate dirty conditions including foetal calf serum, bovine serum albumin and sheep erythrocytes according to dvv, rki, astm and cen standard guidelines [37, 78] . soap hand wash coupled with an alcohol gel sanitiser was shown to be more effective than either agent used alone, and activity persisted for longer [79] . these findings are corroborated by other studies showing increased reduction of murine norovirus with a wash-sanitiser regimen compared to washing with 70% ethanol alone in the presence of a high level of organic loads [80] . however, it is worth noting that hand washing with soap and water alone was found to be more effective than alcoholbased rubs for hands soiled with meat [81] . hand hygiene by washing hands with soap and water or with alcohol-based hand sanitisers are primary preventive measures against the spread of sars-cov-2. this review of the literature shows that several factors are pertinent to the antiviral activity of sanitising agents. alcohol-based agents cause dissolution of the lipid membrane and denature proteins, thereby disrupting the virus membrane and inhibiting metabolism. the concentration of alcohol in hand-cleansing products, the volume used, contact time, degree of soiling, product formulation and use of excipients are some of the critical factors that affect the efficacy of alcohol against viruses. due to its relatively greater lipophilicity, isopropanol is considered more effective than ethanol against sars-cov-2. to ensure a greater than 3-log reduction of sars-cov-2, a hand sanitiser should ideally contain >80% v/v ethanol or >75% v/v isopropanol. however, recent study which suggests that ethanol and isopropanol used above 30% v/v is effective against sars-cov-2 [47] requires confirmation by other investigators. gel-based hand sanitisers are reported to have more efficacy against enveloped viruses while foam-based preparations have the most rapid drying time. it is recommended that at least 3 ml of product should be used with a total contact time of around 45-50 s. soiled hands can limit the efficacy of alcohol-based products as well as the presence of excipients; for isopropanol-based formulations, the replacement of glycerol with other emollients is recommended. similarly, the addition of sodium hydroxide potentiates the antiviral activity of alcohols. further studies are clearly needed on the optimum design and delivery form of agents for efficient hand decontamination of sars-cov-2. such knowledge will prove of benefit for preparedness against other highly infectious viruses. the species severe acute respiratory syndrome-related coronavirus: classifying 2019-ncov and naming it sars-cov-2 clinical features of patients infected with 2019 novel coronavirus in wuhan policy for temporary compounding of certain alcohol-based hand sanitizer products during the public health emergency immediately in effect guidance for industry. u.s. department of health and human services. food and drug administration survival of enveloped and non-enveloped viruses on inanimate surfaces enteric involvement of severe acute respiratory syndrome-associated coronavirus infection public health might be endangered by possible prolonged discharge of sars-cov-2 in stool hand sanitizers: a review on formulation aspects, adverse effects, and regulations monitoring of quality standards of hand sanitizers as per drugs and cosmetics act 1940, and rules made thereunder and expediting the licensing of manufacturers of such products (2020) government of india, directorate general of health services, central drugs standard control organization compounding alcohol-based hand sanitizer during covid-19 coronaviruses: an overview of their replication and pathogenesis the proximal origin of sars-cov-2 cryo-em structure of the 2019-ncov spike in the prefusion conformation genomic characterization and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding who guidelines on hand hygiene in health care (advanced draft) global patient safety challenge spectrum of antimicrobial activity and user acceptability of the hand disinfectant agent sterillium gel interaction of alcohol with phospholipid membrane: nmr and xrd investigations on dppc-hexanol system hand sanitizers: a review of ingredients, mechanisms of action, modes of delivery, and efficacy against coronaviruses adaptation of membrane lipids to alcohols coronavirus envelope protein: current knowledge a review of the 2019 novel coronavirus (covid-19) based on current evidence povidone-iodine hand wash and hand rub products demonstrated excellent in vitro virucidal efficacy against ebola virus and modified vaccinia virus ankara, the new european test virus for enveloped viruses can vaccinia virus be replaced by mva virus for testing virucidal activity of chemical disinfectants? virucidal activity of a new hand disinfectant with reduced ethanol content: comparison with other alcohol-based formulations chemical disinfectants and antiseptics. hygienic hand wash. test method and requirements chemical disinfectants and antiseptics. hygienic handrub. test method and requirements chemical disinfectants and antiseptics -quantitative suspension test for the evaluation of virucidal activity in the medical area -test method and requirements (phase 2/ step 1) chemical disinfectants and antiseptics. quantitative non-porous surface test without mechanical action for the evaluation of virucidal activity of chemical disinfectants used in the medical area. test method and requirements (phase 2/step 2) astm e1838 determining the virus-eliminating effectiveness of hygienic handwash and handrub agents using the fingerpads of adults astm e2011 -virucidal activity for hygienic hand washing. valencian institute of microbiology standard test method for evaluation of hygienic handwash and handrub formulations for virus-eliminating activity using the entire hand mycobactericidal, sporicidal activities of chemical agents astm e2197-17e1 -standard quantitative disk carrier test method for determining bactericidal, virucidal, fungicidal, mycobactericidal, and sporicidal activities of chemicals guideline of deutsche vereinigung zur bekampfungder viruskrankheiten e.v. (dvv; german association for the control of virus diseases) and robert koch institute (rki; german federal health authority) for testing the virucidal efficacy of chemical disinfectants in the human medical area leitline der deutschen vereinigung zur bekämpfung der viruskrankheiten (dvv) e.v. zur quantitativen prüfung der viruziden wirksamkeit chemischer desinfektionsmittel auf nicht-porösen oberflächen development and virucidal activity of a novel alcohol-based hand disinfectant supplemented with urea and citric acid the e protein is a multifunctional membrane protein of sars-cov the relationship of concentration and germicidal efficiency of ethyl alcohol analysis of alcohol-based hand sanitizer delivery systems: efficacy of foam, gel, and wipes against influenza a (h1n1) virus on hands inactivation of surrogate coronaviruses on hard surfaces by health care germicides chemical disinfection of non-porous inanimate surfaces experimentally contaminated with four human pathogenic viruses inactivation of respiratory syncytial virus by detergents and disinfectants inactivation of sars coronavirus by means of povidone-iodine, physical conditions and chemical reagents stability and inactivation of sars coronavirus efficacy of various disinfectants against sars coronavirus inactivation of severe acute respiratory syndrome coronavirus 2 by who-recommended hand rub formulations and alcohols comparison of the efficacy and drying times of liquid, gel and foam formats of alcohol-based hand rubs who recommended hand-rub formulations temporary policy for manufacture of alcohol for incorporation into alcohol based hand sanitizer products during the public health emergency (covid-19); guidance for industry u.s. department of health and human services, food and drug administration, center for drug evaluation and research (cder) cdc statement for healthcare personnel on hand hygiene during the response to the international emergence of covid-19 virucidal activity of world health organization-recommended formulations against enveloped viruses, including zika, ebola, and emerging coronaviruses efficacy of hand rubs with a low alcohol concentration listed as effective by a national hospital hygiene society in europe the relative influences of product volume, delivery format and alcohol concentration on dry-time and efficacy of alcohol-based hand rubs limited efficacy of alcohol-based hand gels efficacy of ethanol-based hand foams using clinically relevant amounts: a cross-over controlled study among healthy volunteers less and less-influence of volume on hand coverage and bactericidal efficacy in hand disinfection impact of the amount of hand rub applied in hygienic hand disinfection on the reduction of microbial counts on hands healthcare infection control practices advisory committee guideline for hand hygiene in health-care settings. recommendations of the healthcare infection control practices advisory committee and the hipac/shea/apic/idsa hygiene task force who guidelines on hand hygiene in health care impact of shortening the duration of application and the standardized rubbing sequence as well as the reduction of the disinfectant volume used for the hygienic hand rub with 2-propanol (60% v/ v) testing of the world health organization recommended formulations in their application as hygienic hand rubs and proposals for increased efficacy comparison of waterless hand antisepsis agents at short application times: raising the flag of concern effect of music on surgical hand disinfection: a video-based intervention study influence of glycerol and an alternative humectant on the immediate and 3-hours bactericidal efficacies of two isopropanol-based antiseptics in laboratory experiments in vivo according to en 12791 glycerol significantly decreases the 3-hours efficacy of alcohol-based surgical hand rubs modified world health organization hand rub formulations comply with european efficacy requirements for preoperative surgical hand preparations glycerol content within the who ethanolbased handrub formulation: balancing tolerability with antimicrobial efficacy hand sanitisers amid covid-19: a critical review of alcohol-based products on the market and formulation approaches to respond to increasing demand effect of ph and temperature on the infectivity of human coronavirus 229e canine coronavirus inactivation with physical and chemical agents monoclonal antibodies to the peplomer glycoprotein of coronavirus mouse hepatitis virus identify two subunits and detect a conformational change in the subunit released under mild alkaline conditions inactivation of the coronavirus that induces severe acute respiratory syndrome, sars-cov the use of bacteriophage ms2 as a model system to evaluate virucidal hand disinfectants inactivation of avian influenza virus using four common chemicals and one detergent resistance of surface-dried virus to common disinfection procedures effectiveness of hand sanitizers with and without organic acids for removal of rhinovirus from hands the effects of test variables on the efficacy of hand hygiene agents a close look at alcohol gel as an antimicrobial sanitizing agent hand hygiene regimens for the reduction of risk in food service environments a method of assessing the efficacy of hand sanitizers: use of real soil encountered in the food service industry acknowledgements. the doctoral research fellowship from the university of petroleum and energy studies to jeevan patra is gratefully acknowledged.conflict of interest. the authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.data availability statement. the datasets supporting the conclusions of this review article are included within the article and in references listed in the paper. key: cord-309001-erm705tg authors: liu, q.; song, n. c.; zheng, z. k.; li, j. s.; li, s. k. title: laboratory findings and a combined multifactorial approach to predict death in critically ill patients with covid-19: a retrospective study date: 2020-06-30 journal: epidemiol infect doi: 10.1017/s0950268820001442 sha: doc_id: 309001 cord_uid: erm705tg to describe the laboratory findings of cases of death with coronavirus disease 2019 (covid-19) and to establish a scoring system for predicting death, we conducted this single-centre, retrospective, observational study including 336 adult patients (≥18 years old) with severe or critically ill covid-19 admitted in two wards of union hospital, tongji medical college, huazhong university of science and technology in wuhan, who had definite outcomes (death or discharge) between 1 february 2020 and 13 march 2020. single variable and multivariable logistic regression analyses were performed to identify mortality-related factors. we combined multiple factors to predict mortality, which was validated by receiver operating characteristic curves. as a result, in a total of 336 patients, 34 (10.1%) patients died during hospitalisation. through multivariable logistic regression, we found that decreased lymphocyte ratio (lymr, %) (odds ratio, or 0.574, p < 0.001), elevated blood urea nitrogen (bun) (or 1.513, p = 0.009), and raised d-dimer (dd) (or 1.334, p = 0.002) at admission were closely related to death. the combined prediction model was developed by these factors with a sensitivity of 100.0% and specificity of 97.2%. in conclusion, decreased lymr, elevated bun, and raised dd were found to be in association with death outcomes in critically ill patients with covid-19. a scoring system was developed to predict the clinical outcome of these patients. since the outbreak of disease caused by a new virus in wuhan, china in december 2019, hundreds of thousands of people have been infected with the virus. as of march 2020, the cumulative number of confirmed diagnoses on the chinese mainland has been over 80 000, but there has been no new domestically transmitted case reported for several days, which indicates that the epidemic in china has been basically controlled. however, the viral disease has swept into at least 190 countries and killed tens of thousands of people. over the last few weeks, the number of cases of this disease outside china has increased dozens of times. therefore, the world health organization (who) had made an assessment on 11 march that the disease could be characterised as a pandemic [1] . the pathogen of the disease was first isolated by chinese scientists on 7 january 2020, and was identified as a new type of coronavirus [2] , which was initially known as 2019-ncov and recently named sars-cov-2 [3] . it has the characteristics of a typical coronavirus family and belongs to the beta coronavirus 2b lineage. although the initial epidemiological investigation suggested that it was transmitted to humans through wildlife, the distinct human-to-human transmission phenomenon has been confirmed in subsequent studies [4] . full-genome sequencing and phylogenic analysis suggest that it is similar in structure to severe acute respiratory syndrome (sars) virus, and mainly causes respiratory tract infections in humans [5] , which was designated as coronavirus disease 2019 (covid-19) later by who. several descriptive studies have confirmed that most infected patients were mildly symptomatic or even asymptomatic [6, 7] . however, there were still many patients who developed severe pneumonia or even death [8, 9] . the relevant indicators of this part of the patients, especially the patients who died, need to be of significant concern. a data-driven analysis stated the hubei province (except wuhan) has an estimated case-fatality rate (cfr) of 1.41% [10] , which was also confirmed by a recent retrospective study involving 1099 patients [11] . however for critically ill patients, the mortality rate is much higher. a study from jin yin-tan hospital of wuhan compared the critically ill and non-severe patients, and the results showed that the mortality of critically ill patients with covid-19 was considerable [9] . however, there is still no specific and effective treatment except for meticulous supportive care for this disease. no solid evidence proved any antiviral agents could improve outcomes in covid-19. even lopinavir−ritonavir treatment, previously thought to be effective, has been shown by studies that no benefit was observed in hospitalised adult patients with severe covid-19 [12] . in such cases, critically ill patients could rapidly develop acute respiratory distress syndrome (ards), organ dysfunction and other serious complications and eventually death. therefore, it is significant to estimate risk factors for severe disease and death, so that we could more efficiently focus on medical resources to treat patients who may have poor prognoses in efforts to reduce mortality. although there have been previous articles describing the clinical course of the disease or some factors that predicted death [13] [14] [15] because many patients were still in the middle stage of the disease at the time of the study and did not reach their clinical outcome, the collected clinical data were inevitably incomplete, and the results were less accurate. under the situation that the epidemic is basically under control, most patients have their definite outcomes, either cured or died, so that the research would be more accurate, which was why we conducted this research study at this time. in addition, we used a combined predictive system with several variables, which also ensures the validity of the prediction. in this research study, we collected the clinical data of patients with covid-19 admitted in the hospital, encompassing laboratory indexes and their clinical outcomes (cure or death). we aimed to describe laboratory findings of cases of death, compare them with cured patients, and finally design a multifactor prediction model, which is expected to provide early identification for patients with clinically severe covid-19. this single-centre, retrospective, observational study included adult patients (≥18 years old) with severe or critically ill covid-19 admitted in two wards of union hospital, tongji medical college, huazhong university of science and technology in wuhan. all patients were diagnosed with covid-19 according to who interim guidance and they all had a definite outcome (death or discharge) between 1 february 2020 and 13 march 2020 ( fig. 1 ). demographical data, laboratory indexes, medical history, underlying diseases and outcome data were extracted from the blood screening test and electronic medical records. all data were examined by one independent physician through the patient's paper charts. the study was conducted in accordance with the principles of the declaration of helsinki. informed consent was not obtained from patients as all data were retrieved retrospectively from the laboratory testing information system and no additional blood samples were taken. this study was approved by the ethics review board of wuhan union hospital. the presence of sars-cov-2 in respiratory specimens was detected by next-generation sequencing or real-time rt-pcr methods to confirm the diagnosis of covid-19 as described in the previous article [16] . underlying diseases included hypertension, cardiovascular disease, diabetes, malignancy, cerebrovascular disease, chronic obstructive pulmonary disease (copd), chronic kidney disease, chronic liver disease, organ transplant history and undergoing general anaesthesia surgery within 1 month. the duration from the onset of disease to hospital admission was recorded. the neutrophil ratio (neur), is the percentage of neutrophils to the total number of all white blood cells (wbcs). similarly, the lymphocyte ratio (lymr) is the number of lymphocytes as a percentage of the number of wbcs. statistical analyses were performed by spss 17.0 (spss inc, chicago il, usa). continuous and categorical variables were presented as median, interquartile range (iqr) and n (%), respectively. the mann−whitney u test, χ 2 test or fisher's exact test were used. binary logistic regression was conducted further. we used both single variable and multivariable logistic regression to verify those factors. variables with statistically significant differences in the single variable analysis were included in the multivariable analysis, and several variables with potential for bias were excluded. this was followed by a backwards stepwise logistic regression. the final model was used to predict death and that sensitivity/specificity/roc curves were produced. a p-value < 0.05 was considered statistically significant. the study included 336 adult patients with severe or critically ill covid-19 eventually. the median age was 65 years (iqr, 51-69), and 169 (50.3%) were men. the median duration from first symptoms to hospital admission was 2 days (iqr, 1-4). in a total of 336 patients, 182 (54.2%) had one or more underlying conditions. in all, 34 (10.1%) patients died during hospitalisation while 302 (89.9%) were discharged. compared with survivors, non-survivor patients were significantly older with a median age of 74 years (iqr, 64-78) vs. 64 years (iqr, 51-68) ( table 1) . (1) laboratory findings laboratory parameters of all patients were recorded on day of hospital admission, then divided into survivor or non-survivor groups according to their clinical outcome. the levels of wbc count, neutrophil (neu), neutrophil ratio (neur), blood urea nitrogen (bun), d-dimer (dd) and c-reactive protein (crp) were higher in non-survivor patients, together with the reduction of lymphocyte ratio (lymr) ( table 1) . (1) results of logistic regression analysis in single variable analysis, age, wbc count, lymr, neur; serum albumin (alb), bun; prothrombin time (pt), fibrinogen (fib), dd; and crp were associated with death (table 1) . wbc, lym, lymr, are all related to the wbc count, and may affect each other or have contained relationships. to avoid bias in the prediction model due to correlation, it is better for the variables in a model to be completely independent of each other. similarly, fib, pt and dd all reflected the coagulation function, and we chose only one to represent it. therefore, we excluded wbc, pt and fib in the subsequent analysis. this was followed by table 2 . eventually, we found that lymr, bun and dd at admission were closely related to death. (1) combination of predictors and development of predictive model three laboratory indicators were combined to provide a predictive probability value for the outcome of death in covid-19 patients, which was expressed in terms of pre. the roc curve was then used to evaluate the predictive efficiency of the combined predictor and individual factors for the outcome of death, which is shown in figure 2 . according to it, the area under the curve (auc) and cut-off values of the three factors were calculated ( table 3) . as demonstrated, the optimal thresholds of lymr, bun and d-dimer were 8.615%, 5.95 mmol/l and 1.56 μg/ml. on the basis of the logistic regression model and the roc curves, the scoring system for prediction of death was developed with three variables including lymr, bun and dd. when lymr < 8.615%, 1 point is counted, otherwise 0 points are counted. similarly, when bun ≥ 5.95 mmol/l or d-dimer ≥ 1.56 μg/ml, 1 point is counted, otherwise 0 points are counted. the total score for each patient was calculated by summing points of each risk factor. we also drew the roc curve of the combination for predicting risk of death in patients with severe or critically ill covid-19 in figure 2 . the combined prediction model had the auc of 0.994 (95% ci 0.979-0.999), with specificity/sensitivity of 97.24%/100.00% and positive predictive value (ppv)/ negative predictive value (npv) of 81.0%/100.0% (table 3 ). the cut-off value of it was 0.115. (1) comparison of the above three indicators of patients who died at different time points to further investigate the changes in d-dimer, bun and lymr in the progression of disease in patients who died, we compared table 4 and figure 3 , the results showed no significant differences among the three time points for these indicators. for example, bun was not significantly different in patients who died at the beginning of hospital admission, at the beginning of mechanical ventilation and before death, which means this indicator did not change significantly during the progression of disease in patients who died. since the outbreak of covid-19, many studies have revealed its virological characteristics, transmission characteristics, clinical manifestations, but there were few studies on patients who died. it is now clear that sars-cov-2, a kind of coronavirus, belongs to the β-coronavirus family, and could cause severe coronavirus disease similar to severe acute respiratory syndrome (sars) and middle east respiratory syndrome (mers) [17] . at the same time, based on previous research, the virus can spread rapidly from person to person. its basic reproductive number (r 0 ) is estimated to be 2.2, which means that each patient has transmitted the infection to 2.2 other people on an average [4] . the clinical manifestations of the disease are variable. in previous studies, most patients had mild symptoms such as fever, fatigue and dry cough [8] , with lower overall mortality than sars and mers [11] , but the mortality rate of severe patients was higher than those two diseases [9] . therefore, in this research study, we analysed the laboratory examination indicators of patients who died and hoped to find out the risk factors that could predict the outcome of death. through analysis, we screened three indicators of dd, lymphocyte ratio and bun as predictors of the outcome of death. dd, as a commonly used clinical and simple test, effectively reflects the activation of the coagulation system. in our study, dd higher than 1.56 μg/ml was closely related to fatal outcome of covid-19. it has been proved that high levels of dd are significantly related to 28-day fatality in patients with infection or sepsis detected in the emergency department [18] . the cohort study from jin yin-tan hospital and wuhan pulmonary hospital (wuhan, china) also found that dd greater than 1 μg/ml was the risk factor of death in covid-19 patients. meanwhile, dd is an activation marker of coagulation cascade, which is considered an early event in patients with infection and sepsis [19] . therefore, elevation of dd may indicate severe infection or sepsis, which is often the cause of death in patients with covid-19. as mentioned in previous studies, lymphocytopenia occurred in most severe patients [8, 9] . this laboratory abnormality is similar to it previously observed in patients with sars and mers [20, 21] . consistent with these research studies, our findings showed that lymphocyte ratio lower than 8.615% was highly associated with death of covid-19 patients. therefore, we speculated that the virus may mainly affect lymphocytes, leading to their apoptosis, thereby triggering immune dysfunction in patients, which may be the reason why some patients rapidly develop sepsis and multiple organ failure. some studies have shown that a drastic reduction in the total number of lymphocytes indicates that the coronavirus could consume many immune cells and suppressed the cellular immune function [22] , suggesting that injury of t lymphocytes may be an important factor leading to worsening of the patient's condition. hence, application of immunomodulators may improve infection status in critically ill patients. bun often implies acute kidney injury and is also considered an important predictor of organ failure. our analysis showed that elevated blood urea was closely related to the poor prognosis of patients with covid-19, which was also mentioned briefly in other studies [8] . we suspected that the elevation of bun may be related to acute kidney injury, which may be caused by the invasion of the virus itself, insufficient tissue oxygen supply and shock. in other research studies, older age has been reported as a favourable risk factor of mortality [9, 13] , but through multivariate analysis, our study excluded age as a predictor of death. the reason was that the main objective of previous research was the comparison between severe and non-severe patients, but our study mainly focused on the cases of death in severe patients, in which case the effect of age may be relatively small. we further collected laboratory indexes performed during the hospitalisation of these patients who died and before their death. the timing of mechanical ventilation in these deceased patients was also the time of their worsening condition, so we chose to retrospectively analyse changes in the corresponding indicators at these time points in order to explore the relationship between these indicators and progression of disease. by comparing the above three indicators at different time points, we found that the difference was not significant (table 4 , fig. 3 ). this may mean that relevant indicators of these patients have not changed obviously from the early stages of admission to the clinical outcome, and existing treatment may not significantly slow the deterioration of the disease. it also implies that these deceased patients showed signs of developing a severe outcome at the beginning of hospital admission, indicating the rationality of our prediction using the initial admission data as well as suggesting that patients with this trend could be screened out early and need to be focused on treatment to reduce the covid-19 morbidity rate. to our knowledge, this is the largest number of studies involving severe covid-19 patients, and it is also one of the few studies focusing on patients who died. compared with single-factor predictions in other studies [13] , our designed multifactor scoring system is significantly more accurate. the roc curve also shows that this method has extremely high specificity and sensitivity. at the same time, the three indicators required by this prediction method are easy to obtain at the time of admission and can be implemented in other medical centres. our study also has some limitations. firstly, there is a certain bias due to retrospective research. some indicators that may be meaningful are not routinely tested in the clinic, such as cytokines, lactic acid, transferrin and so on. secondly, the research sample size is still too small. it is still a single-centre study and could not represent the overall situation well. thirdly, the scoring system is established and evaluated using the same group of patients, which makes the evaluation results may be less accurate, and a prospective research cohort needs to be established to further validate its accuracy. we found that three factors including decreased lymphocyte ratio, elevated bun and raised d-dimer were related to death outcomes in critically ill patients with covid-19. a combined multifactorial prediction model with high accuracy was developed to predict the clinical outcome of these patients. world health organization declares global emergency: a review of the 2019 novel coronavirus (covid-19 a novel coronavirus from patients with pneumonia in china sars-cov-2 is an appropriate name for the new coronavirus early transmission dynamics in wuhan, china, of novel coronavirus-infected pneumonia a pneumonia outbreak associated with a new coronavirus of probable bat origin epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study clinical findings in a group of patients infected with the 2019 novel coronavirus (sars-cov-2) outside of wuhan, china: retrospective case series clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in clinical course and outcomes of critically ill patients with sars-cov-2 pneumonia in wuhan, china: a single-centered, retrospective, observational study. the lancet respiratory medicine early estimation of the case fatality rate of covid-19 in mainland china: a data-driven analysis clinical characteristics of coronavirus disease 2019 in china a trial of lopinavir-ritonavir in adults hospitalized with severe covid-19 clinical course and risk factors for mortality of adult inpatients with covid-19 in wuhan, china: a retrospective cohort study analysis of factors associated with disease outcomes in hospitalized patients with 2019 novel coronavirus disease association of radiologic findings with mortality of patients infected with 2019 novel coronavirus in wuhan, china clinical features of patients infected with 2019 novel coronavirus in wuhan understanding of covid-19 based on current evidence evaluation of a novel prognostic score based on thrombosis and inflammation in patients with sepsis: a retrospective cohort study d-dimer is a significant prognostic factor in patients with suspected infection and sepsis multiple organ infection and the pathogenesis of sars middle east respiratory syndrome coronavirus efficiently infects human primary t lymphocytes and activates the extrinsic and intrinsic apoptosis pathways t-cell immunity of sars-cov: implications for vaccine development against mers-cov acknowledgements. the authors thank all participants for their contribution to the study.financial support. this research received no specific grant from any funding agency, commercial or not-for-profit sectors. ethical standards. the study was conducted according to the principles of the declaration of helsinki. the ethics review board of wuhan union hospital approved this study. informed consent for individual patient was not obtained since all data were retrieved retrospectively from the laboratory test information system without additional blood samples or laboratory analysis.data availability statements. the data that support the findings of this study are available from union hospital, tongji medical college, huazhong university of science and technology. restrictions apply to the availability of these data, which were used under licence for this study. key: cord-325453-5eskj42c authors: haider, najmul; yavlinsky, alexei; chang, yu-mei; hasan, mohammad nayeem; benfield, camilla; osman, abdinasir yusuf; uddin, md. jamal; dar, osman; ntoumi, francine; zumla, alimuddin; kock, richard title: the global health security index and joint external evaluation score for health preparedness are not correlated with countries' covid-19 detection response time and mortality outcome date: 2020-09-07 journal: epidemiol infect doi: 10.1017/s0950268820002046 sha: doc_id: 325453 cord_uid: 5eskj42c global health security index (ghsi) and joint external evaluation (jee) are two well-known health security and related capability indices. we hypothesised that countries with higher ghsi or jee scores would have detected their first covid-19 case earlier, and would experience lower mortality outcome compared to countries with lower scores. we evaluated the effectiveness of ghsi and jee in predicting countries' covid-19 detection response times and mortality outcome (deaths/million). we used two different outcomes for the evaluation: (i) detection response time, the duration of time to the first confirmed case detection (from 31st december 2019 to 20th february 2020 when every country's first case was linked to travel from china) and (ii) mortality outcome (deaths/million) until 11th march and 1st july 2020, respectively. we interpreted the detection response time alongside previously published relative risk of the importation of covid-19 cases from china. we performed multiple linear regression and negative binomial regression analysis to evaluate how these indices predicted the actual outcome. the two indices, ghsi and jee were strongly correlated (r = 0.82), indicating a good agreement between them. however, both ghsi (r = 0.31) and jee (r = 0.37) had a poor correlation with countries' covid-19–related mortality outcome. higher risk of importation of covid-19 from china for a given country was negatively correlated with the time taken to detect the first case in that country (adjusted r(2) = 0.63–0.66), while the ghsi and jee had minimal predictive value. in the negative binomial regression model, countries' mortality outcome was strongly predicted by the percentage of the population aged 65 and above (incidence rate ratio (irr): 1.10 (95% confidence interval (ci): 1.01–1.21) while overall ghsi score (irr: 1.01 (95% ci: 0.98–1.01)) and jee (irr: 0.99 (95% ci: 0.96–1.02)) were not significant predictors. ghsi and jee had lower predictive value for detection response time and mortality outcome due to covid-19. we suggest introduction of a population healthiness parameter, to address demographic and comorbidity vulnerabilities, and reappraisal of the ranking system and methods used to obtain the index based on experience gained from this pandemic. on 31st december 2019, the world health organization (who) china country office was informed about a series of pneumonia cases with unknown aetiology in wuhan city, hubei province [1] . by 15th july 2020, the disease, covid-19, caused by infection with sars-cov-2 had infected 13 150 645 people and resulted in 574 464 deaths (4.4% reported case fatality ratio), affecting >200 countries/territories across the world [2] . published mathematical models identified a number of countries in asia, north america, europe and oceania with a higher risk of importation of the sars-cov-2 via infected people arriving from china [3] [4] [5] [6] . on 22nd february 2020, lebanon and israel reported their first covid-19 cases. no epidemiological link to china could be established through contact tracing, which suggested a link to an ongoing outbreak in iran [7] . before these reports, every country's first case had a history of travel to china. the who declared covid-19 as a pandemic on 11th march 2020 [2] . earlier, the who had characterised covid-19 as a public health emergency of international concern on 30th january 2020, their greatest concern being the potential for the virus to spread to countries with weaker health systems [2] . thus, it was important to know how countries with different degrees of preparedness were responding to the pandemic, to inform epidemiological risk, and how resources could be best deployed and control measures applied in support of this global health emergency. some countries were at a higher risk of importation of covid-19 cases because of a higher volume of air passengers and travellers from china and understanding those countries' responses to this pandemic is also important [3, 4, 6] . these questions remain valid during all phases of the pandemic and especially during the process of removal of lockdown measures and opening of air bridges, when once again, risks of further spread increase. this is critical knowledge for what remains essentially a globally susceptible population, with few countries reporting immunity levels above an average of 5% in the community [8] . the global health security index (ghsi) is a comprehensive assessment and benchmarking of health security and related capabilities of 195 countries that make up the states parties to the international health regulations. the ghsi is a project of the nuclear threat initiative and the johns hopkins center for health security, and was developed jointly with the economist intelligence unit [9] . the ghsi provides an index of preparedness based on the capacity gaps of countries in their potential response to a pandemic, such as covid-19, which most countries are ill-positioned to combat [9, 10] . the ghsi comprises six categories, 34 indicators and 85 sub-indicators based on 140 questions. category 2 is the early detection and reporting of epidemics of potential international concern. we chose this category as our measure of the countries' reporting abilities. we further considered the risk of importation of covid-19 from china to different countries based on air-flight passenger data [4] . since most of the case reports prior to 20th february 2020 were linked to cases imported from china, we considered this approach a reasonable estimate of the relative risk of importing new cases into a given country. we also compared overall ghsi to mortality due to covid-19 in the country (deaths per million), referred to as the mortality outcome hereinafter. the mean overall ghsi score is 40.2 out of 100. the high-income countries have an average score of 51.9 [9]. the joint external evaluation (jee) is a voluntary, externally validated, collaborative assessment of 19 technical areas required to validate countries' capacities to prevent, detect and rapidly respond to public health risks [11] . unlike the ghsi which is an academic tool developed to allow inter-country comparisons on pandemic preparedness, the jee is a formal component of the who ihr monitoring and evaluation framework which all un member states are committed to implementing. the jee is not designed for making inter-country comparisons but instead is a tool created to support who member countries in establishing a quantitative baseline assessment of ihr core capacities from which they can then measure their own progress over time. while the intention of jee scoring has never been to draw inter-country comparisons, these have nonetheless occurred as politicians and national governments seek to assess their preparedness capacities against those of their neighbours or regional rivals. ninety-six countries participated in the jee scoring exercise and in this paper we use readyscore, which is the average of 19 technical areas included in jee, as presented by shahpar et al. [12] . this readyscore is calculated using either jee 1.0 or jee 2.0, depending on which assessment the country completed. the objective of this study was to quantify and compare different countries' detection response times and mortality outcomes within specific dates, as related to the covid-19 epidemic. specifically, we evaluated the effectiveness of ghsi and jee in predicting countries' covid-19 detection response times as of 20th february 2020, and mortality outcomes as of 11th march and 1st july 2020, respectively. the global health security index (ghsi) and joint external evaluation (jee) we used the mean value of ghsi category 2 [9] -'early detection and reporting of epidemic of potential international concern'as an indicator of each country's preparedness for an epidemic/pandemic to evaluate a country's response to covid-19. the countries are ranked between 0 and 100 (where 100 is fully prepared and 0 is not prepared at all). the countries with a score of 66.7 and above are categorised as 'most prepared (msp)', 33.4-66.6 as 'more prepared (mrp)' and 0-33.3 as 'least prepared (lep)' [9]. the mean overall ghsi score is 40.2 out of 100 and the usa lead the rank with 83.5 points. the high-income countries have an average score of 51.9. we also used jee's readyscore to evaluate countries' responses to covid-19 pandemic [12] . the score ranged from 0 to 100 and categorised as (i) better prepared (80-100), (ii) work to do (40-79) and (iii) not ready (0-39) [13] . the mean value of readyscore is 54.00 and canada leads the ranking with an score of 93.00 [12, 13] . the details of these indicators are presented in the supplementary material (tables s1-s3). we used these overall ghsi scores to evaluate each country's preparedness and response to the pandemic and specifically to test associations using the detection response time and the related outcome, the deaths per million (death outcome at two different time points). haider et al. described the relative risk of importing covid-19 cases from four major cities of china (wuhan, beijing, shanghai and guangzhou) to 168 countries and territories by considering the simulated air flight passenger data between 1st and 31st january 2020 [4] . the risk incorporated the probability of passengers being infected based on the outbreak sizes in the departure cities. countries were grouped into four quartiles (q1-q4) based on the risk value (43 countries in the top quartile (q4), 42 countries in the third quartile (q3), 41 countries in the second quartile (q2) and 42 countries in the bottom quartile (q1), representing the lowest risk. we used this as a risk index of covid-19 importation since it represented a proxy for the volume of travel of infected passengers from china into those countries and territories. thus, countries with fewer passengers arriving from china had a lower risk of importation of covid-19 cases. najmul haider et al. we collected covid-19 case and date of first report from who's daily situation updates until 20th february [14] as, during that period, most of the case reports outside of china were linked to the cases imported from china. on 22nd february, the who reported confirmed cases in lebanon and israel, both ultimately linked to iran [7, 14] . the time lag between sampling a suspected case and reporting it at a country level has been reported to be up to 24-72 h [15] . restricting our analysis until 20th february 2020 allowed us to investigate the cases linked to importation from china and enabled us to use the importation risk index based on the countries' air transportation links with china alone, as described above. furthermore, we collected covid-19 cases and death data from first reports until 11th march 2020 when who characterised covid-19 as a pandemic and until 1st july 2020 from worldometer [16] . we collected further data mostly from united nation's sources including country's population density [17] , the percentage of the population aged 65 and above [18] , human development index [19] , gross domestic product (gdp) [20] and worldwide governance indicators (wgis) [21] . for each of the preparedness categories, summary statistics are presented for the median [range] of the number of people infected with sars-cov-2 in the corresponding countries, and median [range] of the times from the beginning of the epidemic in china until detecting the corresponding first cases. kaplan-meier curves were used to illustrate the time to first case detection from 31st december 2019. countries without any case detected until 20th february 2020 were censored. log-rank test was used to compare the rate of detection between the ghsi categories, and between the risk quartiles. spearman's correlation was used to assess association between countries' responses to the covid-19 epidemic and the ghsi, as well as the risk of covid-19 importation. multiple linear regression analysis was utilised to assess the amount of variation in either the time to detection of the first case that can be explained by the ghsi and jee score and the importation risk among countries reporting cases by 20th february 2020. since the values of importation risk were skewed, they were log transformed prior to further analysis. we performed multiple linear regression analysis for mortality outcome until 11th march 2020. since the two indices, ghsi and jee were strongly correlated (r = 0.82), we added them separately in the model. finally, we performed negative binomial regression analysis to predict the impact of ghsi and jee and other variables on countries' mortality outcome until 1st july 2020. although negative binomial regression was preferred, there was not enough data for the period until 11th march 2020 to support the model and thus we ran multiple linear regression. we have excluded china from the pre-pandemic (20th february 2020) analysis because we sought to evaluate countries' surveillance systems' effectiveness in detecting and monitoring the evolution of a potential pandemic and china is where the first cases were reported globally. the median [range] value of ghsi score was 40.5 [3.7-98.2]: 22 countries in this study scored higher than the median value. as of 20th february, 26 countries reported covid-19 cases imported from china: 13 were categorised as most prepared, 11 as more prepared and 2 as least prepared countries in the ghsi. with respect to the risk of importation, 21 countries were in the top risk quartile (q4), 4 countries in q3, 1 country in q2 and none in q1. as of 20th february, the countries in the top importation risk quartile (q4) identified the first covid-19 cases at median 28 [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] days after 31st december 2019, earlier than the countries in q3, median 41 [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] days. nepal was the only country in q2 that identified its first covid-19 case, 25 days after the start of the epidemic in china. the msp countries identified the first covid-19 cases at median 26 [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] days after 31st december, earlier than the mrp countries at 30 days and lep countries at 27 [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] days. the correlation coefficient between the risk of covid-19 importation and the time to first reported covid-19 case was −0.61, and the same coefficient between the risk of importation and the number of cases was 0.64. the correlation coefficient between the ghsi and the time to the first reported covid-19 case was −0.32 (fig. 1 ). there were significant differences in the rates of first case detection between ghsi categories and between risk quartiles. around 30% of countries in the highest ghsi score category had their first case detected by the end of january 2020, while 50% of countries in the top risk quartile (q4) had their first detection by this time (fig. 2) . nepal was identified as an outlier in the multiple linear regression analysis. if nepal was excluded from the regression, the risk of importation alone explained the majority of the variation (p < 0.0001; adjusted r 2 : 0.63). however, the ghsi score had minimum impact in both cases. the patterns remained the same if nepal was included in the analysis although the amount of variation explained was lower (adjusted r 2 : 0.33) ( table 1 ). when we used jee's readyscore instead of ghsi, the model's predictive power remained very similar (adjusted r 2 : 0.66) with lower predictive value for jee (coefficient = −0.05, p = 0.02) ( table 2 ). the two indices, ghsi and jee were strongly correlated (r = 0.82) indicating a good agreement between them. however, both ghsi (r = 0.31) and jee (r = 0.37) had a poor correlation with countries' covid-19-related mortality outcome. among the 20 countries with highest mortality outcome, 10 countries also had the higher score in ghsi and five countries had higher score in jee (mean readyscore) ( table s3 in the supplementary material). the variables explaining the countries' mortality outcome are presented in tables 1 and 2 . for the period until the declaration of the pandemic (11th march 2020), the percentage of the population aged 65 and above was weakly positively correlated (coefficient 0.13, p value = 0.06) but the ghsi was not a significant predictor (table 1) . when we used jee in the model instead of ghsi, similar results were observed, with jee remaining nonsignificant and the percentage of the population aged 65 and above as a significant variable (coefficient 0.14, p value = 0.05) ( table 2 ). in the negative binomial regression model, countries' mortality outcome was strongly predicted by the percentage of the epidemiology and infection population aged 65 and above (incidence rate ratio (irr): 1.10 (95% confidence interval (ci): 1.01-1.21) while the overall score of ghsi (irr: 1.01 (95% ci: 0.98-1.01) was not a significant predictor (table 1 ). when we used jee's readyscore in the model instead of ghsi, countries' mortality outcome were strongly predicted by the percentage of the population aged 65 and above (irr: 1.10 (95% ci: 1.03-1.16)), while jee (irr: 0.99 (95% ci: 0.996-1.02) had little impact on the prediction (table 2 ). more than 200 countries/territories have been affected with covid-19 resulting more than 11 million cases and 500 000 death as of 1st july 2020 [16] . we used datasets for three important time periods of the covid-19 pandemic, including (i) 31st december 2019-20th february 2020 when every country's first covid-19 case was linked to travel to china, (ii) 31st december 2019-11th march 2020 when the who declared the covid-19 pandemic and (iii) 31st december 2019-1st july 2020, 6 months after the reporting of the first case in wuhan, china. in these three periods of this epidemic, we evaluated two outcome variables including (i) time taken until first case detection and (ii) mortality outcome due to covid-19 (deaths/ million). however, in none of the models did ghsi or jee status predict any expected outcome with confidence. the first 26 countries reporting sars-cov-2 are an important subset to study in order to understand how countries responded to this emerging disease. until 20th february 2020, most of the covid-19 case importations were linked to international travel from china. our findings suggest that countries with a higher risk of importation of covid-19, based on flight connectivity [4] , detected cases earlier irrespective of their preparedness level, in contrast to the expectation that the higher ghsi scoring countries ought to more rapidly detect the presence of a novel pathogen in the population. china's notification to who and who's press briefing [22] encouraged by the 'international health regulation 2005' probably helped countries to proactively act on the risk [23, 24] . this implies that the risk of importation can be used as a proxy to account for the time it took covid-19 to make an incursion into a particular country, after which most countries detected their first cases at speeds that were not affected by their ghsi or jee preparedness scores. our study further confirms that the health preparedness indices used either in the ghsi or jee had low predictive value in terms of (i) number of cases detected in the country until 20th february 2020 when most cases were imported from china and (ii) mortality outcome (deaths/million) until either 11th march or 1st july 2020. we do not know the exact dates when countries had their first imported cases, however, we inferred the risk of importing cases using direct and indirect air transportation links with china [4] . data on the amount of technical assistance that the who and other institutions provided to less prepared countries are not readily available. however, our finding that countries with lower ghsi preparedness scores had comparable first case detection times to those that were deemed better prepared suggests that in many lower scoring countries, there is a good uptake of relevant guidelines and the local health staff are well-trained to respond to such outbreaks [25] . although the capability of a country to perform large-scale testing will determine its longterm ability to detect covid-19 cases reliably, and rapid first case detection is not the sole measure of success in containing an epidemic/pandemic, the latter can be seen as a good indicator of readiness and a more detailed investigation of the factors that determine its effectiveness is warranted. early case detection and accurate reporting of cases is essential for containing a pandemic so as to limit the spread of the disease both locally and globally. the current spread of covid-19 outside china (as of 15th july 2020) is now driven mostly by community transmission and countries with higher preparedness scores are experiencing larger local outbreaks. for example, the usa has an overall ghsi of 92.1 (ranked 1st) and has identified more than 3.0 million cases as of 1st july 2020, of which the majority are now assumed to be locally acquired [16] . sars-cov-2 can apparently be transmitted without symptoms from a-/pre-symptomatic patients [26, 27] and super-spreading events may occur [28] . countries that fail to detect cases in a timely manner run the risk of creating secondary outbreak foci [27] . iran is one such example, with higher reported numbers of covid-19-related deaths (n = 10 817) as of 1st july 2020. additionally, it appears iran has exported covid-19 cases to at least 12 other countries including bahrain, kuwait and lebanon that had not reported importing cases from china or other countries by the end of february 2020 [7] . a study by tuite et al. estimated that there could be 18 300 (95% ci: 3770-53 470) cases in iran as of 24th february 2020 for it to export cases to those countries [7] while only 43 cases were reported on that day in iran [2] . all of the above highlights the importance of having accurate pandemic preparedness metrics that would help to direct resources to where they are most needed in the event of another major outbreak. however, it seems that both the ghsi and jee indicators do not correlate well with countries' ability to prevent or respond effectively to an epidemic. our findings showed that the countries with lower preparedness narrowed the gap of duration to detection of the first covid-19 case. the countries with lower economic capacity especially those in south east asia and sub-saharan africa are familiar with infectious disease epidemics. these countries, therefore, have extensive experience in early management of infectious diseases epidemics. for example, vietnam, a country that has faced dengue [29] and chikungunya [30] epidemics in the last few years, has successfully controlled the covid-19 epidemic in their territory and started laboratory preparedness prior to the first case being imported and reported in the country [16] . vietnam had not reported any deaths even after 6 months [16] . the country took early measures to test, trace, isolate and for the period until 20 february 2020, multiple linear regression was performed, and the risk of importation of covid-19 from china had higher predictive value than ghsi. for mortality outcome (deaths/million) until 1 july 2020, a negative binomial regression analysis was performed. the percentage of the population aged 65 and above were strongly associated with mortality rate. the incidence rate ratio (irr) of 1.10 of the variable 'the percentage of the population aged 65 and above' indicates that an increase of 1% of population above 65 years of age increases the risk of death rate by 10%. quarantine suspected people and remains the only country of their size of population in the world without any mortality due to covid-19 as of 1st july 2020 [31] . investment in health infrastructure should reduce the risk of infection to a country and reduces overall risk of pandemic spread [32] , but if not activated for political or other socioeconomic reasons, even well-structured capacity is inadequate in the face of a pandemic. collective coordinated and comprehensive approaches that engage the entire machinery of government and international organisations including who should catalyse such preparedness in order to change the trajectory of the covid-19 pandemic. such coordinated efforts remain at the core of global health security. the 10 countries worst affected with covid-19 in terms of deaths per million are among the top 20 countries in terms of their overall ghsi scores. these scores are not correlated with the times taken to detect the first case in each country. similarly, jee's readyscore did not correlate well with detection of the first case in the country. for example, vietnam ranked 50th (score: 49.1) and sri lanka 120th (score 33.9) within ghsi, and both countries responded well with less than one death per million (vietnam reported no deaths). these rankings do not adequately weigh or consider the importance of universal health coverage, integration of national response services across sub-national jurisdictions and the critical importance of effective political leadership during times of crisis and such parameters are likely strongly correlated with time to detection and outcome, perhaps in part explaining the apparent paradox reported here. high income countries with a high ghsi or jee may rely on professionals with competing interests and institutions with varying degrees of authority and responsibility across national and subnational political boundaries to deal with a health crisis rather than taking individual and communal responsibility. thus, the indicators and their respective weighting used in ghsi and the jee might need to be radically revised in future from the lessons learned from the covid-19 pandemic. our study has several limitations. first, the covid-19 cases and death data reported from different countries might not be accurate for the total number of cases and deaths due to underreporting and inadequate testing, although early in the epidemic this was likely to be true in nearly all countries. there is a common case definition for covid-19 globally [33] , however, the testing and diagnostic capacity varies, which could affect both case and death counts. going forward, the challenge will be to reduce such differences in reporting cases and deaths and the rate of detection is being improved generally. second, the global pandemic awareness increases as a result of greater community engagement and this enables the governments to take the necessary steps to control the pandemic including acquiring and building the necessary diagnostic capacity. therefore, the countries likely to import cases later than others had more time to prepare and were on greater alert. this might have affected the case detection speed in some countries. our study has not incorporated this awareness factor in the model but it is unlikely to have been a major confounder especially as poorer countries are less able to disseminate information with lower media coverage and fewer consumers of electronic media. third, the magnitude of the epidemic in china was growing during the study period, so the absolute risk of importation would have likewise grown with time for all of the countries. however, while the size of the epidemic increases the absolute importation risk, this would not change the relative risk values if the transportation links remained unchanged. on the other hand, travel bans were implemented in a number of countries, which would have affected the risk of importation, and which we could not address in our model. nevertheless, given the relatively short time under consideration, we believe these limitations did not alter our findings substantially. these global health preparedness indices (ghsi and jee) are missing variables and inadequately weighting others that could more accurately capture the likely response of countries in such for the period until 20 february 2020, multiple linear regression was performed and the risk of importation of covid-19 from china had higher predictive value than jee. for deaths per million until 1st july 2020, a negative binomial regression analysis was performed. the percentage of the population aged 65 and above were strongly associated with mortality outcome. the incidence rate ratio (irr) of 1.10 of the variable 'the percentage of the population aged 65 and above' indicates that an increase of 1% population above 65 years of age increases the risk of death rate by 10%. 6 najmul haider et al. an emergency as covid-19. as stated earlier, the who was more concerned with the risk of the pandemic to poor countries than rich ones at the beginning which has proven to be a very costly mistake, and the existence of the ghsi in its present form encouraged this position. one such missing variable could be a measure of vulnerabilities and/or dependencies on health systems from an aged population and from co-morbidities. data are showing a high incidence of obesity, diabetes and cancers amongst fatalities, and the possible role of socioeconomic and environmental health risks from air pollution, relative poverty, poor housing and crowding, some of which have shown to be important drivers of the death rate due to covid-19 in all countries [34] . the healthiness of any society including universal access to quality assured health services is another important measure of resilience as the ability to coordinate rapidly a societal response to a pandemic. a measure of trust in the political system or compliance with international health regulations may be important parameters to include in the index, as would be a one health governance indicator showing potential for an integrated approach to prevention and response to emerging infectious diseases. this would devolve responsibility for public health to a wider disciplinary community, increasing the potential to prevent and respond to a pandemic. it would also improve the breadth of scientific advice to government on the one hand while also supporting a system of improved and more accountable governance of the actions arising from such advice on the other. the ghsi and jee indices did not predict well countries' covid-19 detection times and mortality outcome over the period of the study but it will be some months or years before a full assessment can be done. countries with a higher risk of importation detected their first covid-19 cases earlier, irrespective of their preparedness status as measured by the ghsi or jee scores. some limitations are inevitable in these analyses as all of the factors determining the speed of case detections could not be incorporated into our model. in the current covid-19 pandemic emergency, countries with lower preparedness scores appear to have narrowed the gap of the time taken to detect first covid-19 cases when compared to their so-called betterprepared counterparts. long-term investment in health infrastructure for pandemic preparedness is essential but the true test of its efficacy is in a real pandemic. as shown by this study, especially early on in the pandemic of covid-19, social and political factors and vulnerabilities not built into the ghsi and jee evaluation can undermine effective action and these need to be addressed globally. we recommend the ghsi and jee scoring tools be revised to include additional parameters that better estimate countries true pandemic preparedness and vulnerabilities, based on the lessons learned from the covid-19 pandemic. furthermore, the ongoing covid-19 pandemic remains the most recent illustration of how global powers are ill prepared to lead health crises of international concern when their own societies are threatened, and the experience highlights the enduring need for greater political commitment to global health security and for a strengthening of the multilateral system to support a coordinated and effective response. supplementary material. the supplementary material for this article can be found at https://doi.org/10.1017/s0950268820002046. the continuing 2019-ncov epidemic threat of novel coronaviruses to global health -the latest 2019 novel coronavirus outbreak in wuhan potential for global spread of a novel coronavirus from china 2020) passengers' destinations from china: low risk of novel coronavirus (2019-ncov) transmission into africa and south america preliminary assessment of the international spreading risk associated with the 2019 novel coronavirus (2019-ncov) outbreak in wuhan city nowcasting and forecasting the potential domestic and international spread of the 2019-ncov outbreak originating in wuhan, china: a modelling study estimation of coronavirus disease 2019 (covid-19) burden and potential for international dissemination of infection from iran prevalence of sars-cov-2 in spain (ene-covid): a nationwide, population-based seroepidemiological study risk of covid-19 importation to the pacific islands through global air travel world health organization (who) zoonotic diseases action package conference protecting the world from infectious disease threats: now or never who. coronavirus disease 2019 (covid-19) situation how sars-cov-2 tests work and what's next in covid-19 diagnostics reported cases and deaths by country, territory, or conveyance: covid-19 pandemic population density ( people per sq. km of land area) the world bank. total population ages 65 and above human development index ranking the world bank. data: gdp (current us$) the worldwide governance indicators statement on the meeting of the international health regulations (2005) emergency committee regarding the outbreak of novel coronavirus (2019-ncov) notification and other reporting requirements under the ihr sars to novel coronavirus -old lessons and new lessons presymptomatic transmission of sars-cov-2 -singapore response to 'evaluation of modelling study shows limits of covid-19 importing risk simulations in sub-saharan africa' (epidemiology and infection -hyg-le-10513 puzzle of highly pathogenic human coronaviruses (2019-ncov) epidemiological and clinical features of dengue infection in adults in the 2017 outbreak in vietnam special issue: new developments in major vector-borne diseases. part ii: important diseases for veterinarians.). revue scientifique et technique office international des epizooties here are 4 ways viet nam has managed to control covid-19' report by world economic forum context: lessons learned? lancet planetary health who (2020) global surveillance for covid-19 caused by human infection with covid-19 virus: interim guidance covid-19 mortality increases with northerly latitude after adjustment for age suggesting a link with ultraviolet and vitamin d conflict of interest. the authors declare that they have no conflict of interest.ethical standards. this study does not include any individual level data and thus does not require any ethical approval.data availability statement. all the data used in this article are publicly available. the full set of data we collected is available (on request to the corresponding author) for other researchers to assess and use for other research studies. key: cord-323551-22v2hn3v authors: galanti, m.; birger, r.; ud-dean, m.; filip, i.; morita, h.; comito, d.; anthony, s.; freyer, g. a.; ibrahim, s.; lane, b.; matienzo, n.; ligon, c.; rabadan, r.; shittu, a.; tagne, e.; shaman, j. title: rates of asymptomatic respiratory virus infection across age groups date: 2019-04-15 journal: epidemiol infect doi: 10.1017/s0950268819000505 sha: doc_id: 323551 cord_uid: 22v2hn3v respiratory viral infections are a leading cause of disease worldwide. a variety of respiratory viruses produce infections in humans with effects ranging from asymptomatic to life-treathening. standard surveillance systems typically only target severe infections (ed outpatients, hospitalisations, deaths) and fail to track asymptomatic or mild infections. here we performed a large-scale community study across multiple age groups to assess the pathogenicity of 18 respiratory viruses. we enrolled 214 individuals at multiple new york city locations and tested weekly for respiratory viral pathogens, irrespective of symptom status, from fall 2016 to spring 2018. we combined these test results with participant-provided daily records of cold and flu symptoms and used this information to characterise symptom severity by virus and age category. asymptomatic infection rates exceeded 70% for most viruses, excepting influenza and human metapneumovirus, which produced significantly more severe outcomes. symptoms were negatively associated with infection frequency, with children displaying the lowest score among age groups. upper respiratory manifestations were most common for all viruses, whereas systemic effects were less typical. these findings indicate a high burden of asymptomatic respiratory virus infection exists in the general population. respiratory viral infections are a leading cause of disease worldwide, affecting all age groups and representing a serious threat to human health, particularly for infants, older adults and the immunocompromised. the effects of infection on individuals can vary considerably and include completely asymptomatic manifestations, mild upper respiratory effects and severe symptoms requiring hospitalisation. the epidemiology of respiratory viral infections is generally analysed through passive symptom-based surveillance. when studying the burden of infection, many observational studies focus on severe outcomes, such as cardiac and pulmonary complications in hospitalised patients [1] [2] [3] [4] , or the role of respiratory viruses in the exacerbation of pre-existing respiratory conditions [5] [6] [7] . additionally, community-based longitudinal studies have generally been restricted to young children or households and involve the sampling of specimens to identify viruses after an index symptomatic episode occurs [8] [9] [10] . investigation of the prevalence and effects of respiratory viral infections in the broader population, not just among individuals seeking medical attention, is needed to more fully understand the burden of these infections within the community and to develop adequate preventive measures against these pathogens. in particular, the proportion of the population that is infected and yet does not develop symptoms must be determined to better quantify transmission risk, forecast future disease incidence and pathogen spread, and support public health response efforts. here, we document rates of asymptomatic respiratory virus infection through a large-scale community study across multiple age groups. we use data from a cohort of individuals who were tested weekly for respiratory viruses irrespective of symptom status. we combine these test results with participant-provided daily records of cold and flu symptoms and use this information to characterise symptom outcomes by virus and age category. we enrolled 214 individuals from multiple locations in manhattan borough, new york city. we refer to the participants as healthy as they were enrolled from the general population, as opposed to individuals seeking clinical care. the cohort included children attending two daycares, along with their siblings and their parents, teenagers and teachers from a high school, adults working at two emergency departments (a paediatric er and an adult er), and adults working at a university medical centre. the study period spanned 2 years from october 2016 to april 2018 with some individuals enrolled for a single cold and flu season (october-april) and others for the entire period. all individuals from the selected facilities who were willing to participate were enrolled in the study. enrolled individuals were asked to complete a baseline survey and provide two nasopharyngeal swab samples (one from each nostril). following this preliminary step, two nasopharyngeal samples were again collected weekly from each participant irrespective of symptoms. the baseline questionnaire completed at the time of enrolment included information on ethnicity, general health, daily habits, travel history and household structure. for the entire duration of the study, participants provided a daily report rating nine respiratory illness-related symptoms (fever, chills, muscle pain, watery eyes, runny nose, sneezing, sore throat, cough, chest pain), which were recorded on a likert scale (0 = none, 1 = mild, 2 = moderate, 3 = severe). participants were also asked to note if they had sought medical attention, stayed home or taken influenza-like illness (ili)-related medications as a consequence of their listed symptoms. parents provided consent, baseline questionnaire answers and the daily survey information for their enrolled children. a total daily score was generated for each participant by summing the scores of the individual symptoms (total daily score ranges from 0 to 27). details on the participants are summarised in table 1 . two nasopharyngeal samples per participant were collected on a weekly basis using minitip flocked swabs. both samples were stored jointly in 2 ml dna/rna shield (zymo research, irvine, ca, usa) at 4-25°c for up to 30 days and then stored at −80°c in two aliquots. nucleic acids were extracted from 200 µl of sample and 10 µl of internal control using the easymag nuclisens system (biomerieux, durham, nc, usa). samples were then screened for viruses using the esensor xt-8 respiratory viral panel (rvp; genmark dx, carlsbad, ca, usa) [4] , a multiplex pcr assay. the rvp system separately detects influenza a (any subtype, a/h1n1, a/h3n2, a/h1n1pdm2009) and b; rsv a and b; parainfluenza (piv) 1, 2, 3 and 4; human metapneumovirus (hmpv); human rhinovirus (hrv); adenovirus b/e and c; and coronaviruses (cov) 229e, nl63, oc43 and hku1. samples positive for a particular virus were identified by an electrical signal intensity of ⩾2 na/mm 2 (with the exception of cov oc43 for which positive results were identified by an intensity of ⩾25 na/mm 2 , per manufacturer specifications). we classified all specimens, irrespective of result, as symptomatic or asymptomatic according to the individual symptom score in the days surrounding the date of swab collection. we used multiple definitions as a standard for symptomatic infection does not exist. definitions are summarised in table 2 : definitions 1-4 consider a time window of 7 days around the day of the collection, whereas definitions 5-8 use a window of 11 days. while some definitions use raw metrics, definitions 4 and 8 normalise scores by the average symptom score for each participant (average weekly total symptom scores for each participant ranged from 0 to 39). we refer to infections that do not satisfy one or more specified definitions as asymptomatic infections. the association between reporting respiratory symptoms and testing positive was calculated with the χ 2 test. a 'symptomatic week' was defined as a calendar week where the total symptom score was ⩾10. analyses were conducted using the total number of positive samples, as well as the number of illness-events. we defined an illness-event as a group of consecutive weekly swab specimens for a given individual that were positive for the same virus (allowing for a 1-week gap to account for false negatives and temporary low shedding). the effects of different viruses and the severity of symptoms among different age groups were assessed using analysis of variance (anova). the χ 2 statistic was also used to assess pairwise differences. participants were divided into four groups: children (under 10 years of age), teenagers (10-17 years of age), adults with daily contact with children (parents and pediatric er doctors) and adults without daily contact with children. to assign adult participants to the correct category, we used information on household composition derived from the initial questionnaire. in the analysis of symptoms by virus, specimens positive for more than one virus were excluded. to analyse the specific effects of different viruses, we grouped symptoms into upper respiratory symptoms (runny nose, sneezing, sore throat, watery eye), lower respiratory symptoms (cough, chest pain) and systemic symptoms (fever, chills, muscle pain). of the 4215 nasopharyngeal samples collected, 737 (17.5%) were positive for one or multiple respiratory viruses. among the positive results, between 69% and 74% of the samples were classified as asymptomatic depending on the chosen definition (table 2) . overall, 55% of positive specimens were asymptomatic by all definitions. testing positive for one or more respiratory viruses was associated with an increased likelihood of being symptomatic (p < 0.0001); however, for the majority of symptomatic weeks (67%), rvp did not identify the presence of any respiratory virus. there was a weak association between pre-existing respiratory conditions (asthma, allergies) and the likelihood of experiencing symptomatic infections. however, the association was significant only for some definitions, and the effect on symptom score was marginally significant (p = 0.08 anova). coinfections accounted for 10% of positive samples and were found most frequently among children; they occurred throughout the year and were predominantly a combination of hrv and adenovirus. pairwise comparisons between single infections and coinfections across all eight definitions showed that testing positive for multiple viruses was not associated with more severe symptoms. among the viruses considered, influenza and hmpv were associated with significantly higher symptom scores than rsv, hrv, cov, adenovirus and piv. as shown in figure 1 , more than 50% of influenza and hmpv infections were classified as symptomatic by a majority of definitions, whereas the other viruses were mostly asymptomatic according to all definitions (p < 0.01 and p < 0.001 when comparing, respectively, the ratio of symptomatic influenza and hmpv infections to the other viral infections). the analysis of specific symptoms confirmed the higher severity of influenza and hmpv for lower, upper and systemic 2 m. galanti et al. symptoms (fig. 2a) . upper respiratory symptoms were most commonly associated with positive results for all viruses and for all age groups, followed by lower respiratory symptoms and then systemic symptoms ( fig. 2a and b) . infected children showed a similar percentage of upper and lower respiratory symptoms, and teenagers showed a similar percentage of lower respiratory and systemic symptoms. the majority of hmpv infections presented with both upper and lower respiratory symptoms. higher severity of symptoms was not associated with higher frequency of infection. figure 3 shows that while children were most frequently infected with a respiratory virus (they presented with the highest number of viral shedding events per season), they recorded (as reported by their parents) the lowest symptom scores on average. adults without daily contact with children reported the highest symptom score (the difference was significant only between children and adults without contacts with children, p = 0.003). this finding holds when controlling for length of infection, as longer-lasting infections were more frequent in children. host response to respiratory viruses is heterogeneous: some presumed infections, measured by detectable shedding of virus, exhibit no symptoms or signs of disease, whereas others result in more serious symptoms. when assessing the burden of an [11] [12] [13] . in [11] we showed that most individuals, particularly children and their close contacts, contract multiple respiratory infections per year. here, we analysed the symptoms reported by the same infected individuals and characterised them by virus and by age group. the presence of respiratory symptoms was associated with testing positive for one or more respiratory viruses. however, the majority of symptomatic manifestations were not paired with a positive rvp result. the origin of these symptomatic, negative rvp results could be due to allergies, bacterial infections or potential viral infections with pathogens that have not yet been identified or for which the viral panel does not test. one of the main goals of this analysis was to estimate the asymptomatic ratio, that is, the fraction of infections occurring without symptoms. the asymptomatic ratio is an important indicator for constraining both true respiratory virus prevalence within a community and the potential for further disease transmission [14] . asymptomatic carriers can, in fact, contribute to disease spread by generating ( possibly symptomatic) secondary infections. estimates of the asymptomatic ratio vary widely not only across diseases (from 95% of polio virus infections to <10% of measles), but also within the same disease due to different diagnostic testing procedures (pcr vs. serological tests [15] ) and sampling approaches (household vs. longitudinal studies). among respiratory viruses, the role of asymptomatic infection is poorly understood. for influenza alone, the prevalence of asymptomatic infection has been estimated to be as low as 9.4% and as high as 90% depending on the virus type, study, season and definition of asymptomatic infection [15, 16] . there is also some evidence that viral shedding correlates with symptom severity and that the contagiousness of asymptomatic individuals is less than for symptomatic persons [17, 18] . the rationale is that respiratory symptoms (coughing, sneezing, runny nose) help spread pathogens through droplet transmission, either inhaled or settled [19] . however, the absence of symptoms might bring asymptomatic infected individuals into greater contact with susceptible persons outside the home. this is particularly true for infants and toddlers whose behaviour, hygiene habits and close physical contact typically favour the spread of germs, especially in childcare settings. whether greater contact occurs and makes asymptomatic individuals effectively as contagious as symptomatic persons is not known. an additional difficulty is that a standard, accepted definition of symptomatic infection does not exist. further, perception of symptoms is highly subjective, and it may be difficult to assess whether a symptom is caused by the pathogen. for example, chronic symptoms can occurrunny nose is a common sign in children that does not necessary imply viral infectionand allergies can cause sneezing. fever, muscle pain and chills may also be caused by infections other than respiratory viruses. for these reasons, we adopted multiple definitions of 'symptomatic episodes', including personalised metrics. more than half of the rsv, adenovirus, piv, cov and hrv infections documented here were asymptomatic according to all definitions. influenza and hmpv infections caused significantly more symptoms than these other viruses; however, given its high prevalence, hrv was responsible for more than half of symptomatic episodes. in general, all viral infections presented with similar symptoms, with upper respiratory manifestations being more frequent, whereas systemic symptoms such as fever, muscle pain and chills were less typical. for surveillance in which ili (defined as fever plus cough and/or sore throat) is used to record respiratory infections among people seeking care, this constraint may mean that many viral infections go undetected. consistent with previous findings [20, 21] we did not find an association between viral co-infection and the likelihood of being symptomatic or presenting with more severe symptoms. regardless of the definition, our findings underscore the extremely high proportion of respiratory viral infections that are asymptomatic. further analysis is required to capture the role played by asymptomatic individuals in outbreak transmission dynamics, specifically the relative infectiousness of asymptomatic vs. symptomatic infections. there was considerable heterogeneity in individual immune response: some individuals seemed to be consistently less predisposed to developing respiratory symptoms upon viral infection, whereas others were always symptomatic when testing positive. infections in children were less symptomatic than in adults, even though children proved to be more frequently infected with respiratory viruses [11] . frequent infections may enhance the ability of the immune system to identify and respond to pathogens; hence, the group subjected to the fewest respiratory viral infections (adults without children) reported the highest average symptom scores. however, the apparent lower pathogenicity among children may simply be an artefact introduced due to parents reporting symptoms for their children. some symptoms, such as sore throat, muscle pain and chest pain, are difficult to identify in young children, and in fact they were less reported in this age category than among adults and teenagers. this possible bias in the reporting of symptoms is one limitation of our study. the daycare and workplace connections among some individuals in our cohort are another possible bias: some infections were likely caused by the same strain; thus the symptom profiles could be due to specific features of the pathogen rather than individual immune response. further, all the children in our cohort were attending daycare or were siblings of children attending daycare. interactions within the daycare setting could be a confounder in this analysis. future studies should investigate the genetic basis of heterogeneity in host response to respiratory virus infection in order to identify the regulatory pathways controlling reactions to these infections. moreover, longitudinal studies of this type, involving large networks of connected individuals, are needed to assess the role of asymptomatic infections in the transmission of respiratory viruses. the burden of respiratory syncytial virus infection in young children estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory tract infections in 195 countries: a systematic analysis for the global burden of disease study mortality associated with influenza and respiratory syncytial virus in the united states comparison of the biofire filmarray rp, genmark esensor rvp, luminex xtag rvpv1, and luminex xtag rvp fast multiplex assays for detection of respiratory viruses role of viral respiratory infections in asthma and asthma exacerbations respiratory viruses and exacerbations of asthma in adults role of viruses in exacerbations of chronic obstructive pulmonary disease molecular epidemiology of respiratory syncytial virus transmission in childcare viral shedding and clinical illness in naturally acquired influenza virus infections epidemiology of viral respiratory tract infections in a prospective cohort of infants and toddlers attending daycare longitudinal active sampling for respiratory viral infections across age groups asymptomatic summertime shedding of respiratory viruses asymptomatic shedding of respiratory virus among an ambulatory population across seasons estimating pathogen-specific asymptomatic ratios the fraction of influenza virus infections that are asymptomatic: a systematic review and meta-analysis household transmission of the 2009 pandemic a/h1n1 influenza virus: elevated laboratory-confirmed secondary attack rates and evidence of asymptomatic infections does influenza transmission occur from asymptomatic infection or prior to symptom onset? world health organization writing group (2006) nonpharmaceutical interventions for pandemic influenza, international measures dynamics of infectious disease transmission by inhalable respiratory droplets epidemiology of multiple respiratory viruses in childcare attendees infection with multiple viruses is not associated with increased disease severity in children with bronchiolitis author orcids.m. galanti, 0000-0002-9060-1250.financial support. this work was supported by the defense advanced research projects agency contract w911nf-16-2-0035. the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.conflict of interest. js and columbia university disclose partial ownership of sk analytics. all other authors declare no competing interests. key: cord-349262-gnqbyc6t authors: hemida, maged gomaa; ali, mohammed; alhammadi, mohammed; alnaeem, abdelmohsen title: the middle east respiratory syndrome coronavirus in the breath of some infected dromedary camels (camelus dromedarius) date: 2020-10-14 journal: epidemiol infect doi: 10.1017/s0950268820002459 sha: doc_id: 349262 cord_uid: gnqbyc6t dromedary camels remain the currently identified reservoir for the middle east respiratory syndrome coronavirus (mers-cov). the virus is released in the secretions of the infected camels, especially the nasal tract. the virus shedding curve through the nasal secretions was studied. although human transmission of the virus through the respiratory tract of close contact people with dromedary reported previously, the exact mechanism of transmission is still largely unknown. the main goal of this study was to check the possibility of mers-cov shedding in the exhaled air of the infected camels. to achieve this goal, we conducted a follow-up study in one of the dromedary camel herds, december 2018–april 2019. we tested nasal swabs, breath samples from animals within this herd by the real-time pcr. our results showed that some of the tested nasal swabs and breath were positive from 24 march 2019 until 7 april 2019. the phylogenetic analysis of the obtained s and n gene sequences revealed the detected viruses are clustering together with some human and camel samples from the eastern region, especially from al-hufuf city, as well as some samples from qatar and jordon. these results are clearly showing the possibility of shedding of the virus in the breath of the infected camels. this could explain, at least in part, the mechanism of transmission of mers-cov from animals to humans. this study is confirming the shedding of mers-cov in the exhaled air of the infected camels. further studies are needed for a better understanding of the mers-cov. it has been almost 8 years since the emergence of the middle east respiratory syndrome coronavirus (mers-cov) in 2012 in the arabian peninsula. the majority of human cases were reported in the arabian peninsula [1] [2] [3] . more recently, there are many reports about the potential risk of infection with mers-cov in africa [1] . the only known reservoir so far is the dromedary camels [4] . there are many reports about the animal-human transmission in the context of mers-cov [5, 6] . it is now well documented that mers-cov is released in the respiratory secretions detected in the nasal swabs by several research groups [5, [7] [8] [9] [10] [11] . although no report on the secretion of the virus in the milk, it is highly recommended to boil it before consumption to avoid any potential contamination of the raw milk during the process of milking and handling [12] [13] [14] . similar concept for eating the meats and raw organs, particularly the liver, it should be cooked well before consumption to avoid any potential hazards of infection as suggested by the world health organization (who) [12, 14] . the virus was also detected in the air around some positive camel herd. detection of the mers-cov-rnas of identical sequences from the owner and the herders of this particular herd was also reported [5, 6] . however, some other studies failed to detect the virus in the urine of some infected animals [15] . earlier studies succeeded in the detection of the mers-cov nucleic acids in the air circulated by an infected camel herd as well as in the people of close contact of this particular herd [5, 6] . meanwhile, several reports showed the possibility of detection of the mers-cov nucleic acids in the circulating air in some health care settings during the korean outbreak of mers-cov in 2015 [16, 17] . the main goal of the current study was to check the possibility of mers-cov shedding in the infected animal breath. we conducted this study according to the guidelines of the animal ethics protocols and the national committee of bio-ethics, king abdul-aziz city of science and technology, royal decree no. m/59 (http://www.kfsh.med.sa/kfsh_website/users uploadedfiles%5cncbe%20regulations%20english.pdf). the animal experiments were approved by the animal care committee of the deanship of scientific research, king faisal university, saudi arabia. we conducted molecular surveillance for the mers-cov in a dromedary camel herd during 2019. this herd consists of 52 animals held together in wire fenced yards. the male animals were kept in a separate individual partition while the female was kept together in multiple partitions. there is a shared open yard where all females may be allowed to be mixed together. the males only approach females during the mating season from november to march per each year. during the time from 10 march 2019 until 7 april 2019, we randomly selected nine animals out of the 52 (18%) from the camel herd to conduct our molecular surveillance. all animals were of variable ages ranging from 2 to 8 years old. the nine animals selected from different colour coat-based breeds, including magaheem, wodoah and sofr, were selected. these nine animals were mixed with the rest of the animals in the herd all the time. our molecular surveillance lasts from 10 march to 8 april 2019, with a biweekly sampling interval. we collected nasal swabs, breath samples from these nine animals, as previously described. a paired nasal and breath samples were collected per each animal at each time point, as described in detail below. nasal swabs were collected from at least 15% of the dromedary camel herd from november until march 2019. swabs were transferred into the tubes containing viral transport media including the dmem media containing 10% foetal bovine sera and antibiotic cocktail (penicillin and streptomycin). the processing of the collected swabs was done as previously described. briefly, the swabs were vortexed, then centrifuged at 5000 rpm for 10 min in a cooling centrifuge at 4°c. the supernatants were collected and stored at −80°c for further testing. the breath samples were collected in sterile tubes containing viral transport media as used for the collection of the nasal swabs. the breath collection technique was done in a simple way. briefly, the animals were secured in a specially designed camel crush. the animals were allowed to settle down for 5 min. with the help of an assistant, the nasal orifice is dilated, and the tube containing the media is slightly inserted inside (2-3 cm) the orifice without touching the skin or the mucous membranes of these animals during sampling. the collection tubes were held in a slope position facing the air stream coming from the animal during the exhalation. the tubes were kept in such position until the animal completes at least five cycles of breathing, including five exhalation streams. any container that touched the skin or the mucous membrane of the animals was discarded and not included in our analysis. the collection tubes were shaken immediately for 3 min and inverted upside down several times, then placed into ice tanks until transferred to the laboratory. the viral rnas were extracted from paired swabs and breath samples collected from dromedary camels using the qiagen viral rna (rneasy mini kit, qiagen, hilden, germany) extraction kit, according to the manufacturer's protocol. the concentration of the extracted viral rnas was measured by the nanodrop machine (thermo scientific nanodrop 2000, applied biosystems, 850 lincoln centre drive, foster city, ca 94404, usa). the eluted rnas were stored at −80°c for further testing. the rt-pcr targeting the upstream gene (up-e) of mers-cov was used for screening [13] . confirmation was done using the open reading frame (orf) 1a. five μl of the extracted rna was subjected to the real-time pcr testing using upe primers using lightmix molecular dx mers-cov-upe kits (roche, roche molecular systems, inc, pleasanton, ca 94588, usa) according to the manufacturer's protocol. all positive samples by the upe assay were confirmed by orf1a, as previously described [7] . positive samples were considered when there is an overlapping between the results from two targets. samples (nasal swabs and breath) that had ct values <39 were considered positive per each target. we used several sets of primers to amplify the partial mers-cov-s, mers-cov-n genes. the sequences of these primers are as follow: nf-5 ′ -cct tcg gta cag tgg agc ca-3 ′ and nr-5 ′ -gat ggg gtt gcc aaa cac aaa c-3 ′ , primers for the mers-cov-s gene sf-5 ′ -ccaattta-cgccaggatgat-3 ′ and sr-5 ′ -aatagaggcgg aaatagcac-3. we used the primers listed above to amplify the partial mers-cov-s and n genes using the qiagen one-step rt-pcr kit (cat no./id: 210210) to synthesis the cdna then doing the pcr in one reaction. the procedure of this one-step reaction was carried out as per the kit's instructions as well as previously described. we ran 5 μl per each amplified mers-n and s amplicons on a 1% agarose gel containing sybr® safe dna gel stain (invitrogen, thermo fisher scientific, waltham, ma, usa). the amplicons were visualised under ultraviolet light, then the gel pictures were photographed with the gel documentation system (bio-rad laboratories, inc., hercules, ca, usa). we purified the desired pcr products from the target bands by using the gel-based pcr extraction kits (qiagen, cat no/id: 28704), as per the instructions of the kits. we eluted the purified pcr products in 50 μl elution buffer as suggested by the kits, then stored at −20°c. we sequenced some positive samples from dromedary camels having low ct values (≤29) on the real-time pcr. the sequencing was done by the sanger method using the applied biosystems® we used the obtained sequences from dromedary camels to develop the phylogenetic tree based on the reported partial mser-cov-s and n genes sequences. the phylogenetic trees were built using the neighbour-joining method by the mega-7 software, as previously described. the scale bar represents the tree distance corresponding to 0.6 nucleotide substitution/nucleotides. a series of two by two tables, utilizing the fisher's exact test, were used to test the association between the results of the molecular surveillance, the sampling intervals and the sampling technique [18] . all the statistical data analyses were performed by spss version 21.0 (ibm corp, 2012). the values (<0.05) were considered significant (fig. 1) . our molecular testing for the selected animals at three time points starting 10 march 2019, until 7 april revealed that 1/9 (11%) of the tested nasal swabs were positive; all the nine animals were negative from the breath samples (table 1) . two weeks later, on 24 march, 6/9 (77%) nasal swabs were positive, and 5/9 (55%) breath samples were positive (table 1) . finally, on 7 april, 3/9 nasal swabs (33%) were positive, while 1/9 (11%) of the breath samples were positive ( table 1 ). the prevalence of positive nasal and breath samples was higher on the second sampling batch (24 march) compared to first (10 march) and third (7 april) batches. however, the difference was statistically significant (p < 0.05) only between the first vs. second sampling batches. similarly, consolidated results obtained from both nasal swaps and breath were significantly higher in the second batch vs. the first batch (p < 0.05). no significant statistical differences were observed between the prevalence of positive results obtained from nasal swabs and breath samples. the detection of the mers-cov-rnas in the breath of all three time intervals was statistically significant (p > 0.05) ( table 1) . the sequencing analysis of the partial mer-cov-s and n genes (figs 2 and 3 , respectively) from this positive mers-cov herd showing the detected viruses were closely related to the sequence from dromedary camels and humans from the arabian peninsula. the reported partial mers-cov-s gene clustered together with the other human sequences reported from al-hufuf city in the same regions in 2015 as well as sequences from qatar and jordon-2015 (fig. 2 ). the mers-cov continues to pose a significant risk for humans, especially some of those who may come in close contact with dromedary camels [12, 13] . despite the emergence of sars-cov-2 late 2019 [19] , mers-cov still has the high case fatality rates among the affected populations compared to sars-cov and sars-cov-2. the mers-cov case fatality rate started at 52% in 2012 then dropped down to 32% in 2019 [20] . the presence of mers-cov in the air of the proximity of patients and infected dromedary camels was previously documented independently by many research groups [6, 16, 17] . this result confirms that mers-cov may be transmitted through droplet infection. however, little is still known about the mechanism of transmission of mers-cov from the dromedary camels to humans. the roles of some camel secretions and excretions in the context of the mers-cov infection have been studied [5, 13, 15] . our results show that about 11% of the tested nasal swabs were positive during the first batch of the collection, while none of the breath samples was positive ( table 1 ). the reasons behind the inability to detect the viral rnas in the breath during the first round of sampling may be attributed to the dilution of the viral rna in the breath, and the rna may be present at the nondetectable limit by the real-time pcr. a high consistency was observed between the detection of the viral nucleic acids in both the nasal and breath samples in all three tested batches (p > 0.05), indicating similar opportunities to detect the virus nucleic acid in both methods (table 1 ). this may be attributed to several factors. first, during the early stage of the virus infection, the rna concentration in the breath may be at an undetectable level by the real-time pcr. second, the virus replication reached the peak of each animal; therefore, high virus concentration was achieved as previously reported [7, 8] . these results showed that mers-cov could be shed in the breath of infected animals for a while. however, the nasal swabs are still the sample of choice in the diagnosis of mers-cov among the infected dromedary camel population. spreading of the virus from animal to another in the same herd could potentially happen through the respiratory routes taking into consideration the very close contact between animals within the same herd. it may also occur through sharing the contaminated food and water from infected animals; however, all these aspects still need further investigations. detection of the virus in the air of positive camel's herd [5, 6] may suggest the virus is excreted in the breath of the infected animals in high concentration. however, this hypothesis was never tested before. detection of the foot and mouth diseases virus (fmdv) in the breath of some infected cattle was previously documented [21] . this study confirmed the potential spread of the fmdv between animals and among various herds in close proximity or within a distance. the aim of our study was to test the possibility of mers-cov shedding in the breath of the infected dromedary camels. our results are clearly showing the potential secretion of mers-cov in the breath of infected animals ( table 1) . the phylogenetic analysis based on the partial mers-cov-s and n genes in these infected animals revealed high identity to other mers-cov previously detected in the arabian peninsula (figs 2 and 3) [7, 8, 11] . a recent study showed that the sars-cov-2 could be transmitted through the droplet infection in the air [22] . although detection of the mers-cov-rnas in the breath does not conclude the viability of the detected virus particles. this may require virus isolation and the plaque assay to ensure the virus infectivity in the collected samples. these techniques should be done at a biosafety contaminant-3 laboratory. however, the overlapping between the viral detection in the nasal swabs and breath, as one of the gold standard techniques for the detection of the virus as well as the high degree of identity of the reported sequences in this study, may provide a piece of strong evidence about the potential shedding of the mers-cov in the breath of infected animals at various levels. further large-scale studies are highly recommended to document the curve of the mers-cov shedding in various body secretions and execrations as well as in the breath. this will lead to a better understanding of the dynamics of the virus spread among certain dromedary camel populations as well as in the environment. taken together all these evidence, we may conclude that mers-cov could be transmitted through the breath of infected animals. the virus spread from animal to animal and from animal to human come in their close contact (fig. 4) . however, large-scale controlled studies are still required to further enrich our understandings about the mechanism of mers-cov transmission between animals as well as from animals to humans. middle east respiratory syndrome coronavirus (mers-cov) neutralising antibodies in a high-risk human population comparative serological study for the prevalence of anti-mers coronavirus antibodies in high-and low-risk groups in qatar middle east respiratory syndrome coronavirus during pregnancy middle east respiratory syndrome (mers) coronavirus seroprevalence in domestic livestock in saudi arabia evidence for camel-to-human transmission of mers coronavirus detection of the middle east respiratory syndrome coronavirus genome in an air sample originating from a camel barn owned by an infected patient longitudinal study of middle east respiratory syndrome coronavirus infection in dromedary camel herds in saudi arabia mers coronavirus in dromedary camel herd, saudi arabia middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels in africa and middle east genetic evidence of middle east respiratory syndrome coronavirus (mers-cov) and widespread seroprevalence among camels in kenya prevalence of middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels in abu dhabi emirate mers-cov at the animal-human interface: inputs on exposure pathways from an expert-opinion elicitation middle east respiratory syndrome coronavirus and the one health concept dromedary camels and the transmission of middle east respiratory syndrome coronavirus (mers-cov) failure to detect mers-cov rna in urine of naturally infected dromedary camels risk of global spread of middle east respiratory syndrome coronavirus (mers-cov) via the air transport network airflow as a possible transmission route of middle east respiratory syndrome at an initial outbreak hospital in korea handbook of biological statistics a novel coronavirus from patients with pneumonia in china some one health based control strategies for the middle east respiratory syndrome coronavirus detection of foot-and-mouth disease virus in the breath of infected cattle using a hand-held device to collect aerosols airborne transmission of sars-cov-2: the world should face the reality acknowledgements. we wish to thank the king abdul-aziz city for science and technology (kacst), saudi arabia, for their generous funding through the mers-cov research grant programme (number 20-0004/24-1), which is a part of the targeted research program (trp). data availability statement. data are available upon request. key: cord-355171-oi3ezlsl authors: macintyre, c. r.; seale, h.; yang, p.; zhang, y.; shi, w.; almatroudi, a.; moa, a.; wang, x.; li, x.; pang, x.; wang, q. title: quantifying the risk of respiratory infection in healthcare workers performing high-risk procedures date: 2013-12-05 journal: epidemiol infect doi: 10.1017/s095026881300304x sha: doc_id: 355171 cord_uid: oi3ezlsl this study determined the risk of respiratory infection associated with high-risk procedures (hrps) performed by healthcare workers (hcws) in high-risk settings. we prospectively studied 481 hospital hcws in china, documented risk factors for infection, including performing hrps, measured new infections, and analysed whether hrps predicted infection. infection outcomes were clinical respiratory infection (cri), laboratory-confirmed viral or bacterial infection, and an influenza infection. about 12% (56/481) of the study participants performed at least one hrp, the most common being airway suctioning (7·7%, 37/481). hcws who performed a hrp were at significantly higher risk of developing cri and laboratory-confirmed infection [adjusted relative risk 2·9, 95% confidence interval (ci) 1·42–5·87 and 2·9, 95% ci 1·37–6·22, respectively]. performing a hrp resulted in a threefold increase in the risk of respiratory infections. this is the first time the risk has been prospectively quantified in hcws, providing data to inform occupational health and safety policies. healthcare workers (hcws) are at increased risk of healthcare-associated infections due to the front-line nature of their work. transmission of highly infectious diseases from infected patient to other patients and hcws occurs constantly in hospitals and healthcare centres and has been well documented [1] [2] [3] . although hcws are aware of infection control measures, low levels of compliance with standard precautions by this group are frequent [4, 5] . hcws are less willing to adhere to infection-control practices when they work for extended hours [6] , with probable reasons for low compliance being insufficient time, scarcity of equipment, lack of knowledge and low perception of risk [5] . the three principal routes of transmission of respiratory pathogens are contact transmission (direct and indirect), droplet transmission, and airborne transmission. for any pathogen, more than one transmission route may occur, but many pathogens are known to be transmitted by one predominant mode. in droplet transmission, pathogens or droplets which are larger than 5 μm, such as influenza virus and bordetella pertussis are transmitted from an infected patient to hcws through breathing, talking, coughing, sneezing, as well as through performing high-risk procedures (hrps) [2, 7, 8] . however, influenza virus has also been documented to be transmitted by the airborne route, which results in infectious particles being present in the air for longer periods of time [9] [10] [11] [12] . respiratory infectious diseases, even those with limited airborne transmission, are more likely to be transmitted from patients to hcws during hrps such as suctioning and intubation which generate respiratory aerosols [13] . many studies suggest that both invasive and non-invasive procedures are likely to increase the probability of hcws being infected [13, 14] . some studies have reported that non-invasive positive pressure ventilation (nppv) can be a risk of severe acute respiratory syndrome (sars) transmission to hcws [15] [16] [17] . cardiopulmonary resuscitation, manual ventilation, bronchoscopy and suctioning have also been documented to increase the risk of hcws being infected with sars and tuberculosis (tb) [18] [19] [20] [21] , while tracheal intubation has been significantly associated with risk of sars transmission to hcws [17] . while it has been well documented that tb and sars can be transmitted to hcws during aerosol-generating procedures, there are some data suggesting h1n1 can transmit via such procedures [13] . seasonal influenza also causes outbreaks in healthcare settings [22] . hcws are one of the most vulnerable groups likely to be infected with influenza infection in acute-care facilities due to the high exposure rates in such settings [23] . an attack rate of nosocomial influenza could reach 11-59% in hcws in a healthcare environment [24] . as such, hcws are a priority group for preventive strategies such as influenza vaccination [25, 26] . although various guidelines and policies for infection control measures are implemented in healthcare settings worldwide, the risk of transmission of infectious diseases while performing hrps has not been well quantified. this study aims to describe the range of exposure to hrps in hcws and to quantify the risk of respiratory infections occurring in hcws who perform hrps. we prospectively studied 481 hospital hcws from wards including emergency and respiratory wards from nine hospitals in beijing, china over a 5-week period from 1 december 2008 to 15 january 2009. these 481 subjects were a control group in a larger study [27] . the hospital wards were selected as high-risk settings in which repeated and multiple exposures to respiratory infections are expected. participants were hospital hcws aged 518 years and who were provided with written information about the study. staff who agreed to participate provided informed consent and a copy of the information sheet with the participants' initials was retained as documentation [27] . the study protocol was approved by the institutional review board and human research ethics committee of the beijing ministry for health. participants were asked to record on a daily basis whether they had performed one of the following: provision of nebulizer medications, suctioning, intubation, aerosol-generating procedures and chest physiotherapy. the following information was also collected: number of hours worked, estimated number of daily contacts with patients, number of daily contacts with influenza-like illness (ili) patients, and hand-washing adherence. the use of personal protective equipment such as gloves, gowns, eye shields, foot/hair covers was documented by study participants in a daily selfreport diary, and details of clinical and demographics were also recorded. all study participants were followed up for a period of 31 days and monitored daily for the onset of respiratory symptoms. if any symptom developed, combined nasal and throat swabs (double rayon-tipped, plastic-shafted swab) were taken and tested for respiratory viral or bacterial infection (fig. 1) . the nose and throat swabs were tested at the laboratories of the beijing centres for disease control and prevention. viral dna⁄rna was extracted from 300μl of each respiratory specimen using the viral gene-spin ™ kit (intron biotechnology inc., korea) according to the manufacturer's instructions [27] . we tested nose and throat swabs for the following: adenoviruses, human metapneumovirus (hmp), coronaviruses 229e/nl63 and oc43/hku1, parainfluenza viruses 1, 2 and 3, influenza viruses a and b, respiratory syncytial virus (rsv) a and b, and rhinovirus a/b by nucleic acid testing using a commercial multiplex polymerase chain reaction (pcr), with the seeplex® rv12 detection kit (seegen inc., korea). details of laboratory methods have been described in a previous publication [27] . we also tested for bacterial colonization. a multiplex pcr (seegen inc.) was used to detect streptococcus pneumoniae, mycoplasma pneumoniae, b. pertussis, legionella spp, chlamydophilia and haemophilus influenza type b. after preheating at 95°c for 15 min, 40 amplification cycles were performed under the following conditions in a thermal cycler (geneamp pcr system 9700, applied biosystems, usa): 94°c for 30 s, 60°c for 1·5 min, and 72°c for 1·5 min. amplification was completed at the final extension step at 72°c for 10 min. the multiplex pcr products were visualized by electrophoresis on an ethidium bromide-stained 2% agarose gel. the controls represent hcws in their usual working conditions, without any interventions. this study is a post-hoc analysis of data collected during the primary trial on hrps in the control arm. the prospective data collection and measurement of clinical endpoints in a group of hcws working under usual conditions afforded the opportunity to measure the association of incident infection with hrps. the primary outcomes of the study were: clinical respiratory infection (cri)presence of two or more respiratory symptoms or one respiratory symptom (e.g. cough, runny nose, shortness of breath, sore throat) and one systemic symptom (e.g. fever, lethargy, chills); laboratory-confirmed viral infection (influenza a and b, parainfluenza, rsv, coronavirus, hmp virus, adenovirus, rhinovirus); laboratoryconfirmed viral or bacterial infection (and of the above viruses or a bacterial infectionpertussis, hib, pneumococcus, mycoplasma, legionella); and influenza a or b (categorized as 'influenza' if either strain were present). the outcomes were tested against predictor variables such as age, education, category of hcw, influenza vaccination uptake, and performance of hand washing and hrps. the total number of hrps performed over the study period was calculated. a binary variable defining whether or not hcws performed any hrps during the study period was created and analysed with other predictor variables against incident infection during the study period. poisson regression was used for the analysis of the outcomes, using egret software (cytel, usa). a p value of 40·05 was considered significant in the analysis. a total of 481 hcws were recruited into the study. demographic characteristics of study participants are described in table 1 . of these, 369 (76·7%) were females; and 52% (252/481) of the participants were aged 535 years. the breakdown of participants by area was: respiratory ward 75 (16%); emergency department 72 (15%); respiratory clinic 16 (3·3%); paediatric department 15 (3·1%); infection fever clinic 6 (1·2%); and other wards 297 (62%). of the 481 hcws, 236 (49%) were doctors and 245 (51%) were nurses and others. during the study period, the uptake of influenza vaccine in hcws was low in both 2007 and 2008 (19·3% and 18·1%, respectively). fifty-six (11·6%) out of 481 hcws performed at least one hrp during the study, with the most common activity being airway suctioning (66%, 37/56). figure 2 shows the number of days on which a hcw reported performing a hrp. thirty-four (61%) out of 56 hcws, reported performing a hrp more than once during the study period. the aggregated number of days a hrp was performed was 264. in addition, hcws on the respiratory ward (33%) were more likely to perform hrps than those in the emergency department (16·7%). nurses and other hcws (16%, 39/245) were significantly more likely than doctors (7%, 17/236) to perform hrps (p < 0·01). the weekly incidence of cri was 18/1000 hcws, for viral or bacterial infection, 16/1000; for any viral infection, 6/1000; and for influenza, 2·5/1000. hcws who performed hrps had a significantly higher risk of cri [relative risk (rr) 2·5, 95% (ci) 1·3-6·5, p < 0·01) and laboratory-confirmed viral or bacterial infection (rr 2·6, 95% ci 1·4-5, p < 0·01) than those who did not perform hrps (table 2) . by poisson regression analysis, adjusting for other variables, only hrps determined the risk of an infection outcome, as shown in tables 3-5. the relative risk for cri in hcws who performed a hrp was 2·9 (95% ci 1·42-5·87, p < 0·01) ( table 3 ). the rr for a laboratory-confirmed pathogen (viral or bacterial) in symptomatic hcws was 2·9 (95% ci 1·37-6·22, p = 0·01), in those who performed a hrp (table 4 ). for the outcome of any respiratory viral pathogen, the rr was 3·3 (95% ci 1·01-11·02, p = 0·05) ( table 5 ). hand washing, influenza vaccination and use of surgical or cloth face masks did not affect the risk of infection outcomes. we also tested for other variables, e.g. number of hours worked, number of patients the hcw was in contact with during the study period, and number of contacts with patients with ili; however, none of these had a significant association with infection outcomes. there were no significant association between laboratory-confirmed influenza and hrps (data not shown); but the numbers of influenza-positive cases were low (six cases) in the study. we examined the association between hrp and the risk of respiratory infection in hcws. our findings demonstrated that hcws who perform a hrp have a greater risk of respiratory infections than those who did not perform a hrp. this is consistent with observational findings of other studies [15, 16, 18, 21, 28] ; however, we have been able to quantify the magnitude of this risk in our study as a threefold increase in risk. many factors influence the nosocomial spread of infectious diseases, and hcws are the initial point of contact with patients in both acute and long-term healthcare settings. our findings suggest that targeted interventions and policies are warranted to offer greater protection to hcws who perform hrps. this has occupational health and safety implications for hcws routinely engaged in hrps. more than 10% of hcws performed a hrp during a 1-month period, and the majority of those performed more than one hrp. this suggests that interventions to reduce transmission of respiratory infections may be more efficient if targeted to hcws performing hrps. there may be certain settings such as emergency wards, intensive-care units and respiratory wards where hrps are more commonly performed, making these important targets for interventions. there are some limitations to our study. our study was conducted in china, so the results, particularly around frequency of performing hrps, may not be generalizable to different hcw populations in other contexts. there are variations in infection control practices from hospital to hospital, even within china. however, the quantification of risk for hcws who perform hrps has implications for hcws everywhere. the fact that this was a control group in a larger trial is a strength, rather than a limitation, in that this group had rigorous follow-up and documentation of incident infection as well as risk factors (including hrps). this provides more robust data than, for example, an observational study such as a case-control study, because it was prospective and measured infection in a group that went about usual practice. to date, there is much policy debate and direction about hrps, but no data whatsoever to inform the actual risk associated with hrps. despite the limitations of the analysis, we believe that the data we present in this paper are a useful addition to current knowledge. hcws are at higher risks of contracting respiratory infections, and are subject to generic guidelines around infection control. these include hand hygiene before and after the patient care; wearing of personal protective equipment such as gowns, goggles, gloves, n95 respirators or surgical masks; presence of minimum number of hcws when performing a procedure in a single room; and in addition, it is recommended that such procedures should be performed in a sterilized room [13, 29, 30] . we have shown in two large randomized controlled trials that the risk of respiratory infection in hcws can be reduced with the use of n95 respirators [27, 31] . we also show that in highrisk wards, targeted use in situations of self-identified risk, such as when performing hrps or barrier nursing a patient, is less effective than continuous use of a respirator in that ward while on shift [31] . this suggests that hcws are unable to identify all situations of risk when left to decide whether or not they should wear a respirator. this is the first time the risk of hcws performing hrps has been prospectively quantified, and this finding has important occupational health and safety implications for hcws, particularly in settings such as emergency and respiratory wards where hrps are frequently performed. the traditional approach to hospital infection control has not consistently categorized staff in terms of whether they perform hrps in order to apply guidelines. we found that the majority (89%) of hcws do not perform hrps. this proportion may vary in different country, hospital and ward settings; our study suggests that categorizing hcws by whether or not they perform hrps in their work may serve as a useful classification in order to tailor guidelines appropriately or increase the attention to adherence with existing guidelines. the minority of hcws performing hrps should receive optimal respiratory protection, and high-risk wards should have guidelines in place to minimize the risk to hcws. transmission of bacterial infections to healthcare workers during intubation and respiratory care of patients with severe pneumonia the 2003 sars outbreak and its impact on infection control practices healthcare workers and health care-associated infections: knowledge, attitudes, and behavior in emergency departments in italy practices and an assessment of health care workers' perceptions of compliance with infection control knowledge of nosocomial infections compliance with universal precautions among health care workers at three regional hospitals transmission of influenza virus via aerosols and fomites in the guinea pig model guideline for isolation precautions: preventing transmission of infectious agents in health care settings high humidity leads to loss of infectious influenza virus from simulated coughs detection of infectious influenza virus in cough aerosols generated in a simulated patient examination room airborne transmission of influenza: implications for control in healthcare and community settings influenza virus in human exhaled breath: an observational study interim guidance on infection control measures for 2009 h1n1 influenza in healthcare settings, including protection of healthcare personnel should noninvasive ventilation be considered a high-risk procedure during an epidemic? transmission of severe acute respiratory syndrome during intubation and mechanical ventilation illness in intensive care staff after brief exposure to severe acute respiratory syndrome aerosol generating procedures and risk of transmission of acute respiratory infections to healthcare workers: a systematic review the transmission of tuberculosis in confined spaces: an analytical review of alternative epidemiological models possible sars coronavirus transmission during cardiopulmonary resuscitation a practical approach to airway management in patients with sars occupational tuberculous infections among pulmonary physicians in training influenza in the acute hospital setting requiring influenza vaccination for health care workers: seven truths we must accept vaccination against classical influenza in health-care workers: self-protection and patient protection acip recommendations: introduction and biology of influenza, 2010-11 influenza prevention & control recommendations world health organisation a cluster randomized clinical trial comparing fit-tested and non-fit-tested n95 respirators to medical masks to prevent respiratory virus infection in health care workers which preventive measures might protect health care workers from sars? infection prevention and control during health care for confirmed, probable, or suspected cases of pandemic (h1n1) 2009 virus infection and influenza-like illnesses epidemic and pandemic prone acute respiratory diseases -infection prevention and control in health care a randomised clinical trial of three options for n95 respirators and medical masks in health workers dr seale is in receipt of an nhmrc australian-based public health training fellowship (1 012 631) . professor macintyre receives funding from influenza vaccine manufacturers gsk and csl biotherapies for investigator-driven research. dr seale has received funding from sanofi pasteur, gsk and csl biotherapies for investigator-driven research and for conference presentations. key: cord-347916-9suvf3ln authors: kong, man; zhang, hongmei; cao, xiaocui; mao, xiaoli; lu, zhongxin title: higher level of neutrophil-to-lymphocyte is associated with severe covid-19 date: 2020-07-09 journal: epidemiol infect doi: 10.1017/s0950268820001557 sha: doc_id: 347916 cord_uid: 9suvf3ln in december 2019, cases of severe coronavirus 2019 (covid-19) infection rapidly progressed to acute respiratory failure. this study aims to assess the association between the neutrophil-to-lymphocyte ratio (nlr) and the incidence of severe covid-19 infection. a retrospective cohort study was conducted on 210 patients with covid-19 infection who were admitted to the central hospital of wuhan from 27 january 2020 to 9 march 2020. peripheral blood samples were collected and examined for lymphocyte subsets by flow cytometry. associations between tertiles of nlr and the incidence of severe illness were analysed by logistic regression. of the 210 patients with covid-19, 87 were diagnosed as severe cases. the mean nlr of the severe group was higher than that of the mild group (6.6 vs. 3.3, p < 0.001). the highest tertile of nlr (5.1–19.7) exhibited a 5.9-fold (95% ci 1.3–28.5) increased incidence of severity relative to that of the lowest tertile (0.6–2.5) after adjustments for age, diabetes, hypertension and other confounders. the number of t cells significantly decreased in the severe group (0.5 vs. 0.9, p < 0.001). covid-19 might mainly act on lymphocytes, particularly t lymphocytes. nlr was identified as an early risk factor for severe covid-19 illness. patients with increased nlr should be admitted to an isolation ward with respiratory monitoring and supportive care. in december 2019, an ongoing outbreak of unexplained pneumonia in wuhan drew attention globally [1] . deep gene sequencing confirmed that the patients with pneumonia were infected with a novel β-coronavirus, which was identified as a severe acute respiratory syndrome coronavirus 2 (sars-cov-2) by the international committee on taxonomy of viruses. the pneumonia caused by sars-cov-2 was referred to as coronavirus disease 2019 (covid-19) by the world health organization on 30 january 2020 [2] . as of 9 march 2020, 80 735 confirmed cases, including 5111 severe cases and 3119 deaths, were reported in china. most patients with covid-19 infection reportedly experienced mild cold symptoms, including fever, cough and fatigue. although computed tomography (ct) of the chest indicated increased infection after 3-5 days, the prognosis was good [3] . in patients with severe infection, the disease developed rapidly into acute respiratory distress syndrome, coagulopathy, septic shock and even death [4, 5] . for mild cases, symptomatic treatment and general isolation were provided at centralised isolation points or non-critical designated hospitals. however, for severe cases, efforts were directed towards preventing the development of critical diseases; risk factors were identified as early as possible; appropriate supportive care was provided; and transfer to intensive special hospital was coordinated if necessary to reduce mortality. rational use of medical resources was also implemented. a study indicated that large amounts of pro-inflammatory cytokines in serum were associated with pulmonary inflammation and extensive lung damage in covid-19, similar to severe acute respiratory syndrome (sars) and middle east respiratory syndrome coronavirus (mers-cov) infection [6] . the current study aimed to investigate the association between different laboratory data (including lymphocyte subsets and inflammatory biomarkers) and clinical characteristics of hospitalised patients with mild and severe covid-19 infection to reveal a potentially useful prognostic factor associated with severe morbidity. this study was a retrospective single-centre study among patients treated at the central hospital of wuhan, a specific hospital for the treatment of patients with covid-19. lymphocyte subset analysis was conducted on the patients who were included in the final analysis and admitted from 27 january 2020 to 9 march 2020. the clinical diagnosis of viral pneumonia was initially based on their clinical symptoms, including fever, cough or respiratory illness and typical changes in chest ct. all patients in this study lived in wuhan during the outbreak period of covid-19. this study was approved by the ethics review committee of the central hospital of wuhan. all patients regularly signed informed consent when they were admitted to wuhan union hospital, china. the clinical classification of covid-19 was in accordance with the diagnosis and treatment protocol for covid-19 (trial version 7) issued by the national health committee of the people's republic of china (http://www.nhc.gov.cn/). the condition was considered as severe-type covid-19 when one of the following criteria was present: (1) respiratory distress with respiratory rate >30/min; (2) oxygen saturation ≤93% in the resting state; or (3) arterial blood oxygen partial pressure (pao 2 )/oxygen concentration (fio 2 )≤300 mmhg (1 mmhg = 0.133 kpa). if follow-up of these cases revealed progression into cases requiring admission to intensive care unit (icu), the prognosis was considered very poor. throat swab samples were collected from all patients for covid-19 viral nucleic acid detection via real-time reverse transcription-polymerase chain reaction assay. this test was performed in a clinical laboratory at the central hospital of wuhan in accordance with the chinese center for disease control and prevention protocol. the data collected included demographic data, symptoms, chronic disease, medical history, outcome events from the electronic medical record system and findings from the laboratory information system. laboratory results included blood routine, coagulation routine, biochemical indicators, infectionrelated biomarkers and lymphocyte subsets. the percentages of the lymphocyte subsets were analysed using the bd facscanto flow cytometer. the laboratory data of some patients were missing because of delayed results or the absence of test types. the data that support the findings of this study are available from the central hospital of wuhan. restrictions apply to the availability of these data, which were used under licence for this study. descriptive statistics were presented as means ± s.d. for continuous variables and as percentages for categorical variables. student's t-test was used to compare normally distributed continuous variables, whereas the mann-whitney u-test was applied for data not obeying a normal distribution, which was represented by the median (lower quartile, upper quartile). the nonparametric chi-squared test was used to compare categorical variables. a linear regression model was used to analyse the correlation between neutrophil-to-lymphocyte ratio (nlr) levels and the incidence of severe illness. multivariate logistic regression was used in the three models to obtain the odds ratios (ors) and the corresponding 95% confidence intervals (95% cis) between the tertiles of nlr and the incidence of severe cases. ors were not adjusted for any variable in the crude model. model 1 was adjusted for age, c-reactive protein, interleukin-6, procalcitonin, diabetes and hypertension. subsequently, lactic dehydrogenase (ldh), white blood count (wbc), d-dimer, cd4+ t cells and cd8+ t cells were further adjusted. the lowest tertile of nlr was used as the reference under each model to calculate the ors and the corresponding 95% cis. all data were analysed using spss version 21 (spss inc., chicago, il). p < 0.05 was regarded as statistically significant in all statistical analyses. table 1 lists the demographic and clinical characteristics of covid-19 infection. among the 210 patients diagnosed with covid-19, 87 (41.4%) were categorised into the severe group and 123 (58.6%) into the mild group upon admission. in the severe group, 38 (43.7%) progressed into icu or even death; in the mild group, only 1 (0.8%) developed into a critical case. compared with patients in the mild group, patients in the severe group were older (67.9 ± 12.3 vs. 53.2 ± 15.6, p = 0.005), particularly those older than 70 years (39.1%). no significant difference in the proportion of women was found between the severe group (50.5%) and the mild group. a total of 72 (82.8%) severe cases and 82 (66.7%) mild cases had a fever, and a significant difference in body temperature was determined between the two groups. compared with mild cases, severe cases were more likely to experience mild shortness of breath (p < 0.001), myalgia (p = 0.009), fatigue (p = 0.011) and chest congestion (p < 0.001). meanwhile, 99 (47.1%) patients had at least one underlying comorbidity, and the proportion of severe cases with underlying comorbidity was higher than that of mild cases (65.5% vs. 34.1%, p < 0.001). comparison between the severe cases and the mild cases indicated significant differences in diabetes (p = 0.004), hypertension (p < 0.001), chronic renal disease (p = 0.024) and tumour (p = 0.007), but not in cardiovascular disease, hyperlipidaemia, tuberculosis and chronic obstructive pulmonary disorder (copd). the results of the peripheral blood test of the patients on the day of hospital admission are listed in table 2 . absolute wbcs and neutrophil counts were significantly higher in the severe group than in the mild group (6.1 vs. 5.3, p = 0.002 and 4.3 vs. 3.1, p < 0.001, respectively), whereas the absolute lymphocyte count was significantly lower in the severe group than in the mild group (0.8 vs. 1.2, p < 0.001). no significant differences in monocyte, haemoglobin and platelet count were found. with regard to the coagulation of all patients, activated partial thrombin time (aptt) and d-dimer were significantly elevated in the severe group (29.9 vs. 28.9, p = 0.005 and 0.7 vs. 0.4, p < 0.001, respectively). compared with the mild group, most patients in the severe group showed higher levels of infection-related indicators, such as procalcitonin (0.07 vs. 0.05, p < 0.001), c-reactive protein (3.2 vs. 0.6, p < 0.001), interleukin-6 (9.4 vs. 3.2, p < 0.001) and erythrocyte sedimentation rate (52.0 vs. 34.0, p < 0.001), indicating that inflammation was more prominent in severe cases. other significant abnormal findings related to blood biochemistry included gamma-glutamyl transferase (ggt), hydroxybutyrate dehydrogenase (hbdh), lactic dehydrogenase (ldh) and blood urea nitrogen (bun), which could be a direct influence of the covid-19 virus or an indirect influence of hypoxia. man kong et al. the lymphocyte subsets in all cases were then analysed. the total b cells, nk cells and t cells were significantly lower in the severe group than in the mild group (0.7 vs. 1.2 p < 0.001); specifically, the cd3 cell count was lower in the severe group than in the mild group (0.4 vs. 0.7 p < 0.001). no significant differences in b cell count and nk cell count were found between the two groups. the different subsets of t cells were further analysed. both the suppressor t cells (cd3 + cd8+) (0.2 vs. 0.3 p < 0.001) and helper t cells (cd3 + cd4+) (0.2 vs. 0.5 p < 0.001) were significantly lower in the severe group than in the mild group. linear regression analysis was used to assess the relationship between nlr and the incidence of severe illness in all participants ( table 3) . nlr was positively correlated with the incidence of severe illness (β = 0.056, p = 0.000). the potential of the aforementioned parameters as potential prognostic predictors of case severity was evaluated. nlr was significantly higher in the severe group than in the mild group (6.6 vs. 3.3 p < 0.001) ( table 2) . thus, nlr was selected as a potential predictive factor and further analysed. multivariable logistic regression was used to analyse the correlation between nlr levels and the progress of severe cases (table 4 ). nlr was divided into tertiles. the man kong et al. lowest nlr level was used as the reference, and the categorical variables were analysed. the unadjusted logistic regression analysis indicated that upper categorical levels had a statistically significant higher risk of severity than that of lower levels: or 95% ci 4.1 (1.7-10.2) and or 95% ci 10.1 (4.0-25.6), respectively. after adjustments for age, c-reactive protein, interleukin-6, procalcitonin, diabetes and hypertension, the highest tertile remained statistically significant, and the relative risk was or 7.8, 95% ci 2.3-26.4. further adjustments for other factors, including ldh, wbc, d-dimer, cd4+ t cells and cd8+ t cells showed that the highest level of relative risk (or 5.9, 95% ci 1.3-28.5) remained significant. this study included 210 patients infected with covid-19. their clinical characteristics and laboratory findings were analysed. compared with the mild group, the severe group was older, mostly had a high fever and had at least one underlying disorder. these clinical features were mostly similar to those in previous studies [1, 7] . this finding suggested that owing to their weakened immune function, older than younger patients with chronic diseases were more likely to be infected with covid-19. early identification of risk factors for severe patients is vital to afford appropriate supportive care or access to icu if necessary. severe cases presented lower lymphocyte counts and higher neutrophil levels. the severe group also showed elevated biomarkers for infection. as a widely used factor for systemic infection and inflammation, nlr was used to assess the severity of bacterial infection and the clinical prognosis of pneumonia [8] [9] [10] . a study (n = 61) recently reported that nlr was the most useful factor affecting the incidence of severe covid-19 and indicated the incidence of severe illness with nlr ≥ 3.13 [11] . in this study, the patients in the highest nlr tertile presented a 5.9-fold increased risk of incidence of severe covid-19 after adjustments for potential confounders were applied. qin et al. indicated that nlr increased in several patients with covid-19 infection, and surveillance of nlr helped in the diagnosis and treatment of covid-19 infection in the early stages, which was consistent with our findings [12] . an elevated nlr was also found in an early cohort (n = 138) with higher neutrophil counts and marked lymphopenia in severe cases [7] . therefore, nlr could be a useful factor in reflecting the degree of imbalance between inflammatory and immune responses in patients with covid-19. the biological mechanism underlying this association has yet to be determined, and several plausible explanations exist. one of the most convincing explanations is based primarily on the physiological link between neutrophilia and lymphopenia with systemic inflammation and stress. another explanation is that neutrophils are the important cellular components of the host defenses in the innate immune system, whereas lymphocytes are considered as the major cells involved in adaptive immunity [9] . lymphocytes play a key role in the regulation of inflammatory response, and sustained reduction in severe cases is associated with the non-resolution of inflammation [13] . the mechanism underlying this regulation requires further research. a well-coordinated innate immune response is known to be the first line of defense against viral infections. however, when the first line of defense is dysregulated, excessive inflammatory cell infiltration, inflammatory storm and even death may occur [14] . previous studies on sars-cov and mers-cov showed that t cells, particularly cd4+ and cd8+ t cells, played a crucial role in inhibiting or weakening overactive innate immune responses during viral infection [14] [15] [16] . cd4+ t cells coordinate the deletion and amplification of immune cells to regulate immune responses. cd4+ t cells facilitate virus-specific antibody production via the t-dependent activation of b cells [17] . moreover, cd8+ t cells primarily exert their effects via two mechanismscytokine secretion and cytolytic activities against target cells [18] . secretion of cytokines such as ifn-γ is essential to resist viral and bacterial infections [19] . zhao et al. indicated that cd8+ t cells played a crucial role in viral clearance and immune-mediated injury in most infiltrative inflammatory cells in the pulmonary interstitium [20] . a comparison between b celldeficient mice and t cell-deficient mice showed that t cells in lungs infected with mers-cov survived and killed virus-infected cells [21] . these data emphasise the importance of t lymphocytes rather than b cells in controlling the pathogenesis and outcomes of sars-cov and mers-cov infection. in the current study, t lymphocytes decreased more in the group with severe covid-19, similar to sars-cov and mers-cov. covid-19 may attack t cells and destroy our immune system, leading to serious infection. the severity of pathological damage during sars-cov and mers-cov was related to the extensive infiltration of pulmonary neutrophils and the increase in neutrophils in peripheral blood [22] . therefore, covid-19 could mainly act on lymphocytes, particularly t lymphocytes. on the basis of the current study, patients with covid-19 who are suffering from pneumonia and those with increased nlr should be admitted to an isolation ward with respiratory monitoring and supportive care rather putting them into centralised isolation. this finding should largely reduce the progression of critical illness caused by untimely treatment to reduce mortality. our study has several advantages. one is the previously determined biological plausibility of a strong association between nlr and the risk of incidence of severe cases. the second is the analysis, which eliminates several potential confounding variables to avoid bias. this study also has some limitations: it was a smallsized, single-centre and retrospective study (a larger cohort would be better to eliminate potential bias), and for some patients, repeated measurement data were provided on the first day. the data are presented as n (%) or or (95% ci). model 1 was adjusted for age, c-reactive protein, interleukin-6, procalcitonin, diabetes and hypertension. model 2 was further adjusted for ldh, wbc, d-dimer, cd4+ t cells, cd8+ t cells. covid-19, coronavirus disease 2019; nlr, neutrophil-to-lymphocyte ratio. first data point was always used, resulting in potentially incomplete information on variations in intraday cell count. covid-19 could mainly affect lymphocytes, especially t lymphocytes. nlr was an early risk factor affecting the prognosis of patients with severe covid-19 illness. patients with a higher nlr should be admitted to an isolation ward with respiratory monitoring and supportive care. clinical features of patients infected with 2019 novel coronavirus in wuhan world health organization (2020) a public health emergency of international concern over the global outbreak of novel coronavirus declared by who a familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study clinical characteristics of 140 patients infected with sars-cov-2 in wuhan, china. allergy. online ahead of print clinical characteristics of coronavirus disease 2019 in china clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan, china inflammation biomarkers in blood as mortality predictors in community-acquired pneumonia admitted patients: importance of comparison with neutrophil count percentage or neutrophillymphocyte ratio prognostic significance of neutrophil-to-lymphocyte ratio in patients with sepsis: a prospective observational study the role of neutrophil to lymphocyte count ratio in the differential diagnosis of pulmonary tuberculosis and bacterial community-acquired pneumonia: a cross-sectional study at ayder and mekelle hospitals neutrophil-to-lymphocyte ratio predicts severe illness patients with 2019 novel coronavirus in the early stage dysregulation of immune response in patients with covid-19 in wuhan, china. clinical infectious diseases. online ahead of print failure to normalize lymphopenia following trauma is associated with increased mortality, independent of the leukocytosis pattern age-dependent dysregulation of innate immunity clinical and immunological features of severe and moderate coronavirus disease 2019 clinical course and outcomes of critically ill patients with middle east respiratory syndrome coronavirus infection immunological responses against sars-coronavirus infection in humans cd4+ t cells orchestrate both amplification and deletion of cd8+ t cells p47 gtpases: regulators of immunity to intracellular pathogens pathology and pathogenesis of severe acute respiratory syndrome rapid generation of a mouse model for middle east respiratory syndrome pathogenic human coronavirus infections: causes and consequences of cytokine storm and immunopathology acknowledgements. we thank the men who have fought bravely against the virus on the front line during the covid-19 epidemic; some of them have even lost their lives. we also thank the clinics that allowed us access to their databases.financial support. none.conflict of interest. the authors declare that they have no conflicts of interest. key: cord-346673-kyc1wks5 authors: nickbakhsh, s.; thorburn, f.; von wissmann, b.; mcmenamin, j.; gunson, r. n.; murcia, p. r. title: extensive multiplex pcr diagnostics reveal new insights into the epidemiology of viral respiratory infections date: 2016-03-02 journal: epidemiol infect doi: 10.1017/s0950268816000339 sha: doc_id: 346673 cord_uid: kyc1wks5 viral respiratory infections continue to pose a major global healthcare burden. at the community level, the co-circulation of respiratory viruses is common and yet studies generally focus on single aetiologies. we conducted the first comprehensive epidemiological analysis to encompass all major respiratory viruses in a single population. using extensive multiplex pcr diagnostic data generated by the largest nhs board in scotland, we analysed 44230 patient episodes of respiratory illness that were simultaneously tested for 11 virus groups between 2005 and 2013, spanning the 2009 influenza a pandemic. we measured viral infection prevalence, described co-infections, and identified factors independently associated with viral infection using multivariable logistic regression. our study provides baseline measures and reveals new insights that will direct future research into the epidemiological consequences of virus co-circulation. in particular, our study shows that (i) human coronavirus infections are more common during influenza seasons and in co-infections than previously recognized, (ii) factors associated with co-infection differ from those associated with viral infection overall, (iii) virus prevalence has increased over time especially in infants aged <1 year, and (iv) viral infection risk is greater in the post-2009 pandemic era, likely reflecting a widespread change in the viral population that warrants further investigation. acute respiratory infections are the commonest cause of illness in all ages, and a leading cause of mortality in children aged <5 years, creating a significant global healthcare burden [1] [2] [3] . various aetiological pathogens (viruses, bacteria and some fungi) are recognized, causing largely indistinguishable symptoms. in most settings, viruses are the most frequently detected agent [4, 5] . although most infections are mild, respiratory viruses have the potential to cause severe illness in high-risk groups. although influenza is a major research focus [6] , the advent of polymerase chain reaction (pcr) technology has led to improved awareness that non-influenza viruses are also important contributors to disease burden, and of the role of viral subtype in clinical severity [7] [8] [9] . the use of pcr testing as part of routine diagnostics provides an important resource for monitoring respiratory viruses as part of national surveillance [10] . multiplex pcr methods in particular provide a valuable resource for epidemiological enquiry [11] . all patients requiring microbiological diagnosis are tested for all pathogens included in the panel, ensuring consistency in testing across patients. the collation of multiplex diagnostic data from a large patient population and over an extended time-frame therefore enables robust comparisons of infection trends temporally and across patient subgroups. furthermore, when testing is implemented over multiple years, sufficient data can be accrued to investigate the clinical relevance of co-infections and their epidemiological patterns [12] . although the utility of diagnostic data in the epidemiology of respiratory infections has been demonstrated [11, [13] [14] [15] [16] , studies that cover all major viruses, patient age and illness severity groups, and that span multiple years, are lacking. the largest nhs health board in scotland, greater glasgow and clyde (nhsggc), has used multiplex pcr testing as part of their routine diagnostic services since 2005. this health board serves ∼1·1 million people, representing ∼1·7% of the total uk population [17] . the resultant accumulation of data provides a novel opportunity to investigate viral respiratory infections in a more comprehensive fashion than previously possible. these data also provide a unique opportunity to compare the periods before and after the introduction of the novel pandemic influenza virus [a(h1n1) pdm09] into scotland (see [18] ). we analysed diagnostic data generated by nhsggc using multiplex pcr from 2005 to 2013 with the following objectives: (i) to describe testing and virus prevalence trends, (ii) to examine temporal and patient subgroup distributions for each individual virus, and (iii) to compare factors associated with overall viral infection and co-infection using statistical modelling, in order to provide robust and timely estimates of who is most at risk of viral-associated respiratory illness, and when, within a major urban uk population. in this study we used virological diagnostic data generated by the west of scotland specialist virology centre (wossvc) for nhsggc during 2005-2013 [19] . during this period, a total of 61 427 clinical samples were received from 40 962 patients attending primary and secondary healthcare services for respiratory diagnostic purposes (i.e. excluding pathology-origin samples). most (98%) clinical samples were taken from the upper or lower respiratory tract: primarily nasal and/or throat swabs (67%), gargles (13%), nasopharyngeal aspirates (7%), sputum (5%), bronchoalveolar lavage (3%) and nasopharyngeal/ endotracheal secretions (2%). in a minority of cases (n = 142 samples), plasma was additionally taken for follow-up investigation; most (89%) of these samples related to the 2009 influenza a pandemic period which was excluded from statistical modelling analyses. each sample was tested by real-time rt-pcr for 11 groups of respiratory viruses: human rhinovirus (rv); influenza a virus [iav; a generic assay detecting seasonal h3n2 and h1n1 subtypes and one specific to a(h1n1)pdm09], influenza b virus (ibv), human respiratory syncytial virus (rsv), human coronavirus (cov; aggregating 229e, nl63, hku1 and oc43 species), adenovirus (adv), human metapneumovirus (mpv) and human parainfluenza types 1-4 (piv1-4). details of nucleic acid extraction methods and the real-time pcr assays are provided elsewhere [20] . complete testing coverage across viruses was largely maintained throughout the study period. however, high frequencies of partial testing did arise due to the burden placed on laboratory resources during the major waves of a(h1n1)pdm09 virus circulation. the laboratory protocols were consistent throughout the study period, with the exception of the rv assay which was modified during 2009 to detect a wider array of rv and enteroviruses (including d68), and the cov-hku1 assay which was discontinued in 2012. for each of the 61 427 clinical samples, positive/negative pcr test results were recorded by the laboratory for each virus group. information was also provided on the sampling date, patient's age at sampling, gender, and the origin of the sample [whether the patient had attended a general practice (gp), hospital outpatient or non-critical-care inpatient services, or was admitted to a critical care ward]. in the case of inconclusive/absent test results or other patient information, the corresponding data were coded as missing. all patient identifiers were anonymized. of the 40 962 patients, 8394 had multiple samples submitted for virological testing during the study period (range 1-37 samples, median 1, s.d. = 1·22). for 70% of these patients, the samples were received within a 30-day window. we aggregated the pcr test results to within this time-frame generating single 'episodes' of respiratory illness, using the collection date of the first sample when assigning temporal information. episodes were classified as positive for a given virus if at least one sample tested positive. following data exclusions, 44 230 patient episodes, representing 36 157 individual patients, were retained for analysis of temporal distributions. we conducted descriptive statistical analyses of viral infection prevalence in the patient population providing time-and agestratified estimates. by the end of april 2009, scotland was afflicted by the influenza pandemic [20] . figure 1a highlights the resultant upsurge in testing frequencies during the summer and autumn waves of 2009, and during a third wave of a(h1n1)pdm09 virus circulation in the winter of 2010/2011. during these periods, testing was primarily directed towards iav and only subsets of iav-negative patients were tested for other viruses. due to this disruption in regular testing procedures, we focused our description of viral infection distributions across patient subgroups on the 26 974 patient episodes tested outside this period, and refer readers to a previous report for details of viruses detected during the 2009 pandemic [20] . for each virus group, we compared the frequency of mono-infection episodes (one virus group detected) and co-infection episodes (more than one virus group detected). to correctly classify episodes into these subgroups, we excluded all partially tested patients. in more detailed analyses, we counted the frequency of each possible virus pair and quantified the statistical correlation between mono-infection and co-infection frequencies across viruses. we investigated statistical associations between time period, season, patient age, gender, and gp/general hospital/critical care origin (a proxy for illness severity), and two outcomes: (i) virus-positive vs. virusnegative episodes, and (ii) co-infection vs. monoinfection episodes. with respect to time, we split sampling dates into two major periods either side of the influenza pandemic and periods of high partial testing: pre-pandemic [prior to may 2009 when the a(h1n1)pdm09 virus was established in scotland] and post-pandemic [following subsidence of the third major wave of the a(h1n1)pdm09 virus in january 2011]. associations with each factor were first assessed by crude unadjusted odds ratios, and then adjusted for confounding using multivariable logistic regression models that included all factors to assess their independence. statistical interactions were examined using mantel-haenszel stratification methods (based on p < 0·05, results not shown). the potential interactions were added to the main effects models and their significance assessed based on an interaction parameter p < 0·05. model fit was assessed by le cessie van houwelingen global goodness-of-fit tests [21] . all statistical analyses were carried out in r v. 3·1·1 [22] . to correctly classify patients into outcome groups, all partially tested patients were excluded. of the 36 157 fully tested patients, 90% sought healthcare facilities once during the study period thereby contributing a single episode. however, 4218 patients had attended healthcare facilities more than once, providing information for multiple episodes (range 2-26 episodes, median 2, s.d. = 2·04). we retained the first observed episode per patient in the statistical analyses to ensure the patient-level interpretation of statistical associations was not influenced by the nonindependence of data relating to the same individual. see supplementary figure s1 for full details of data preparation. we analysed 44 230 episodes of respiratory illness tested by wossvc during 2005 to 2013. full details of patient distributions across subgroups and per study year are provided in supplementary table s1. patients' median age was 27 years (range 0-98 years, s.d. = 25·5 years) and 49% were male. excluding the three major waves of influenza a(h1n1)pdm09 virus circulation, episode frequencies increased year-by-year from 2472 cases tested in 2005 to 6149 cases tested in 2013. however, the age patterns were not consistent over this period; the percentage of adult episodes was greater in 2013 than in 2005 (e.g. 21% vs. 8% in patients aged 565 years), while the percentage of child episodes was fewer in 2013 than 2005 (e.g. 16% vs. 26% in patients aged 1-5 years) ( fig. 1b) . at least one virus was detected in 35% (15 302/44 230) of tested patients; these patients had a median age of 17 years (range 0-96 years, s.d. = 25 years) and 49% were male. the prevalence of confirmed viral infection in the patient population was greater in the 2013 influenza season than in 2005 in all age groups (fig. 1c) ; the absolute difference in prevalences was 22% (infants aged <1 year), 12% (1-5 years), 14% (6-16 years), 18% (17-45 years), 12% (46-64 years) and 17% (565 years). overall virus-specific prevalences in the patient population were ranked as follows: rv (14%, n = 4847); iav (9·7%, n = 4244); rsv (4·9%, n = 1786); cov (4·1%, n = 1339); adv (3·6%, n = 1221); ibv (3%, n = 1019); mpv (2·6%, n = 345); piv-3 (2·2%, n = 757); piv-4 (0·86%, n = 286); piv-1 (0·84%, n = 295) and piv-2 (0·35%, n = 122). age distributions for each viral infection group are presented in supplementary table s2 . the most common infection in each 6-month period (excluding 2009) was rv, constituting a low of 19% of infections during the typical influenza period of 2005/2006, to a high of 59% during the typical non-influenza period of 2010 (fig. 1d) . for most virus groups, detections were most frequent in the 1-5 years group (with the exception of iav, ibv and cov), males, and hospital attendees not admitted to a critical care ward (fig. 2) . seasonally, virus detections were most common in december (45% in gp attendees, 43% in hospital attendees) and least common in august (11% in gp attendees, 22% in hospital attendees) (fig. 3a,c) . the most commonly detected viral infection in each month was rv, peaking in september in both gp and hospital attendees (fig. 3b,d) . influenza a and b were the most common detections in december-march in gp attendees (combined proportion: range 31-45%), and in january-february in hospital attendees (combined proportion of 30%). of the remaining non-influenza viral infections, a large proportion was attributed to rsv, rv and cov during periods of high influenza activity; their combined proportions ranged from 39% to 52% in gp attendees (december-march) and from 51% to 55% in hospital attendees (january-february). of 9094 positive patients (in 26 974 patients outside of the pandemic period), 1952 were gp attendees, 6560 were general hospital attendees (outpatients and non-critical-care inpatients), and 1282 were inpatients admitted to a critical care ward [an intensive care unit (icu), intensive therapy unit (itu), high dependency unit (hdu), or coronary care unit (ccu)]. the latter group provided a proxy for classifying episodes of severe respiratory illness. eighty-eight percent (n = 4443) of gp attendees and 69% (n = 15 027) of hospital attendees were aged >5 years. as shown in figure 4 , the prevalence of severe episodes in all viruspositive patients, regardless of origin, was greater in patients with rv (7·5%), rsv (7·5%), piv1 (11·8%) and piv4 (7·4%) infections than in virus-negative patients or other viral infections including iav (5·5%) and ibv (4·1%). investigating further the rv/ iav and rv/piv1 comparisons, we found the observed difference in prevalence was statistically significant based on pearson's χ 2 tests (p = 0·036 and p = 0·05, respectively). age-specific prevalence of severe episodes was greatest at the extremes of age (<5 and 565 years) for all viruses except hpiv2 (we note the particularly small sample size for this virus group). of 9654 virus-positive patients from 27 284 episodes tested for all 11 viruses, 11% (1086/9654) had a co-infection. the median age in co-infected patients was 3 years (range 0-91 years, s.d. = 22 years) and 58% were male. co-infections were more commonly detected in those aged ≤5 years overall (18% compared to 7% in the >5 years group) and for each viral infection, particularly rv, rsv, adv and cov (detected in 6%, 3%, 3% and 2% of these infections, respectively, in those aged >5 years) (fig. 5a,b) . a total of 1389 virus pairs were detected in 1086 episodes of co-infection; most episodes involved two viruses (87%, 964/1086), the remaining involved three (n = 105), four (n = 15) and five (n = 2) viruses. all viruses were detected with most others at least once (fig. 5c) ; however, a clustering pattern was evident in which rv, adv, rsv and cov were frequently detected with one another. the most common virus detection in a co-infection was rv (56% of coinfections), the majority of which were with adv (n = 195, 25%) and rsv (n = 181, 23%). other viruses relatively frequently detected in co-infections were adv, rsv and cov; constituting 31%, 30% and 28% of co-infections, respectively. we found a significant positive correlation between virus detection frequencies in mono-infections and coinfections [pearson's product-moment correlation = 0·88 (95% ci 0·60-0·97, p < 0·001) and fitted linear regression model slope = 0·85 (p < 0·001)] (fig. 5d) . however, iav and ibv were identified in co-infections at relatively low frequencies (n = 121 and n = 68, respectively) compared to non-influenza viruses (e.g. rv, n = 678) (fig. 5d ). table 1 summarizes the results of univariable and multivariable logistic regression analyses for associations with viral infection. season, age group, and patient origin were significantly associated with the odds of viral infection based on unadjusted odds ratio estimates. in the multivariable analysis, several independently significant factors were identified based on the adjusted odds ratios. viral respiratory infections were more likely to be detected in winter, in children aged 1-5 years, and in gp attendees, irrespective of the other factors. following adjustment for multiple factors, time period was also a significant predictor (because of a negative confounding by age): the odds of viral infection were significantly greater postpandemic than pre-pandemic. significant statistical interactions (based on p < 0·05) revealed that the effect of age was not homogeneous across gender or patient-origin subgroups. this variation in age association across other factors is shown in figure 6a ,b where age-specific infection prevalences are stratified by the third factor. these figures show that the age distribution of infection differed according to gender and patient-origin subgroups. table 2 summarizes the results of univariable and multivariable logistic regression analyses for associations with co-infection. several differences were found in comparison with viral infections overall. based on unadjusted odds ratio estimates, time period, season (autumn only), age group, gender and patient origin were significantly associated with co-infection. however, in the multivariable analysis time period and gender were confounded by age and were therefore not identified as significant independent factors. in contrast to viral infection overall, co-infections were equally likely to be detected in spring and winter, were less likely to be detected in the 1-5 years age group than infants, and were more likely to be detected in general hospital attendees (outpatients and those not admitted to critical care wards) than gp attendees. significant statistical interactions (based on p < 0·05) revealed that the effect of age on co-infection status was not homogeneous across gender and patient-origin groups. in contrast to viral infection overall, co-infections were relatively more common in males than females in those aged 46-64 years and hospital attendees in all age groups (fig. 6c-d) . there was no evidence of a poor model fit based on the global goodness-of-fit tests: (i) p values = 0·147, 0·07, 0·07 for the main effect model and two models with interaction terms, respectively, for associations with viral infection overall, and (ii) p values = 0·940, 0·985, 0·746 for the main effect model and two models with interaction terms, respectively, for associations with co-infection. the advent of multiplex pcr as part of routine diagnostics provides an unprecedented opportunity for studying the epidemiology of multiple respiratory viruses simultaneously within a single population. previous uk-based studies have highlighted the utility of laboratory-based surveillance for monitoring respiratory infection trends, and in comparing the relative burdens between viruses [10, 13, 23] . our study is the first to compare the epidemiologies of different respiratory virus groups utilizing extensive diagnostic data generated by multiplex rt-pcr from patients attending both primary and secondary healthcare services. the collation of test-negative results by diagnostic laboratories provides valuable denominator information for measuring disease occurrence, to estimate the relative contribution of different pathogens to healthcare usage (such as gp consultations) and to provide an early warning for periods of increased healthcare pressures. importantly, the diagnostic test data utilized in this study were generated by a single laboratory, permitting a more consistent comparison of trends across patient and virus groups because testing methods were on the whole standardized throughout the study. our study has revealed changes in the frequency of virological testing of respiratory illnesses in the nhsggc health board during 2005-2013, with adults representing an increasingly greater percentage of episodes. however, age-specific prevalences were greater in the 2013 influenza season than in 2005 for all age groups. it is possible that there is raised awareness in the public and/or clinicians, and consequently greater healthcare seeking and/or sampling behaviour in adults. alternatively these results could reflect a true increase in non-viral causes of respiratory illness in this age group. we note that a shift in the demography of the glasgow population has been reported [24] . our observations might indicate the impact of an ageing population on respiratory-related healthcare services, through an increase in gp/hospital consultations, or a genuine increase in the incidence of adult respiratory infections. rv was the most prevalent virus overall, corroborating previous uk-based studies that include patients attending both primary and secondary healthcare services [10, 12] . the clinical significance of rv is disputed, although severe cases of disease are recognized depending on virus species, patient subgroups, and season [7, [25] [26] [27] . in additional analyses (fig. 4) we found the prevalence of severe respiratory illness (patients located in critical care wards) was significantly greater in rv infections than iav, supporting the proposition that rv is associated with more severe disease than traditionally accepted. of the other non-influenza viruses, rsv and cov were relatively highly prevalent. we note that the extent of research into the commonly circulating covs is small compared to iav and rsv, although severe clinical cases are recognized [28] . our study is the first comparative analysis in the uk to include cov, providing an important opportunity to quantify its temporal and patient subgroup distributions and co-infection patterns in comparison to the other common virus groups. we confirm that cov contributes a large fraction of infections during periods of high influenza activity and that cov is relatively frequently co-detected with other viruses. the contribution of different respiratory viruses to the healthcare burden in scotland has previously been studied [23] . further investigation on a seasonal basis is needed to help elucidate the public health relevance of rv and cov, particularly since cov has a similar age distribution as the influenza viruses. the remaining viruses (adv, mpv, piv1-4) were detected in comparatively smaller numbers on a yearly basis and during months of high influenza activity. the 9-year study period provided a novel opportunity to compare the epidemiology of respiratory viruses before and after the 2009 influenza a pandemic [18] . in our multivariable statistical analysis we found viral infections to be more likely in the post-pandemic era. this result was independent of other factors such as patient's age implying non-patient factors, such as a change in the underlying virus population, have increased the likelihood that a patient seeking healthcare services will have a viral infection (as opposed to non-viral causes). whether this is a direct consequence of the pandemic virus, its impact on the epidemiologies of others viruses, or a consequence of long-term changes in the non-influenza virus population, remains to be elucidated. seasonal and patient-related factors corroborate existing knowledge and were independent of time, indicating the generality of these factors as predictors of viral infection. it is well recognized that the burden of viral respiratory illness lies predominantly in young children [29] . we found that in patients with respiratory illness attending healthcare facilities, those aged 1-5 years were more likely than other age groups to have a viral infection independent of season or time period. the most commonly detected viruses in this age group were rv, rsv, adv and mpv (20%, 9·3%, 9·1% and 4·7% of infections, respectively) corroborating previous reports [23, 30] . together with a recent study that found bacterial-viral co-infections were relatively uncommon in children with pneumonia [31] , these findings support the concern regarding the overprescription of antibiotics in children [32] . that the increasing trend in virus prevalence was most notable in infants (<1 year) also warrants further attention. while it is possible that these findings are influenced by changes in clinical testing decisions, we note that this trend is particularly pertinent in relation to recent european outbreaks of enterovirus d68 in children [33] ; investigation into the contribution of individual viruses will be the focus of future work. we further note that, based on the multivariable statistical analyses, the increasing trend in prevalence in children explained why co-infections were more likely detected in the post-2009 pandemic era. there are very few studies describing co-infection patterns in respiratory viruses. our study provides the largest examination to date, confirming that around 11% of viral infections in patients attending healthcare services in an urban setting involve more than one virus, similar to the 10·4% reported by a previous uk-based study [12] . that nearly all respiratory viruses were co-detected with all other viruses highlights the sufficient opportunities for co-infections. we would expect co-infection frequencies to reflect individual virus prevalences. indeed, in line with the aforementioned study [12] , rv was the most common detection in co-infections, rv/rsv was a frequent pairing, and most co-infections were in children aged <5 years. our study also reveals that covs are relatively frequently involved in co-infections. however, co-infections with influenza viruses were relatively few, perhaps explained by differences in their age and seasonal distributions, or an inter-viral interference [34] . we found that the average age of co-infection was 3 years, compared to 17 years for viral infections overall, and co-infections were more likely in infants than in those aged 1-5 years. that co-infections were more likely in young children is probably explained by (i) a greater opportunity for co-infection due to a shorter exposure lifetime and consequently greater susceptibility to a wider array of viruses, and (ii) a greater chance of co-infections being detected because children tend to shed virus for longer periods. in adults, the age distribution of co-infections differed according to gender and patient origin; the prevalence was greatest in males and in general hospital attendees not admitted to critical care wards for those aged 46-64 years (fig. 6c,d) . this result provides insight into an age-dependent factor in co-infection patterns in adults but must be viewed with some caution; it is potentially influenced by a bias in multiple specimens submitted in relation to single episodes of illness in adults, most likely as a result of comorbidities. interestingly, co-infections were more likely in general hospital attendees not admitted to critical care wards than gp attendees, supporting the potential role of co-infections in illness severity [35] . there are several limitations to our study to be noted. detection of viral nucleic acid may not represent active infection for all viruses in all cases [36] , potentially introducing detection biases temporally and across patient groups. furthermore, the timing of infection events, and variation in shedding duration across virus and patient groups [37, 38] , could potentially bias the observed co-infection patterns. we also note that our study lacked information on the presence/absence of bacterial pathogens which are also significant contributors to respiratory infections. one further important consideration is that laboratory diagnostic data cannot inform on the epidemiology of asymptomatic infections in the community, or in symptomatic people who do not attend healthcare services. furthermore, that viral populations are not static could also impact on the generalisability of the observed trends and associations; the introduction of new strains can alter disease outcomes, and consequently healthcare seeking behaviour, influencing the stability of healthcare consultation rates in patient subgroups. given the dynamic nature of virus populations, the epidemiological information generated through surveillance must be maintained to ensure future vaccine and antiviral developments are directed to where they are most needed [39, 40] . 1·15 (0·75-1·79, p = 0·521) or, odds ratio; ci, confidence interval. * distribution of patient numbers, with corresponding % in parentheses, across factor levels for all patients (summary) and for co-infection and mono-infection groups. † unadjusted or based on univariable logistic regression. ‡ adjusted or based on multivariable logistic regression. § patients' location corresponding with first clinical sample: gp, general practitioner's surgery; hospital (general), outpatients and non-critical-care patients; hospital (critical care), patients admitted to an intensive care, intensive therapy, high dependency, or coronary care unit. our study provides the most comprehensive description of viral respiratory infections in the uk to date, revealing new epidemiological insights with public health relevance. of particular concern is a greater viral prevalence in 2013 compared to 2005, particularly in infants, and a greater risk of viral infection in the post-2009 pandemic era. further investigation into the long-term temporal dynamics of individual viruses and the epidemiological consequences of virus cocirculation is needed. for supplementary material accompanying this paper visit http://dx.doi.org/10.1017/s0950268816000339. global burden of respiratory infections due to seasonal influenza in young children: a systematic review and meta-analysis global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: a systematic review and meta-analysis disability-adjusted life years (dalys) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the global burden of disease study viruses and bacteria in the etiology of the common cold adults hospitalised with acute respiratory illness rarely have detectable bacteria in the absence of copd or pneumonia; viral infection predominates in a large prospective uk sample investments in respiratory infectious disease research 1997-2010: a systematic analysis of uk funding association between human rhinovirus c and severity of acute asthma in children human metapneumovirus: review of an important respiratory pathogen newly discovered coronavirus as the primary cause of severe acute respiratory syndrome a new laboratory-based surveillance system (respiratory datamart system) for influenza and other respiratory viruses in england: results and experience from seasonal variations of 15 respiratory agents illustrated by the application of a multiplex polymerase chain reaction assay single, dual and multiple respiratory virus infections and risk of hospitalization and mortality assessing the burden of influenza and other respiratory infections in england and wales epidemiology characteristics of respiratory viruses found in children and adults with respiratory tract infections in southern china viral etiologies of hospitalized acute lower respiratory infection patients in china epidemiology and viral etiology of the influenza-like illness in corsica during the 2012-2013 winter: an analysis of several sentinel surveillance systems impact of the 2009 h1n1 pandemic on age-specific epidemic curves of other respiratory viruses: a comparison of pre-pandemic, pandemic and postpandemic periods in a subtropical city during the summer 2009 outbreak of 'swine flu' in scotland what respiratory pathogens were diagnosed as h1n1 a goodness-of-fit test for binary regression models, based on smoothing methods r: a language and environment for statistical computing. r foundation for statistical computing disease burden of the most commonly detected respiratory viruses in hospitalized patients calculated using the disability adjusted life year (daly) model population trends by age group in glasgow human rhinovirus species and season of infection determine illness severity clinical features and complete genome characterization of a distinct human rhinovirus (hrv) genetic cluster, probably representing a previously undetected hrv species, hrv-c, associated with acute respiratory illness in children the significance of rhinovirus detection in hospitalized children: clinical, epidemiological and virological features epidemiology and clinical presentations of the four human coronaviruses 229e, hku1, nl63, and oc43 detected over 3 years using a novel multiplex real-time pcr method respiratory viral infections in infants: causes, clinical symptoms, virology, and immunology respiratory viral infections during the 2009-2010 winter season in central england, uk: incidence and patterns of multiple virus co-infections community-acquired pneumonia requiring hospitalization among u.s. children respiratory virus infections in febrile children presenting to a general practice out-of-hours service emergence and epidemic occurrence of enterovirus 68 respiratory infections in the netherlands in 2010 do rhinoviruses reduce the probability of viral co-detection during acute respiratory tract infections? single and multiple respiratory virus infections and severity of respiratory disease: a systematic review clinical utility of pcr for common viruses in acute respiratory illness influence of age, severity of infection, and co-infection on the duration of respiratory syncytial virus (rsv) shedding epidemiology of multiple respiratory viruses in childcare attendees emerging respiratory viruses: challenges and vaccine strategies respiratory viruses other than influenza virus: impact and therapeutic advances the authors are grateful to dominic mellor, emma thomson, louise matthews and richard reeve for their helpful critique of the manuscript and discussions. a subset of the clinical samples was provided by health protection scotland as part of the scottish enhanced respiratory viral infection surveillance programme.this work was supported by the medical research council uk (grant g0801822). none. key: cord-261282-r1nprlne authors: chughtai, a. a.; wang, q.; dung, t. c.; macintyre, c. r. title: the presence of fever in adults with influenza and other viral respiratory infections date: 2016-10-03 journal: epidemiol infect doi: 10.1017/s0950268816002181 sha: doc_id: 261282 cord_uid: r1nprlne we compared the rates of fever in adult subjects with laboratory-confirmed influenza and other respiratory viruses and examined the factors that predict fever in adults. symptom data on 158 healthcare workers (hcws) with a laboratory-confirmed respiratory virus infection were collected using standardized data collection forms from three separate studies. overall, the rate of fever in confirmed viral respiratory infections in adult hcws was 23·4% (37/158). rates varied by virus: human rhinovirus (25·3%, 19/75), influenza a virus (30%, 3/10), coronavirus (28·6%, 2/7), human metapneumovirus (28·6%, 2/7), respiratory syncytial virus (14·3%, 4/28) and parainfluenza virus (8·3%, 1/12). smoking [relative risk (rr) 4·65, 95% confidence interval (ci) 1·33–16·25] and co-infection with two or more viruses (rr 4·19, 95% ci 1·21–14·52) were significant predictors of fever. fever is less common in adults with confirmed viral respiratory infections, including influenza, than described in children. more than 75% of adults with a viral respiratory infection do not have fever, which is an important finding for clinical triage of adult patients with respiratory infections. the accepted definition of ‘influenza-like illness’ includes fever and may be insensitive for surveillance when high case-finding is required. a more sensitive case definition could be used to identify adult cases, particularly in event of an emerging viral infection. respiratory infections are common and one of the leading causes of morbidity and mortality, particularly in the extremes of age [1] [2] [3] . influenza a and b, human rhinoviruses (hrv), respiratory syncytial virus (rsv), adenoviruses (adv) and parainfluenza virus (piv) are common respiratory viruses in adults and children [1] [2] [3] [4] [5] . of respiratory infections, influenza is the most well studied viral infection, and is commonly reported (around 50%) as the cause of epidemics of respiratory infection, including nosocomial outbreaks [6] . influenza virus is commonly isolated from febrile paediatric and elderly patients presenting with influenza-like illness (ili) and acute respiratory illness (ari) symptoms [1] . the accepted clinical case definition of ili includes fever, which may be suitable for identifying paediatric cases, but less so for adults. fever is thought of as the most common presenting symptom of influenza in hospital emergency departments; however, the presence of fever depends on the age of person and the type of virus [7] [8] [9] [10] . it is known that fever is less common in adults than children with influenza, and that adults may have atypical presentations [5, 6, 11] . in a matched case-control study in finland, 317 laboratoryconfirmed influenza cases and 353 controls with respiratory symptoms were recruited in children aged 413 years. fever was present in 89·8% (317/353) and 35·7% (126/353) of cases and controls, respectively [12] . in contrast to this, fever is not a common presentation in adults with laboratory-confirmed influenza. monto et al. [13] examined clinical trial data of 3744 adult ili cases (defined as body temperature 537·8°c or patients subjective feeling of feverishness) and of those 2470 (66%) had laboratory-confirmed influenza. fever (537·8°c) was reported in 68% of laboratory-confirmed influenza cases, compared to 40% other ili cases [13] . during a randomized clinical trial (rct) around the efficacy of facemask and hand hygiene in the household setting, 44% (15/34) of secondary cases with influenza a and 32% (8/ 25) of cases with influenza b had fever history [14] . the rate of fever was 66% (137/207) in hospitalized influenza cases in a us study [15] . another us study showed that less than half (42·4%) of healthcare workers (hcws) with laboratory-confirmed influenza presented with fever [5] . fever is a less common presenting symptom in elderly people which may lead to diagnostic and treatment delays [16] . in patients admitted with myocardial infarction, 9% had unrecognized and undiagnosed influenza on testing at admission, highlighting the low level of clinical suspicion of influenza [17] . the rate of fever also varies between influenza strains, being more common in influenza a strains than b, and higher in h3n2 [7] [8] [9] [10] 18] . fever is a commonly reported symptom during influenza outbreaks and pandemics due to novel and more virulent nature of strains. in china 67·4% of the patients infected by influenza a(h1n1)pdm09 had fever [19] . in another study in beijing 465 suspected ili cases were tested and of those 318 (68%) were positive for influenza virus (pandemic h1n1-165 and seasonal influenza h3n2-153) and all had history of fever [20] . the aim of this study was to compare the rates of fever in adult subjects with confirmed influenza and other respiratory virus infections and examine predictors of fever. we analysed a dataset of laboratory-confirmed viral respiratory infections collected from three clinical trials of hcws where active surveillance for respiratory viral illness was conducted in prospective follow up [21] [22] [23] . the same methods, data collection forms and outcome measures, were used across the three studies, allowing the data to be pooled [21] [22] [23] . two studies were conducted in beijing china: trial 1 (2008/2009) and trial 2 (2009/2010) and another study (trial 3) was conducted in hanoi, vietnam in 2010/2011 [21] [22] [23] . in all clinical trials, participants were asked to complete diary cards on a daily basis to collect information on number of working hours, patients seen, mask use hours, high-risk procedures performed and appearance of respiratory symptoms. thermometers were given the participants to measure their temperature daily and at symptom onset. symptomatic cases were asked to complete sick patient follow-up forms and detailed information was collected on the following symptoms: chill or fever, cough, congestion, runny nose, sore throat, sneezes, lethargy, loss of appetite, abdominal pain, muscle or joint aches. swabs of both tonsils and the posterior pharyngeal wall were collected on the day of reporting. in all rcts, fever was defined as having body temperature 538°c. clinical respiratory illness (cri) and ili were in the primary outcomes in three clinical trials. cri was defined as two or more respiratory symptoms or one respiratory symptom and a systemic symptom and ili was defined as fever 538°c plus one respiratory symptom [21] [22] [23] . we analysed data from all subjects with a positive isolation of a respiratory virus by multiplex polymerase chain reaction (pcr). descriptive analysis was conducted for rates of fever by virus type. a logistic regression analysis was used to determine the predictors of fever. a multivariable log binomial model was fitted, using a generalized linear model to estimate relative risk (rr). all variables were included in initial model. in the final model, we included only those variables that were significant (p < 0·25) in initial analysis. a backward elimination method was used to remove the variables that did not have any confounding effect, i.e. could not make meaningful change (±10%) in the rr of the comparison arm. finally we estimated the rates of cri and ili in the laboratory-confirmed viral respiratory infections and laboratory-confirmed influenza infections. the data was analysed using sas v. 9.4 (sas institute inc., usa). the demographic characteristics of 158 cases with laboratory-confirmed viral infections are presented in table 1 . ninety (57%) cases were from china and 68 (43%) were from vietnam. the mean age of hcws was 32·8 years and most participants were nurses (65%) and female (87%). most cases were non-smokers (92%) and had not received influenza vaccine (86%). viruses isolated included rhinovirus (n = 75, 47%), rsv (n = 28, 18%), influenza (n = 13, 8%), piv (n = 12, 8%), human metapneumovirus (hmpv; n = 7, 4%), coronavirus (n = 7, 4%) and adv (n = 1, 1%). more than one virus was isolated in 15 cases (9·5%), including nine cases with influenza co-infection. fever was documented in 23·4% cases (37/158) with a positive laboratory viral diagnosis. table 2 details rates of fever (538°c) associated with individual viral respiratory infections. hrv was the most common infection and 25·3% (19/75) of these had a fever. in 28 cases of rsv, four (14·3%) had fever; 8·3% (1/12) of piv and 30% (3/10) of influenza a cases had fever. seven cases of coronavirus and hmpv each were confirmed and of those two (28·6%) had fever. when cases with influenza and a co-infection were included, 36·4% (8/22) had fever. in univariate analysis, country, gender and smoking were significant predictors of fever. country and smoking remained significant predictors in multivariate analysis while gender became non-significant. fever rate was significantly higher in hcws in vietnam compared to hcws in china [rr 2·99, 95% confidence interval (ci) 1·24-7·20]. smokers were around five times more likely to have fever compared to non-smokers (rr 4·65, 95% ci 1·33-16·25). virus type was not associated with fever in univariate analysis; however, after adjusting for other variables, rates of fever were significantly higher in hcws co-infected with more than one virus compared to all other viruses excluding influenza (rr 4·19, 95% ci 1·21-14·52) ( table 3) . cri symptoms were present in 84·8% (137/158) of hcws with laboratory-confirmed viral infections and 90·9% (20/22) laboratory-confirmed influenza infections. the corresponding rates of ili in the two groups were 9·5% (15/158) and 13·6% (3/22), respectively. we have shown, using prospectively collected data, that the rate of fever in adults with confirmed viral respiratory infections is much lower than described in children [1, 9] . the standard clinical case definition of ili requires fever to be presentthe majority of influenza cases in this series would have been missed using the ili definition. this has implications for effective triage, early antiviral treatment and preventive measures for adults with influenza, particularly during outbreaks and pandemic situations. for other respiratory infections, clinical case definitions need to be more sensitive, or >75% of cases will be missed. the main implication for future surveillance, measurements and research studies is that the ili case definition in adults may be highly insensitive. for some types of surveillance systems, this may not be an issue, but for diagnostic screening in event of an emerging viral infection (such as for triage and implementation of infection control protocols) [24, 25] , a more sensitive case definition is needed. rates of fever in influenza and other viral respiratory infections in this study were lower compared to other studies which report fever in around 50-70% adult cases [1, 5, 13, 15] . however, this variation may be due to different study base, case definition and viral strains, as well as the prospective measurement of incident infections. many research studies use fever as an inclusion criterion for laboratory testing [13, 26] . while this may be suitable for studies in children, it is not adequately sensitive for studies of adults, as we have shown the majority of confirmed cases will be missed. the cut-off point for fever could be another factor in sensitivity. some studies have set lower cut-off points for fever, and report higher rates of fever in laboratory-confirmed influenza cases [13, 27] . carrat et al. collected data of cases presented in 35 general practices in france and collected nasal swabs from suspected influenza cases and defined fever as 537·8°c. they found fever in influenza a(h3n2), influenza a(h1n1) and influenza negative cases in 95·2%, 77·5% and 72·7%, respectively. applying a cut-off of 538·2°c, the corresponding rates are 82·2%, 59·3% and 43·9% [27] . symptoms of feverishness (subjective feeling of fever) are included in ili definitions in some cases [13] . previous studies report high rates of fever in children compared to the adults [11] . low rates of fever in adults may also be due to protection via cross-reactive antibodies due to age-dependent differences in the immunity [7, 28] . continued exposure to influenza throughout life may result in a broader protection with age. infection may provoke a stronger immune response in children with minimal to no exposure history compared to adults. therefore influenza infection history might help explain potential differences in clinical symptom severity (and presence of fever) between children and adults. a recent study reported high rates of influenza and other respiratory virus in afebrile hcws with only respiratory symptoms [5] . of 22 laboratory-confirmed influenza cases in this study, only three (13·6%) had ili symptoms, which is very low compared to other studies. in a prospective influenza surveillance study, ili symptoms were present in 48% of adults and 61% of children with laboratory-confirmed influenza virus [1] . cri symptoms were present in 90·9% (20/22) of laboratory-confirmed influenza cases in this study. a highly sensitive definition of influenza may be required to diagnose most of adult influenza cases in the clinical setting to ensure rapid treatment and isolation, and prevention of nosocomial transmission. inclusion of ili cases may overestimate the proportion of febrile cases in influenza surveillance given fever is included in the definition. pre-symptomatic and asymptomatic influenza cases will also be missed, although infectivity and transmissibility of these cases is yet to be proven [29] . longitudinal studies, where all participants are tested, provide similar estimates around rates of fever as in our study [14] . we propose a more sensitive clinical case definition without fever as a requisite criterion. clinical signs and symptoms are less studied for other viral respiratory infections, but available evidence suggests that other respiratory viruses are associated with a lower rate of fever compared to influenza [5, [30] [31] [32] [33] . putto and colleagues [30] examined the clinical records of 258 children (>3 months) in a large hospital in finland, including adv (25 cases), influenza a and b (74 cases), piv (99 cases) and rsv (60 cases). fever (539·0°c) was recorded in 68% cases with adv, 84% influenza a virus, 65% influenza b, 41% piv-1, 50% piv-2, 47% piv-3, and 52% rsv. van den hoogen and colleagues estimated the prevalence and clinical symptoms of hmpv infection, in the netherlands and fever was reported in 61% of the hmpv-positive cases [31] . in hong kong, hmpv was found in 5·5% (32/587) of children admitted in hospitals and all had fever [32] . manoha et al. examined nasal wash specimens from 931 hospitalized children and found hmpv (6%), rsv (28·5%), rhinoviruses (18·3%), influenza a (6%), piv-1 (0·2%) and piv-3 (0·3%). fever was reported in 39·2% cases with hmpv, 37·8% cases with rsv and 30·2% with rhinovirus [33] . of the 210 elderly patients with influenza and 145 with rsv, fever was reported in 65% and 50%, respectively [3] . a us study also reported low rates of fever in hcws infected with coronavirus 229e (13·5%), coronavirus hku (11·4%), coronavirus nl63 (31·3%) and rsv (12·9%) and all cases of hmpv were without fever [5] . rate of fever for all other viruses (excluding influenza) was 21·5% (28/130) in this study. co-infection with more than one virus was the strongest predictor of fever for adults with confirmed viral respiratory infections in the present study. previous studies also show high rates of fever in cases with dual respiratory viral infections compared to single viral infection [34, 35] . rates of hospitalization and icu admission are also reported to be higher in cases with dual respiratory viral infections [36] [37] [38] . increased severity of symptoms in co-infection cases might be due to an altered immune response [34] . around 10% (15/158) of cases in our dataset were infected with more than one virus. drews et al. reviewed the data of eight prospective epidemiological studies and reported the rate of co-infection was 5% [37] . studies in children generally report higher rates of co-infection cases (17-20%) [34, 35, 38, 39] . clinicians should consider the possibly of co-infection if a patient presents with fever; however, further epidemiological and clinical studies are required. smoking was also a significant predictor of fever in this study. smoking increases the risk of viral and bacterial infections through changes in respiratory epithelial and altered immune response [40] [41] [42] [43] .the risk of influenza also increases several times in smokers, compared to non-smokers [40] . atypical clinical presentation of influenza and other respiratory infections in adults could be due to altered structural and immune response associated with active/passive smoking and other environmental hazards. the mechanism by which smoking increases the risk of fever is not clear. high rates of fever in smokers may also be due changes in immunoglobulin levels which could increase viral load. the severity of symptoms generally increases when high viral load is detected in the blood [44] . the difference in fever rates between china and vietnam may be due to prevalence of viruses and co-infection. rsv was the most commonly isolated pathogen from china (31%), followed by rhinovirus (20%) and influenza virus (13%). in contrast to this hrv was the most commonly isolated pathogen from vietnam (85·3%). the number of cases with co-infection were also different in the two countries -13 (14%) in china and two (3%) in vietnam. in multivariate analysis, we adjusted for country and type of virus. limited data are available regarding the prevalent viruses circulating in china during the study period. for the trial 1 period, all influenza was influenza a(h1n1)pdm. for the trial 2 period, 21·3% were h1n1pdm, 2·9% were h3n2, 3·0% were influenza b victoria, 2·6% were influenza b yamagata, 71·2% were influenza a unsubtyped (y. zhang, beijing centre for disease prevention and control, personal communication). we could not obtain data on the viruses circulating in vietnam during the study period. there are some limitations to this study. we did not subtype the influenza strains, and studies show that the rate of fever also varies between influenza strains [7] [8] [9] [10] 18] . fever data was self-reported but self-measured in three trials using a traditional glass and mercury thermometer. lower fever rates in chinese hcws in this study might be due to due to differences in circulating viruses (and their pyrogenicity) between the two countries when the studies were conducted. a japanese study of children with influenza reported a tendency towards shorter duration of fever with increasing age in children [18] ; however, age and other demographic characteristics were not significant in that study. compared to children, this study shows that adults are less likely to have fever with a respiratory viral infection, even influenza. the implication of this finding is that for rapid treatment and reducing the risk of transmission of infection, clinicians should be aware that a diagnosis of viral respiratory infection, even influenza, is possible in the absence of fever. many of these infections are transmissible even when infected persons are asymptomatic or presymptomatic, and greater vigilance for respiratory symptoms in hcws could reduce nosocomial transmission of respiratory viral infections. the absence of fever should not preclude a differential diagnosis of influenza or other respiratory viruses in adults. influenza surveillance in communitydwelling elderly compared with children viral respiratory infections in the institutionalized elderly: clinical and epidemiologic findings respiratory syncytial virus and influenza a infections in the hospitalized elderly respiratory viruses transmission from children to adults within a household influenza among afebrile and vaccinated healthcare workers clinical manifestations and consequences of influenza tecumseh study of illness. xiii. influenza infection and disease differing virulence of h1n1 and h3n2 influenza strains influenza a and b virus infections in children differences in clinical features between influenza a h1n1, a h3n2, and b in adult patients seasonal influenza in adults and children -diagnosis, treatment, chemoprophylaxis, and institutional outbreak management: clinical practice guidelines of the infectious diseases society of america signs and symptoms predicting influenza in children: a matched case-control analysis of prospectively collected clinical data clinical signs and symptoms predicting influenza infection viral shedding and clinical illness in naturally acquired influenza virus infections is influenza an influenza-like illness? clinical presentation of influenza in hospitalized patients fever in the elderly ischaemic heart disease, influenza and influenza vaccination: a prospective case control study natural course of fever during influenza virus infection in children clinical features of the initial cases of 2009 pandemic influenza a (h1n1) virus infection in china clinical predictors for diagnosing pandemic (h1n1) 2009 and seasonal influenza (h3n2) in fever clinics in beijing a cluster randomised trial of cloth masks compared to medical masks in healthcare workers a cluster randomized clinical trial comparing fit-tested and non-fit-tested n95 respirators to medical masks to prevent respiratory virus infection in health care workers a randomized clinical trial of three options for n95 respirators and medical masks in health workers hospital triage system for adult patients using an influenza-like illness scoring system during the 2009 pandemic -mexico pandemic influenza triage tools: user guide predicting influenza infections during epidemics with use of a clinical case definition evaluation of clinical case definitions of influenza: detailed investigation of patients during the 1995-1996 epidemic in france influenzavirus infections in seattle families, 1975-1979. i. study design, methods and the occurrence of infections by time and age does influenza transmission occur from asymptomatic infection or prior to symptom onset? fever in respiratory virus infections prevalence and clinical symptoms of human metapneumovirus infection in hospitalized patients children with respiratory disease associated with metapneumovirus in hong kong epidemiological and clinical features of hmpv, rsv and rvs infections in young children single versus dual respiratory virus infections in hospitalized infants: impact on clinical course of disease and interferon-gamma response correlation of viral load of respiratory pathogens and co-infections with disease severity in children hospitalized for lower respiratory tract infection single, dual and multiple respiratory virus infections and risk of hospitalization and mortality dual respiratory virus infections multiple simultaneous viral infections in infants with acute respiratory tract infections in spain multiple versus single virus respiratory infections: viral load and clinical disease severity in hospitalized children cigarette smoking and infection smoking and the outcome of infection cigarette smoke extract suppresses human dendritic cell function leading to preferential induction of th-2 priming the associations of race, cigarette smoking, and smoking cessation to measures of the immune system in middle-aged men correlation of rhinovirus load in the respiratory tract and clinical symptoms in hospitalized immunocompetent and immunocompromised patients the authors thank the staff at the beijing centre for disease control and national institute of hygiene and epidemiology. thanks are also due to the staff from the hospitals in china and vietnam which participated. 3m helped in fit testing during the first rct in china, no financial support was provided. the second rct in china was funded through the australian national health & medical research council of australia (grant no. 630787). funding to conduct the vietnam trial study was received from the australian research council (arc) (grant no. lp0990749). dr abrar chughtai had testing of filtration of masks by 3m for his phd. professor c. raina macintyre has held an australian research council linkage grant with 3m as the industry partner, for investigator-driven research. 3m have also contributed supplies of masks and respirators for investigator-driven clinical trials. she has received research grants and laboratory testing as in-kind support from pfizer, gsk and bio-csl for investigator-driven research. the remaining authors have no competing interests to declare. key: cord-325226-8zrtjuwf authors: biswas, raaj kishore; afiaz, awan; huq, samin title: underreporting covid-19: the curious case of the indian subcontinent date: 2020-09-11 journal: epidemiol infect doi: 10.1017/s0950268820002095 sha: doc_id: 325226 cord_uid: 8zrtjuwf covid-19 has spread across the globe with higher burden placed in europe and north america. however, the rate of transmission has recently picked up in lowand middle-income countries, particularly in the indian subcontinent. there is a severe underreporting bias in the existing data available from these countries mostly due to the limitation of resources and accessibility. most studies comparing cross-country cases or fatalities could fail to account for this systematic bias and reach erroneous conclusions. this paper provides several recommendations on how to effectively tackle these issues regarding data quality, test coverage and case counts. since the inception of the covid-19 pandemic, both the media and research focus were on china, europe and the usa primarily due to the large cluster of cases in these regions during the early days. however, despite low fatality rates and total cases, the focus then shifted to the low-and middle-income countries (lmics) soon after the outbreak had hit the indian subcontinent (isc) and their covid-19 response dynamics. in the meantime, academic studies started making inferences on the covid-19 response effectiveness through comparing the disease prevalence and fatality rates between higher and lower income nations in order to investigate the curious case of low covid-19 infection rates among the lmics. conducting research on lmics with limited data could often lead to erroneous findings and biased interpretations, which is becoming a concern with the avalanche of studies published daily. the reasons behind inadequate data in lmics or even low fatality/detection rates could be qualitatively discussed. while these would not impede academicians in conducting research, caution should be exercised during interpretation and in proposals of 'evidencebased' policies and the development of operational plans focusing on mitigation and response in respective contexts. for a greater focus and authors' area of expertise, this paper is limited to the countries in the isc, which is a sample representation of lmics; however, country-wise discussion of lmics based on the respective socioeconomic context is required for a reasonable generalisation. india, pakistan and bangladesh are among the worst 20 countries affected by the covid-19 pandemic in terms of total number of cases; however, they are ranked 138, 139 and 147, respectively, in tests per million population, as of 18 june 2020 [1] . it is worth noting that countries such as bangladesh reached the threshold of 10 000 tests per day on 20 may, an astonishing 74 days after the detection of the first confirmed case which was still a mere fraction of the number of tests conducted by countries such as the usa, uk and italy. this lack of testing capabilities during the early days accompanied by the limited protective gears for health personnel and low implementation capacity related to the response of such pandemics could have concealed the true rate of infection and disease spread in the lmics of the isc. due to limited testing facilities and availability of trained personnel, long delays in both sample collection and dissemination of test results are observed in these countries [2, 3] . furthermore, these inadequate facilities could compromise sample handling and storage as they require strict low temperature preservation for an optimum result [3] , which is a challenge for lmics including the isc. evidently, the official press releases in bangladesh reflected that all collected samples were not tested daily with long backlogs leading to curbing sample collection [4] and resulting in public distress as many non-covid medical facilities require certification of a negative test result before admitting new patients. another issue for maintaining a rigorous score of transmission rates is to adequately define both cases and deaths from covid-19, which varies across borders resulting in inconsistencies among reports [5] . for example, countries such as china and bangladesh, have changed the definition of confirmed cases or recoveries during the on-going pandemic [6, 7] . moreover, leadership and political goodwill during such crisis play a crucial role in data collection, testing quality and country-wide coverage [8] . testing coverage in the indian subcontinent has yet to reach the peripheral areas. the government testing facilities are mostly free but are consequently overwhelmed with backlogs, whereas the costs of private tests are well out of the reach of most people [9] . particularly with countrywide lockdowns and reduced patient transport facilities, it is hard to acquire free-public amenities as an observed 60% increase in extreme poverty in bangladesh and 90% of people (around 400 million) working in the informal economy in india are at the risk of deeper poverty due to the lack of work during lockdowns [10, 11] . in such scenarios, people living in urban slums or rural areas are likely to prioritise wage-earning activities meeting the heightened unmet basic needs in the midst of low economic flow considering the risks of covid-19 infection instead of a 14-day post-test quarantine. furthermore, given that the testing centres are mostly located in metropolitan areas, the coverage is often centralised to a few locations. for example, as of 18 june 2020, 64.1% of all tests in bangladesh were conducted in dhaka district compared to 0.01% in jessore district [12] , and 38.3% tests in pakistan were conducted in sindh compared to 0.50% in gilgit-baltistan [13]. this decidedly undermines the coverage across these countries and limits the cases to few privileged cohorts challenging identification of the disease at the community level and underrepresenting cluster transmission for an extended period of time. the lack of testing facilities in peripheral areas requires samples to be transported over long distances in a limited timeframe, which could jeopardise the reliability of samples and test results. the presence of centralised laboratories in few locations contributes to inadequate risk assessments across the country and impacts subsequent decision making. the bbc has put forward an interesting idea of calculating the underreported deaths of covid-19 using fatality data from previous years [14] . while this could be a way forward, the burden of diseases varies yearly. one particular example could be the expected decrease in road fatalities in 2020 and an increase in suicide rate due to the adverse effects on mental health from stress during lockdown. similarly, as covid-19 is not identical with regard to seasonal outbreaks such as measles or dengue in bangladesh or wild polio in pakistan, we cannot predict covid-19 fatality rate from the mortality of the previous seasonal outbreaks, which is likely to lead to a dubious understanding of covid-19 numbers in these countries. instead the total increase in death count after adjusting for typical seasonal diseases observed in the previous years could provide a better crude estimate of the impact of covid-19. another avenue of estimating some of the deaths by covid-like symptoms is data from graveyards and crematoriums. in west bengal, india, the number of bodies cremated is nearly seven times more than the typical rate, whereas in dhaka and narayanganj districts of bangladesh have seen twice or three times more burials in may 2020 compared to march or april [15, 16] . disease misclassifications based on the differentiation in information regarding covid-19 mortality apart from its comorbidities across multiple sources can adversely impact comparative analyses. this can also undermine subsequent resource mobilisation and evidence-based decision making in generating appropriate covid-19 management and response worldwide. moreover, the deceased with covid-like symptoms are often untested in these countries, which although is understandable considering the resource limitations, but again considerably undermines the overall death tally. another method of scrutinising the underreporting of cases is to assess the data of frontline workers since they are more likely to be tested alongside the politicians. as of june 2020, 3.28%, 2.88% and 1.45% of covid-19 fatalities in bangladesh, afghanistan and pakistan were health workers respectively, whereas they were 0.37%, 0.55% and 0.50% in the usa, uk and italy, respectively. these indicate that the health workers lacked adequate protective gear and knowledge about infection prevention and control measures in the isc during the preparation phase as well as the fact that they were likely to be over-represented in the tests conducted, leading to an overall underreporting. as of 19 june 2020, a total of 14 members of the bangladesh parliament out of 350 had tested positive with two fatalities and over 100 staff of the parliament secretariat, which resulted in a truncated budget discussion in the parliament [17] . moreover, 8.1% of the covid-19 infected belongs to the bangladesh police with 28 fatalities so far [18] . their testing is expectedly prioritised and the rapid increase in these numbers indicates that community transmission has been severely underreported in the isc. in the isc, people over 60 generally do not go out of home much, and often their external visits are limited to their familial circle [19] . furthermore, a considerable number of women are homemakers [20] . thus, a large portion of the society is used to staying at home, where able males mostly go out for work. while these scenarios are gradually changing, it could partially explain the slower transmission in isc compared to the developed nations where all household members are more likely to go out increasing the speed of infection. social dogma regarding the covid-19 is also playing a role in these nations. people in pakistan and india are typically religious. there exist concerns among them that some of the funeral rituals, such as bathing the deceased, cannot be performed if they died of covid-19 which has created public resentment towards testing [15] . furthermore, neighbourhood protests were observed in bangladesh where locals denied the covid-19 deceased to be buried in their local graveyards [21] . thus, comparison of case prevalence and fatalities across countries need to consider the cross-cultural and demographic factors. there exists a major cause for concern regarding the data quality in the isc. the subsequent use of these data in their raw form could lead to biased findings [22] . davies et al. rightly found that younger age could be a protective factor in lmics [23] ; however, it is still too early to extrapolate any generalised conclusions. non-random sampling has been conducted in the isc, and their limited capacity forces them to test mostly the symptomatic individuals and foreign returnees from the high risked countries in the early days. thus, statistical or epidemiological modelling might be statistically unprincipled with marginalised results when not taking into account the weaknesses of the data generating mechanism [24] . some findings are developing on a regular basis. for example, in the early days of the pandemic, a hypothesis was shown to be 'statistically significant' that temperature is associated with the infection rate [25] without adequate information on true r 0 value and its impact over covid-19 transmission, which gave a misleading hope to politicians who used it to assure the general mass in bangladesh leading to a sense of nationwide complacency [26] . another assumption was that they are genetically immune to the coronavirus, or bcg vaccine might work as a protective factor [27] ; however, the large death tolls of bangladeshi expatriates in saudi arabia, singapore and new york have evidently debunked it [28] . this false optimism has led to relaxation of social distancing policies in public transport and consideration of opening schools across the country [29, 30] . while it is inevitable that modelling with data from lmics on covid-19 would continue, a few cautions should be exercised: • an appropriate definition of 'death from covid-19' is essential before collapsing deaths from the covid-like symptoms with the covid-19 fatalities. for example, bangladesh changed the definition of covid-19 recovery a month after the detection of the first case [7] . • for validating the covid-19 fatality scores of a region, specific mortality causes of comorbid conditions such as respiratory and cardiovascular complication or communicable diseases representing similar manifestation of symptoms linked to covid-19 could be coded to calibrate from the total deaths during the pandemic period. however, this also needs to consider the seasonal outbreaks of diseases in specific regions. • comparison among seemingly random countries based on convenience or data availability might lead to a systematic bias. a comparative assessment on countries with similar testing coverages, analogous socioeconomic context, close geographical borders (e.g. eu or isc), cultural resemblances and contextspecific priorities with homogeneous health systems might be more insightful. • contrasting country-wise performances and covid-19 infection timeline, where the covid-19 prevalence curve has started to flatten with countries that are yet to reach its peak (figs 1 and 2) , are unreasonable as true propensity of the pandemic is yet to be observed in lmics such as isc. • instead of modelling covid-19 incidence rates across borders, cultures and demographics, this could be limited to regions with homogeneous attributes. comparison between neighbourhoods in the same locality might be better suited for hypothesis testing on mortality and disease prevalence, with the utmost care in avoiding narratives that might mislead the public opinion. • validating the covid-19 official data with random sampling, hospital data, disease burden trends and local news outlets could account for some of the underreporting biases. with thousands of phd dissertations and research articles developing the evidence-base are expected on covid-19 in years to come, it is imperative that the data validity is constantly epidemiology and infection questioned, and cross-border comparisons are routinely scrutinised given the definition of fatalities from covid-like symptoms and quality of non-random data vary worldwide. coronavirus cases lower dir residents complain of delay in virus test results as tests pile up, infections spread, patients suffer dhaka tribune. covid-19: bangladesh lags behind in sample testing despite being among 20 worst affected; dhaka tribune case-fatality rate and characteristics of patients dying in relation to covid-19 in italy a review of coronavirus disease-2019 (covid-19) the daily star. explaining the jump in number of covid-19 recovered patients in bangladesh a systematic assessment on covid-19 preparedness and transition strategy in bangladesh the caravan. india's private covid-19 tests cost highest in south asia rapid-perception-survey-on-covid19-awareness-and-economic-impact.pdf 11. international labor organization (2020) covid-19 and the world of work distribution of reported covid-19 fatalities in three higher income countries (uk, italy and germany) and three countries of isc (bangladesh, india and pakistan) for the first 169 days coronavirus info bangladesh: press release coronavirus: what is the true death toll of the pandemic? the telegraph. asia's hidden deaths: coronavirus fatalities are being covered up and undercounted dhaka tribune. burials in dhaka rose by a third in may budget session to be cut short as mps united news of bangladesh grandparents in bangladesh, india, and pakistan gender, 'race' and patriarchy: a study of south asian women dhaka tribune. panic continues to obstruct burials of coronavirus patients covid-19 in bangladesh: data deficiency to delayed decision age-dependent effects in the transmission and control of covid-19 epidemics covid-19: a massive stress test with many unexpected opportunities (for data science). harvard data science review high temperature and high humidity reduce the transmission of covid-19 new age. warm weather may buy bangladesh some time to fight coronavirus could tb vaccine protect medics from covid-19? the daily star. 375 expatriate bangladeshis died of covid-19, symptoms in saudi arabia road transport association leaders want to operate vehicles at full capacity bangladesh plans to reopen schools combining online acknowledgements. the authors acknowledge european centre for disease prevention and control, who regularly published worldwide covid-19 data. we convey thanks to the media for the situational analysis. we extend our appreciation to the three anonymous reviewers who have significantly improved the focus and clarity of the paper. key: cord-297326-n0fpu8s3 authors: álvarez, e.; donado-campos, j.; morilla, f. title: new coronavirus outbreak. lessons learned from the severe acute respiratory syndrome epidemic date: 2015-01-16 journal: epidemiol infect doi: 10.1017/s095026881400377x sha: doc_id: 297326 cord_uid: n0fpu8s3 system dynamics approach offers great potential for addressing how intervention policies can affect the spread of emerging infectious diseases in complex and highly networked systems. here, we develop a model that explains the severe acute respiratory syndrome coronavirus (sars-cov) epidemic that occurred in hong kong in 2003. the dynamic model developed with system dynamics methodology included 23 variables (five states, four flows, eight auxiliary variables, six parameters), five differential equations and 12 algebraic equations. the parameters were optimized following an iterative process of simulation to fit the real data from the epidemics. univariate and multivariate sensitivity analyses were performed to determine the reliability of the model. in addition, we discuss how further testing using this model can inform community interventions to reduce the risk in current and future outbreaks, such as the recently middle east respiratory syndrome coronavirus (mers-cov) epidemic. middle east respiratory syndrome (mers), is a respiratory illness caused by a novel coronavirus (cov) [1] . the disease was reported for the first time in saudi arabia in june 2012 and spread to several countries in africa, asia, americas and europe [2, 3] . the capability of human-to-human transmission has been observed in at least four hospital settings [4] [5] [6] [7] . significantly, mers-cov shares certain similarities with the severe acute respiratory syndrome (sars)-cov that produced a global epidemic with more than 8000 human cases in 2002-2003 [8, 9] . first, a number of patients infected with both viruses developed an acute respiratory disease that in some cases resulted in death [7, 9] . in this sense, mers-cov appears to be highly pathogenic with an estimated case-fatality rate of around 50%, although this might be an overestimation as many infected patients may not have sought hospital assistance [2] . second, both mers-cov and sars-cov, belong to the genus betacoronavirus and are closely related to coronaviruses isolated from bats [10] [11] [12] . this strongly suggests that mers-cov and sars-cov may have been transmitted from bats to humans through intermediate species (e.g. camels for mers-cov and civet cats for sars-cov). third, since they are new emerging viruses, there are no effective vaccines or antiviral treatments. the sars epidemic is a clear example of how a networked health system can respond to a new threat to human health. one of the best-characterized outbreaks during the sars epidemic was in hong kong in 2003 where there were 1755 confirmed cases with 299 deaths (who; http://www.who.int/csr/sars/country/en/index. html). the outbreak began in mid-february caused by an infected person who travelled from guangdong to hong kong [13] . an important fact in the generation of a model of the outbreak is that the hong kong health authorities quickly implemented contagion control procedures [14] . in general, two interventions were introduced to prevent the spread of sars-cov. the first was implementation of quarantine measures to isolate healthy people who had been in contact with infected people and therefore potentially in contact with the virus; isolating those that could be infected and asymptomatic during the incubation period and isolating and treating patients who had developed the disease. the other intervention was the application of protective measures by healthy people who were in contact with infected people to avoid becoming infected, such as respiratory protection for healthcare workers and daily disinfection of the environment of affected rooms [14] . these control interventions were implemented progressively in the hong kong special administrative region from mid-march to late april [15] . the application of these procedures allowed the rapid control of the outbreak in the subsequent months. in this sense, it is estimated that the epidemic in hong kong ended in late june. system dynamics has been proved to be a powerful instrument for analysing social, economic, ecological and biological systems [16] . in addition, disease epidemiology has been studied using this approach, whereby system dynamics offers the practical application of concepts by computerized models that allow the systematic test of different scenarios and alternative policies [17] [18] [19] . in this work, we performed a modelling of the hong kong sars-cov outbreak using system dynamics. the developed model contains five states, four flows, eight auxiliary variables and six parameters that interact through five differential and 12 algebraic equations. the parameters of the model were optimized following an iterative process of simulation to obtain a model that largely fits the data available to the epidemic. moreover, the credibility of our model and its parameters are supported by both univariate and multivariate sensitivity analyses. the model reproduces how the implementation of control measures was effective in preventing the spread of infection to the rest of the population. basically, these measures result in a sustained reduction in the frequency of contacts. at present, the application of similar measures for infection containment can help to prevent the spread of new emerging epidemics, such as the outbreak caused by mers-cov. data on the total population of hong kong in 2003 was obtained from statistics of the census and statistic department, the government of the hong kong special administrative region (http://www.censtatd.gov.hk/home/). data on cumulative cases, deaths and recoveries during the sars epidemics in hong kong was obtained from global alert and response databases of the who (http://www.who. int/csr/sars/country/en/index.html). the model was developed following the four-step sequence proposed by system dynamics methodology [16] . first, the real data from the hong kong sars epidemics ( fig. 1 ) together with other evidences and our professional experience were used to create a mental modelling of the reality of the outbreak. second, the model structure that is able to explain the evolution of the epidemics was represented as a forrester diagram (fig. 2) . third, the outbreak was mathematically modelled as a continuous dynamic process represented by a set of differential and algebraic equations (tables 1-3) . finally, the model was optimized to fit the real data from figure 1 . a dynamic compartmental model provides a framework for the study of transport between different compartments of a system, i.e. well known epidemiological compartmental models [20] . to explain how the protective measures taken by the government of hong kong allowed the rapid control of the epidemic we consider a dynamic model with five compartments (states) and four transitions (physical flows) between them. this model is based on two assumptions. first, the individuals are classified into five subgroups (susceptible, latent, infected, recovered, dead). although, the last subgroup is not strictly needed, it is used to keep an account of those dead. second, every day there is a different number of people: who are infected without symptoms of the disease (incidence); who develop signs and symptoms of the illness (sick per day); who recover from the disease (daily recovered), and who die (daily deaths) as consequence of the outbreak. the model structure built with vensim software (ventana systems inc., usa) ( fig. 2) contains the five states and the four flows mentioned above and also eight auxiliary variables (inside circles) and six parameters (in bold). these elements are linked by the physical flows (double line with arrow) and by the information transmissions (single line with arrow) according to the mathematical model represented by the set of differential and algebraic equations (tables 1-3 ). the five differential equations in table 1 establish the mass balance (inflows minus outflows) in the compartments, and as such, they describe the changes in the number of people (stocks) in the five subgroups. the four algebraic equations of table 2 express that the physical transitions depend directly from the stock in the compartment of origin and indirectly from other stocks and parameters through the corresponding auxiliary variables. these equations involve three stocks (susceptible, latent, infected), three auxiliary variables (prevalence, contagion rate, recovery rate) and three parameters (incubation period, case fatality, disease duration). for instance, the incidence flow is proportional to the number of susceptible people, the prevalence and the contagion rate. the contagion rate or transmission coefficient (1) prevalence recovery rate recovery rate(t) = (1 − case fatality)/(disease duration) contagion rate contagion rate(t) = frequency of contacts(t) × infectivity cumulative cases cumulative cases(t) = dead(t) + recovered(t) + infected(t) people (6) cumulative attack rate cumulative attack rate (t) = cumulative cases(t)/population(t) dimensionless (7) frequency of contacts frequency of contacts(t) = daily contacts /(1 + (cumulative attack rate/ threshold of cumulative attack rate) 3 ) basic reproductive number(t) = contagion rate(t)/recovery rate(t) dimensionless theoretically depends on the number of contacts per unit time and the probability of effective contact, i.e. the probability that a contact between an infectious source and susceptible host results in a successful transfer whereby the susceptible host becomes infected. the first three algebraic equations of table 3 express the auxiliary variables (prevalence, contagion rate, recovery rate) used in the equations of table 2 . in turn, these depend on other auxiliary variables, stocks and parameters. for instance, prevalence is defined as the ratio between number of infected people and the total population [ table 3 , equation (1)]. the recovery rate depends on the illness duration and case fatality [ table 3 , equation (2)]. equation (3) in table 3 expresses the contagion rate which is directly dependent on the auxiliary variable 'frequency of contacts' and the parameter 'infectivity'. one of the key variables in the model is frequency of contacts because it tries to reproduce the control measures carried out by the hong kong government to control the sars outbreak. we assumed that: (1) the control measures gave, as consequence, a marked reduction of daily contacts in hong kong; (2) the control measures were based on the cumulative attack rate, measured as the ratio between cumulative cases and total population [ table 3 , equation (6)]. therefore, we decided to use the frequency of contacts as the daily contacts modulated by the repression hill function (fig. 3 ), in accordance with the equation (7) of table 3 . despite the complexity level of biological systems several cases have been modelled using the hill function, in order to simulate repressor activities of enzymatic reactions and the regulation mechanisms of several transcription factors [21] . in our model this repressor function allows the establishment of the relationship between frequency of contacts and the cumulative attack rate. note that the cumulative attack rate is used as an active repressor, so halfmaximal repression occurs when the cumulative attack rate is equal to the threshold, and almost total repression occurs when this cumulative attack rate is double the threshold. equation (8) in table 3 is used to report on the basic reproductive number (r 0 ), which is defined as the expected number of secondary infectious cases generated by an average infectious case in an entirely susceptible population [20] . in our model this number is calculated as the ratio between the contagion rate and the recovery rate. if r 0 < 1, then the infected individuals in the total population fail to replace themselves, and the disease does not spread. however, if r 0 > 1, the number of cases generally increases over time and the disease spreads. one of the most challenging issues in system dynamics modelling is to establish the value of the model parameters. the parameters can be estimated taking advantage of the known information available in the literature and can be optimized by means of an iterative process of simulation. using this approach, we set values for the six parameters of our model which are summarized in table 4 . infectivity expresses the ability of the pathogen to penetrate, survive and multiply in the host and it is measured through the secondary attack rate which is defined as the probability that infection occurs in susceptible persons within a reasonable incubation period following known contact with an infectious person or an infectious source. epidemiological studies in (7) of table 3 ]. the dotted line corresponds to the threshold repression function (when n tends to infinity). the representation is normalized to the threshold of cumulative attack rate on the abscissa axis and the value of the discontinuity (daily contacts) on the ordinate axis. singapore showed that 80% of individuals infected with sars-cov did not cause secondary infections, suggesting low infectivity [22] . after the optimization process, the final value set for infectivity was 0·022, which is in agreement with a previous work examining the probability of transmission for sars-cov [23] . the incubation period of the disease, during which time the individual is asymptomatic appears to be between 2 and 10 days with an average value of 5·3 days [24] . similarly, although the duration of the disease is variable depending on the case, the average of the most severe and the mildest cases is 26 days [24] . during this time period the virus can be transmitted to other people. several studies have shown that the case fatality of sars-cov was also variable in the 2003 outbreak depending on different factors. it has been estimated that the mean case fatality for sars in hong kong was around 0·17 [25] . social contact patterns have been shown to be highly relevant on the transmission dynamics of respiratory infections such as measles, rubella and influenza [26] [27] [28] . quantifying this parameter is critical for estimating the impact of such infections, for designing and targeting preventive interventions, and for modelling their impact. in our model the daily contacts parameter, indicating the mean number of contacts of a person in 1 day, was set to 16·8 after the optimization process, which is in agreement with previous studies quantifying these social mixing patterns [29, 30] . the parameter 'threshold of cumulative attack rate' is critical in our model since it allows setting the value of the cumulative attack rate at which control measures were established by the hong kong authorities. this parameter was finally set at 7·8 × 10 −5 . therefore, according to the hill function of figure 3 , for any number of the cumulative attack rate below the 78 cases by million, the frequency of contacts will be greater than 8·4 (half of the daily contacts), and in the opposite case the frequency of contacts will be markedly reduced. as explained before, our dynamic model was subjected to successive rounds of simulation and optimization in order to fine-tune the parameters of the model. in all these simulations we assumed that the hong kong epidemic originated from a traveller from guangdong in china [13] , and at the beginning of the outbreak the entire population of hong kong was susceptible to sars-cov infection. these assumptions are represented by the initial values in the five stocks (table 1) . moreover, based on the reported data, the simulation period was set to 146 days with a time step of 1 day. parameters of system dynamics models are subject to uncertainty. sensitivity analyses were conducted to provide insight into how uncertainty in the parameters affects the model outputs and which parameters tend to drive these variations. in this task it is essential to define the probabilistic distribution patterns of the model parameters, which are shown in table 5 , together with the references that support these patterns. the most influential parameters were estimated by univariate analyses, in which changes in the model output were studied after disturbance in each parameter value independently. in complex models, univariate sensitivity analysis can be insufficient for a comprehensive study of the model. simultaneous fluctuations in the value of more than one parameter may create an unexpected output change due to nonlinear relationships in different model components. thus, to test the influence of simultaneous changes in the model parameters, the univariate analyses were followed by monte-carlo multivariate sensitivity analysis, in which the values of the six parameters were altered at the same time. modelling process, simulations and sensitivity analyses were performed using vensim dss software v. 5.7a (ventana systems). figure 4 shows a graphic comparison between the simulation results using the parameters of table 4 and the real data of the hong kong outbreak. we compared only the variables of the model for which records were found in the public databases, which are the same six variables shown in figure 1 . the simulation output for the variable 'sick per day' fit the data reported by the hong kong authorities (fig. 4a) , suggesting that the model was able to reproduce the epidemic curve. we observed that the number of new cases per day obtained in the simulation grew during the first 46 days. from this time, the number of new infections gradually fell to values <1 at later stages of the outbreak. as a consequence, the auxiliary variable that stores the cumulative sars cases showed a characteristic sigmoidal growth (fig. 4b) , consistent with the real data. we observed that the number of sars cases grew from one at the beginning of the epidemic to around 1800 at the final stage of the outbreak, similar to the 1755 cases reported by the authorities. moreover, we observed a high fit between the model predictions and the real data. as expected, the number of deaths and recovered people in the simulation also grew with a sigmoidal shape to reach values similar to those found in the public databases (fig. 4c, d) . however, we observed a partial fit of these stock variables to the data reported during the outbreak. these slight mismatches were also observed in the flow variables for recovered and daily deaths (fig. 4e , f) that may be due to delays in reporting of cases by the authorities. in short, looking the simulation results of figure 3 we can conclude that our model is able to reproduce largely the most important indicators of the sars epidemic that occurred in hong kong in 2003. focusing on the evolution of the basic reproduction number (fig. 5) , we note that during the early stages of the epidemic, r 0 >1, which is consistent with the observed disease spread. moreover, after day 30, r 0 starts to drop to values <1, probably due to the implementation of containment measures by the hong kong authorities after the issuance of the first global alert against sars on 12 march 2003. these results are consistent with a previous report showing the basic reproductive numbers for different sars epidemic curves, which supports the notion that our model is able to largely replicate the disease outbreak in hong kong [31] . the results of the univariate sensitivity analysis are shown in figure 6 . we focused our attention on the epidemic wave although the analysis is possible in other variables, as shown in the multivariate analysis of figure 7 . variations in case fatality, threshold of cumulative attack rate and disease duration induced little changes in the epidemic curve (fig. 6-c) , while more extensive alterations in the epidemic wave were observed after changes in infectivity, daily contacts and incubation period (fig. 6d-f) . variations in the case-fatality parameter do not alter the output of the variable sick per day (fig. 6a) , although other variables from the model such as recovered and dead are highly impacted (data not shown). small changes in the shape and the maximum of the curve are observed after modification of the parameters 'threshold of cumulative attack rate' and 'disease duration' (fig. 6b, c) . by contrast, changes in infectivity, daily contacts and to a lesser extent in incubation period significantly alter both the position and height of the maximum of the epidemic curve (fig. 6d-f ) . the results of the multivariate sensitivity analysis are shown in figure 7 , we analysed the output of four variables: sick per day, infected, recovered, and dead ( fig. 7a-d, respectively) . variations in the model parameters clearly change the shape of the epidemic curve, altering both the position and the height of the maximum of this variable (fig. 7a) . similarly, the output of the variables infected and recovered is highly impacted by the changes in parameters carried out in the multivariate sensitivity analysis (fig. 7b, c) . the alterations of variable outputs are clearly exemplified by the variable 'dead' (fig. 7d) . certain changes in the parameters of the model can explain an increase in the total number deaths during the epidemic rising from about 300 to about 700. the observed variations in model output when the value of the parameters is changed support the idea that this model might be able to simulate different scenarios and epidemic conditions. system dynamics modelling has been successfully applied to study complex public health issues such as the design of optimal policies in healthcare [32] , the impact of public health intervention in different situations [33, 34] , and to study disease epidemiology [35] [36] [37] . in the latter, system dynamics technology has become a powerful tool to understand and predict the impact of infectious diseases. epidemiological models can help health authorities to make recommendations regarding intervention to fight the spread of directly transmissible pathogens, especially when empirical data is limited. in this sense, mathematical models have been previously used to advise health policies against diseases such as pandemic influenza [38, 39] and sars [31, 40, 41] . several models studying sars transmission and interventions have been published. these are detailed hybrid stochastic and compartmental models that successfully explain the behaviour of the epidemic [14, 31, 42] . here, trying to follow ockham's razor principle, we have built a simpler deterministic model, which is also able to reproduce the behaviour of the epidemic based on its natural history and the intervention measures taken in hong kong. the use of a less complicated model could be helpful in understanding the disease epidemic and also facilitating its reuse under other conditions. the epidemiological models depend on the consistency of the chosen parameters. therefore, the accurate quantification of these parameters is critical to estimate the path of a disease, to predict the impact of possible interventions, and to inform planning and decision making. here, we have combined reported information from the sars epidemic with an iterative optimization process to set the final values for the model parameters. under these conditions, the model output fits the epidemic curve observed in the hong kong sars-cov outbreak (fig. 4) . of the factors that influence the dynamics of infectious diseases, the person-to-person contact pattern has been shown to be essential in disease spread [43] . our model takes this essential factor into account through the auxiliary variable 'frequency of contacts', which is dependent on the auxiliary variable 'cumulative attack rate' and the parameters 'daily contacts' and 'threshold of cumulative attack rate' (fig. 2) . a previous work showed that mixing patterns and contact characteristics were remarkably similar across the different european countries analysed in that table 5 . the solid black line represents the simulation output and the grey area represents the 95% confidence bounds. study even though the average number of contacts recorded differed. interestingly, the authors suggests that the results may well be applicable to other countries with similar social structures, and that the initial epidemic phase of an emerging infection in susceptible populations, such as sars was in 2003, is likely to be very similar [30] . during the sars outbreak, health authorities, hospitals, and the overall population progressively implemented quarantine and protection policies to prevent the transmission of the disease [15] . with this in mind, we made the assumption that the intervention of the health authorities caused a decrease in the frequency of contacts, which in turn led a decrease in the rate of contagion. to mimic this event in our model, the number of daily contacts is regulated by means of a repression hill function, which has been used to simulate repressor activities in complex biological systems [21] . the basic reproductive number is a key epidemiological variable that characterizes the potential of a disease to spread. several works have estimated that prior to the first global alert the basic reproductive number for sars was >1, correlating with an exponential increase in the number of cases. however, the implementation of effective control measures, such as quarantine, isolation, and strict hygiene practice in hospitals led a sudden decrease in r 0 [14, 31] . the fact that, in our model, r 0 drops to values <1 around the date of the global alert reinforces the idea that a decrease in the frequency of contacts is able to effectively simulate the effects of the control measures established in the first stages of the epidemic. it is important to note that estimations of r 0 in previous publications are considerably lower (around 2-4) than ours, which is almost 12 at the beginning of the outbreak [14, 31, 42] . nevertheless, the estimated value of r 0 also differs in these works and the credible intervals surrounding these deterministic estimations were wide, reaching superior values of almost 8. this high variability can be explained in part by the superspreading events that occurred in sars epidemics. superspreading is an unusual situation, in which a single individual directly infects a large number of other people that has a large influence on the early course of the epidemic [42] . interestingly, the fact that our model does not explicitly account these events could partially explain the very high estimated value of r 0 at the beginning of the outbreak. the reliability of the model parameters is supported by univariate and multivariate sensitivity analyses (figs 6 and 7) . furthermore, sensitivity analysis is a powerful tool to analyse the influence of certain decision making in the evolution of the epidemic. taken together, these findings strongly suggest that table 5 . the solid black line represents the simulation output and the grey area represents the 95% confidence bounds. our model, together with other system dynamics models can be used by epidemiologists to investigate the likely consequences of future re-emergences of sars-cov based on analysis of the previous known epidemics. in addition, by adapting the key parameters of these models or with a little change in the model structures, they can be used to face emerging outbreaks of infectious diseases, such as the recent mers-cov epidemic. in this regard, there are several similarities and differences which should be taken into account when using this model. both sars-cov and mers-cov may cause severe respiratory failure, extrapulmonary features such as diarrhoea and also mild or asymptomatic cases. in contrast with sars, mers has lower human-to-human transmission potential, affects predominantly older people with more comorbid illness and has a higher case-fatality rate [44] . these factors would affect the output of the model variables (e.g. epidemic curve, cumulative cases, number of dead, etc.). middle east respiratory syndrome coronavirus (mers-cov): announcement of the coronavirus study group update: severe respiratory illness associated with middle east respiratory syndrome coronavirus (mers-cov) -worldwide isolation of a novel coronavirus from a man with pneumonia in saudi arabia evidence of person-to-person transmission within a family cluster of novel coronavirus infections hospital outbreak of middle east respiratory syndrome coronavirus first cases of middle east respiratory syndrome coronavirus (mers-cov) investigations and implications for the prevention of human-to-human transmission clinical features and virological analysis of a case of middle east respiratory syndrome coronavirus infection the genome sequence of the sars-associated coronavirus coronavirus as a possible cause of severe acute respiratory syndrome human betacoronavirus 2c emc/ 2012-related viruses in bats, ghana and europe full-genome deep sequencing and phylogenetic analysis of novel human betacoronavirus bats are natural reservoirs of sars-like coronaviruses update: outbreak of severe acute respiratory syndrome-worldwide transmission dynamics of the etiological agent of sars in hong kong: impact of public health interventions monitoring the severe acute respiratory syndrome epidemic and assessing effectiveness of interventions in hong kong special administrative region business dynamics: systems thinking and modeling for a complex world system dynamics modeling for public health: background and opportunities understanding diabetes population dynamics through simulation modeling and experimentation new data and tools for integrating discrete and continuous population modeling strategies infectious diseases of humans: dynamics and control modeling and analysis of the dynamic behavior of the xlnr regulon in aspergillus niger severe acute respiratory syndrome-singapore estimating variability in the transmission of severe acute respiratory syndrome to household contacts in hong kong bayesian modelling of an epidemic of severe acute respiratory syndrome the mixed trunsored model with applications to sars measles in england and wales-i: an analysis of factors underlying seasonal patterns quantitative investigations of different vaccination policies for the control of congenital rubella syndrome (crs) in the united kingdom age-related changes in the rate of disease transmission: implications for the design of vaccination programmes who mixes with whom? a method to determine the contact patterns of adults that may lead to the spread of airborne infections social contacts and mixing patterns relevant to the spread of infectious diseases different epidemic curves for severe acute respiratory syndrome reveal similar impacts of control measures using system dynamics to help develop and implement policies and programmes in health care in england work disability related to musculoskeletal pain: a system dynamics approach evaluating hmo policies with a computer simulation model prevention and rehabilitation as a means of cost containment: the example of myocardial infarction modelling the epidemiological consequences of hiv infection and aids: a contribution from operational research model based scenarios for the epidemiology of hiv/aids: the consequences of highly active antiretroviral therapy containing pandemic influenza at the source cost-effective strategies for mitigating a future influenza pandemic with h1n1 2009 characteristics real-time estimates in early detection of sars entry screening for severe acute respiratory syndrome (sars) or influenza: policy evaluation transmission dynamics and control of severe acute respiratory syndrome mathematical models of infectious disease transmission severe acute respiratory syndrome vs. the middle east respiratory syndrome this research was supported by an intramural contract from universidad autónoma de madrid awarded to e. álvarez. the institutional grant awarded to the centro de biología molecular 'severo ochoa' by the fundación ramón areces is also acknowledged. none. key: cord-292502-m76rne1l authors: cheema, s.; ameduri, m.; abraham, a.; doraiswamy, s.; mamtani, r. title: the covid-19 pandemic: the public health reality date: 2020-09-22 journal: epidemiol infect doi: 10.1017/s0950268820002216 sha: doc_id: 292502 cord_uid: m76rne1l the coronavirus disease (covid-19), while mild in most cases, has nevertheless caused significant mortality. the measures adopted in most countries to contain it have led to colossal social and economic disruptions, which will impact the mediumand long-term health outcomes for many communities. in this paper, we deliberate on the reality and facts surrounding the disease. for comparison, we present data from past pandemics, some of which claimed more lives than covid-19. mortality data on road traffic crashes and other non-communicable diseases, which cause more deaths each year than covid-19 has so far, is also provided. the indirect, serious health and social effects are briefly discussed. we also deliberate on how misinformation, confusion stemming from contrasting expert statements, and lack of international coordination may have influenced the public perception of the illness and increased fear and uncertainty. with pandemics and similar problems likely to re-occur, we call for evidence-based decisions, the restoration of responsible journalism and communication built on a solid scientific foundation. the number of confirmed infections with sars-cov-2 has reached 20 million worldwide, and mortality from covid-19 is estimated to be above 850 000 [1] . all the evidence thus far available quite clearly shows that those at highest risk of a severe illness and death are the elderly, individuals with existing co-morbidities and the immunocompromised. undeniably, the covid-19 pandemic has resulted in loss of human life; it has wreaked havoc on healthcare systems worldwide, highlighting inequities in healthcare availability and access; it has resulted in drastic public health measures in most countries of the world. low-and middle-income nations with weak health systems, dwindling economies, high population density, a high reliance on informal employment, poor technological infrastructure and the double burden of non-communicable and communicable disease are, in particular, more vulnerable to the covid-19 challenge than high-income nations. as additional information about the infection and its effects becomes increasingly available, a number of questions which require an explanation arise. while these questions might have been premature a few months ago when very little was known about the epidemiology of the infection, in this commentary we argue that they are now very timely and that it is imperative these questions be addressed. the questions we specifically explore are: how serious is the covid-19 pandemic? how does it compare with the death burden from other causes? what have been the indirect health and social effects of the covid-19 pandemic? we also raise questions surrounding misinformation and its negative consequences on health. in exploring these questions and seeking possible answers, we first present data in two parts: (a) epidemiology of covid-19 and (b) comparison of covid-19 mortality with mortality from previous pandemics and other causes (for comparison, at the time of writing this paper, the total number of worldwide documented cases and deaths are 27 510 526 and 897 231, respectively) [1] . subsequently, we summarise the indirect repercussions of the covid-19 pandemic on non-communicable diseases, economy and lives of people. in the conclusion, we offer a few comments, share thoughts and raise some questions to help open a debate. based on a review of recent covid-19 literature, it is clear that the disease is minor in most cases [2, 3] . the estimated infection fatality rate is in the range of 0.66-1.33% [4, 5] . the most recent systematic review and meta-analysis found a pooled infection fatality of covid-19 to be around 1% among studies with a low risk of bias (meyerowitz-katz and merone, 2020, unpublished). the covid-19 case fatality rate, in principle an indicator of the virulence of the virus and severity of disease, has been a subject of debate. we now know that this rate may not accurately reflect the true infection fatality rate for a variety of reasons, examples of which include inadequate testing, the high number of mild/asymptomatic cases and failure to include those cases in computing the final rate and the country-specific methods of attributing deaths to covid-19. a number of recent studies, primarily in the usa and in spain, which used antibody testing of population samples indicate that the number of undocumented infections is significantly high. these undocumented infections are often not included in computing the published case fatality rates. while the epidemiological implications of these results remain uncertain, they nevertheless strongly suggest that the infection fatality rate is much lower than the currently reported crude case fatality rate of 3.67% [4] . data are becoming available on the number of deaths per million population in the world health organization (who) weekly epidemiological reports. as of 6 september 2020, the who reported 855 deaths per million in belgium, 612 in the uk and 564 deaths per million in the usa [6] . this may be a truer reflection of the severity of covid-19. we cannot and should not understate the severe disease paradigm in those at higher risk, which includes elderly individuals and those with underlying chronic conditions such as obesity, diabetes, heart disease, cancer, chronic lung conditions and an immunocompromised status. additionally, clinical presentation characterised by underlying pathological changes such as thromboembolism, cytokine release and inflammatory syndrome resulting in damage to the lungs, cardiovascular system, liver, kidneys, pancreas and nervous system, have been noted and described [7] . here, we present data that pose questions on the magnitude of attention that the covid-19 pandemic has garnered compared to other public health issues that are in dire need of prevention and response. table 1 compares the mortality of covid-19 with past pandemics of the 20th and 21st centuries. the mortality rate ratios (between past pandemics and covid-19) ranged from 0.2 times (for the lower estimate of the 2009 'swine flu' pandemic) to over 483 times (the upper estimate of the 1918 'spanish' flu pandemic) that of covid-19, after adjusting for population size. while coronavirus infection and death rates continue to escalate in some communities and decline in others, most experts agree that covid-19 continues to present a significant risk especially to the elderly and those with chronic conditions. it should be emphasised that the other causes of death during the covid-19 pandemic cannot be ignored. according to the institute for health metrics and evaluation (ihme), noncommunicable diseases account for over 41 million deaths globally, while communicable and nutritional diseases claim over 10 million lives [17] . of the latter, 8.1 million deaths were from hiv/aids, tuberculosis, enteric infections, measles and other communicable diseases, most of which are preventable or effectively managed [17] . in 2017, there were 219 million cases of malaria (95% confidence interval (ci): 203-262 million) worldwide, causing an estimated 435 000 deaths [18] . furthermore, we observe that deaths due to some acute and largely preventable causes far exceed covid-19-related deaths. ihme 2017 data on mortality suggest that deaths due to injuries exceed those of covid-19, as of 8 september 2020 [17] . road fatalities, including motor vehicles, cyclists and pedestrians, account for the largest proportion of these, at over 1.2 million. over 50% of injury-related fatalities and more than 80% of communicable and nutritional disease-related fatalities occur in low-income and low-middle-income countries. also, the who estimated that in 2019, iatrogenic or medical errors caused 2.6 million deaths in the lower-and middle-income countries alone [19] . these figures demonstrate that there are other concurrent problems causing distressingly high fatality rates that should not be overlooked as we continue to battle the covid-19 pandemic. while mortality is an important measure to ascertain the seriousness of covid-19, its indirect serious health, social and financial consequences cannot be ignored. the presented data also suggest that the world today may be facing bigger public health challenges than covid-19. is the world's reaction to the pandemic in terms of lockdown and travel restrictions disproportionate? we express our concern on the impact that these prevention measures have had, particularly on the mental health and livelihood of the poor and the most vulnerable populations. more importantly, the current scenario risks compromising the physical, mental and social health of individuals and communities [20] . there are reports that persons with non-communicable diseases are failing to seek timely care due to fear of breaking lockdown rules, the threat of acquiring covid-19 during visits to healthcare facilities, and the choice made by hospitals to treat emergencies only [21] . the risk of adverse health effects due to postponement of routine and elective care along with the severe mental stress and depression caused by this largely unprecedented situation is of grave concern. isolation, unemployment and loss of income may further compound the misery of already lonely individuals and families leading to a rise in self-harm and suicidal ideation, gender-based and domestic violence and the risk of substance use [22] . the evidence of the dramatic economic impact of the measures undertaken in many countries to fight the spread of the disease is apparent. for example, in the usa, unemployment is at a record high and the economy is tumbling. nationwide, women, people of colour and the young are affected the most [23] . the loss of income is likely to result in an increase of adverse health outcomes for many of the individuals affected, and the overall economic crisis will negatively impact the ability of entire countries to provide effective healthcare to their citizens. for individuals in low-and middle-income nations, loss of income, separation from loved ones and social isolation may be legitimately viewed as a bigger threat to long-term survival than the doom and gloom associated with the covid-19 pandemic. such a phenomenon has been observed during the economic crises faced by countries prior to the covid-19 pandemic. the financial crisis in greece, for instance, is estimated to have caused an additional 242 deaths per month between september 2008 and december 2013, due to cardiovascular disease, suicide and mental health illness disproportionately affecting women and people older than 65 [24] . job loss during a recession in the usa was associated with significant increases in mortality (hazard ratio: 1.6; 95% ci 1.1-2.3) [25] . in brazil, a middle-income country, an analysis by hone et al. determined that a 1% rise in the unemployment rate was associated with 0.50 increase per 100 000 each quarter in all-cause mortality and that unemployment resulted in 31 415 additional deaths between 2012 and 2017 [26] . hence, we believe that the mortality and disease burden during and after the covid-19 pandemic due to the social and economic consequences of the preventive measures and other factors can be substantially high. in addition to the direct effects on mortality, it is also feared that the economic disruptions could lead to the doubling of malnourished children in africa in the next 6-9 months [27] . in a recent interview with the washington post, mark lowcock, united nations undersecretary general for humanitarian affairs, said, 'there's a huge covid-19 impact which is economic, and that is drowning out the disease itself' [28] . it is hence critical to have an eye on the overall effects of the pandemic both on the short-and long-term. it is hard at this stage to reconstruct the sequence of events leading to the haphazard and incoherent response of most countries to the spread of the pandemic. however, we caution against fearmongering associated with sensational narratives and inappropriate media reporting, which can result in political pressures that global leaders, policymakers, employers and even some healthcare professionals may have been under, along with the initial uncertainties concerning the severity and nature of the disease. sensationalism, confusion stemming from contrasting statements from authority figures and the lack of international coordination have influenced the public perception of the illness, increasing fear and uncertainty. as an example, we cite the hydroxychloroquine saga. the sale of this medication in the usa jumped leaps and bounds with just a mention of its potential benefit from the us president [29] . similarly, the differing recommendations on the use of masks from the who and the us centers for disease control have contributed to the public's confusion [30] . in addition, the pervasive and increasing role that social media play in how people obtain and share information increases the risks of misinformation and confusion. misinformation can imperil the health of public in other ways. in a recent online us survey, it was observed that us adults are engaging in more frequent cleaning and disinfection of their home to prevent sars-cov-2 infection. the study points out that 40% use cleaning agents or disinfectants in an unsafe manner that presents health risks. for example, 19% reported using bleach on food (fruit or vegetables) and 18% reported using cleaning products on their skin [31] . we should neither downplay nor overstate the pandemic risk. those at increased risk of severe disease should receive priority and be effectively managed. from a public health perspective, it is our opinion, that the lack of a timely internationally coordinated evidence-based approach, the inadequate preparedness of health systems and the absence of effective global leadership has driven us to the current health, economic and social disruptions. the lack of control and coordination over who is saying what, how, where and when, can propel misinformation, leading to fragmented decision-making and public confusion. should there not be an agreed upon deontological code to discourage sensational reporting? why are there not globally acceptable guidance statements on commonly used measures such as the use of face masks and chloroquine? the covid-19 pandemic continues to evolve. moving forward and with pandemics likely to re-occur, we call for health decisions to be made on the basis of science and public health evidence. restoration of responsible journalism and communication driven by scientific truth and valid data is of paramount importance. imparting public health education in school, college and community settings to inform learners about health, disease risks and general aspects of public health challenges such as infectious diseases is vital. worldometer database centers for disease control and prevention (cdc) database. available at imperial college london covid-19 response team estimates of the severity of coronavirus disease 2019: a model-based analysis estimating the infection fatality rate among symptomatic covid-19 cases in the united states world health organization database world population history database reassessing the global mortality burden of the 1918 influenza pandemic the 1918 influenza pandemic: insights for the 21st century updating the accounts: global mortality of the 1918-1920 'spanish' influenza pandemic world health organization (who) the 2009 h1n1 influenza outbreak in its historical context novel swine-origin influenza a virus in humans: another pandemic knocking at the door world health organization (who) estimated global mortality associated with the first 12 months of 2009 pandemic influenza a h1n1 virus circulation: a modelling study institute for health metrics and evaluation (ihme) database world health organization (who) who-calls-for-urgent-action-to-reducepatient-harm-in-healthcare patients with chronic illness urgently need integrated physical and psychological care during the covid-19 outbreak covid-19 pandemic will have a longlasting impact on the quality of cirrhosis care the psychological impact of the covid-19 epidemic on college students in china unemployment soars to 14.7%, job losses reach 20.5 million in april as coronavirus pandemic spreads total and causespecific mortality before and after the onset of the greek economic crisis: an interrupted time-series analysis recessions, job loss, and mortality among older us adults effect of economic recession and impact of health and social protection expenditures on adult mortality: a longitudinal analysis of 5565 brazilian municipalities world food program database the nutrition crisis of covid-19 will be even worse than the disease. the washington post association between us administration endorsement of hydroxychloroquine for covid-19 and outpatient prescribing covid-19: what is the evidence for cloth masks? more than 1 in 3 us adults use disinfectants unsafely acknowledgements. we would like to thank ms. danielle jones (dj), lecturer, english as second language, pre-medical education weill cornell medicine-qatar for her english editing services. financial support. this research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors. ethical standards. not applicable. data availability statement. the dataset(s) supporting the conclusions of this article is (are) included within the paper. key: cord-330954-ft14aa2n authors: liu, b. m.; yang, q. q.; zhao, l. y.; xie, w.; si, x. y. title: epidemiological characteristics of covid-19 patients in convalescence period date: 2020-06-03 journal: epidemiol infect doi: 10.1017/s0950268820001181 sha: doc_id: 330954 cord_uid: ft14aa2n this study aimed to investigate the clinical characteristics and to analyse the epidemiological features of coronavirus disease 2019 (covid-19) patients during convalescence. in this study, we enrolled 71 confirmed cases of covid-19 who were discharged from hospital and transferred to isolation wards from 6 february to 26 march 2020. they were all employees of zhongnan hospital of wuhan university or their family members of which three cases were <18 years of age. clinical data were collected and analysed statistically. forty-one cases (41/71, 57.7%) comprised medical faculty, young and middle-aged patients (aged ⩽60 years) accounted for 81.7% (58/71). the average isolation time period for all adult patients was 13.8 ± 6.1 days. during convalescence, rna detection results of 35.2% patients (25/71) turned from negative to positive. the longest rna reversed phase time was 7 days. in all, 52.9% of adult patients (36/68) had no obvious clinical symptoms, and the remaining ones had mild and non-specific clinical symptoms (e.g. cough, sputum, sore throat, disorders of the gastrointestinal tract etc.). chest ct signs in 89.7% of adult patients (61/68) gradually improved, and in the others, the lesions were eventually absorbed and improved after short-term repeated progression. the main chest ct manifestations of adult patients were normal, ggo or fibre streak shadow, and six patients (8.8%) had extrapulmonary manifestations, but there was no significant correlation with rna detection results (r = −0.008, p > 0.05). the drug treatment was mainly symptomatic support therapy, and antibiotics and antiviral drugs were ineffective. it is necessary to re-evaluate the isolation time and standard to terminate isolation for discharged covid-19 patients. severe acute respiratory syndrome coronavirus 2 (sars-cov-2) belongs to β-coronavirus, and humans may present non-specific clinical manifestations such as fever, cough, fatigue and diarrhoea after infection [1] . coronavirus disease 2019 (covid-19) is highly infectious, mainly transmitted by patients infected with sars-cov-2. its incubation period ranges from 1 to 14 days, mostly 3−7 days, and patients in the incubation period were also infectious [2] . according to the 'diagnosis and treatment protocol for covid-19 (trial version 7)' issued by the national health commission of the people's republic of china (prc), patients meeting the discharge criteria of covid-19 should continue to be isolated for 14 days for management and health monitoring [3] . however, the epidemiological characteristics of covid-19 patients in isolation period have not been clarified. in this study, we retrospectively analysed the clinical characteristics and treatment regimens of covid-19 convalescent patients. the study was approved by the medical ethics committee of the zhongnan hospital of wuhan university (ethics number: lyl2020056k), and a written informed consent was waived. we enrolled 71 cases of covid-19 patients in isolation wards from 6 february to 26 march 2020. they were convalescent patients who were transferred to isolation wards after meeting discharge criteria. they were all medical employees of zhongnan hospital of wuhan university or their family members. the admission diagnosis and discharge criteria of covid-19 patients all met the requirements of covid-19 diagnosis and treatment protocol issued by the national health commission of prc [3] . the discharged patients must meet the following criteria: (1) the body temperature returns to normal for more than 3 days; (2) respiratory symptoms improve significantly; (3) pulmonary imaging shows that the acute exudative lesions were significantly absorbed and improved; (4) the sars-cov-2 ribonucleic acid (rna) detection results of two consecutive respiratory specimens are negative (sampling interval should be at least 24 h). according to the internal standard for employees of zhongnan hospital of wuhan university to be released from isolation, the convalescent patients who were consistently negative in five sars-cov-2 rna detections of oropharyngeal swabs (sampling interval should be at least 24 h) could be released. even though the patients had been released, they should continue to be isolated at home for 14 days, during which the antibodies (igm + igg) were detected. the samples for sars-cov-2 nucleic acid detection were taken with oropharyngeal swabs. two plastic rod swabs with polypropylene fibre heads were used to wipe the bilateral pharyngeal tonsils and posterior pharyngeal wall simultaneously. then, the heads of the swabs were immersed in a tube containing 3 ml of virus preservation solution (isotonic salt solution, tissue culture solution or phosphate buffer could also be used), the tails were discarded and the tube cap was tightened. this sampling method followed the laboratory testing technical guidelines for covid-19 (fourth edition) issued by the national health commission of prc (http://www.nhc.gov.cn/jkj/s3577/202002/ 573340613ab243b3a7f61df260551dd4.shtml). general data of patients were collected, including age, gender, underlying disease, medication history, isolation time and frequency of rna detection. we used real-time reverse transcription polymerase chain reaction (rt-pcr) to detect sars-cov-2 in the oropharyngeal swab of respiratory tract specimens. we recorded the clinical manifestations, rna detection results of oropharyngeal swab specimens, sars-cov-2 igm−igg antibodies detection results (data were collected on the 28th day after discharge), chest computed tomography (ct) images and medication of these patients. according to patients' age, they were divided into young and middle-aged group (⩽60 years old) and elderly group (>60 years old). according to rna results in isolation period, patients were divided into the re-detectable positive (rp) group and non-rp (nrp) group. the above subgroups were designed to compare their symptoms, rna detection results, igm−igg antibodies detection results, ct results and treatment. all statistical analyses were processed by spss 22.0 software. continuous variables were expressed as means ± standard deviation (s.d.). the measurement data conforming to normal distribution were compared between groups by t test, while those not conforming to normal distribution were compared between groups by mann−whitney u test. categorical variables were summarised as frequency and percentage, and chi-square test was used for comparison between groups. the correlation between age, rna detection results, frequency of positive rna and symptoms, ct were conducted by spearman's correlation analysis. all the statistical tests were two-sided, and significant differences were considered at p < 0.05. for all 68 adult patients, the average age was 44.3 ± 16.4 years. in all, 80.9% of patients (55/68) were younger than 60 years of age, the majority were female (43/68, 63.2%), 41 cases were clinical first-line medical staff (41/68, 60.3%). there were 14 patients (14/68, 20.6%) complicated with underlying diseases (p < 0.05), and most of them had a related drug use history. the underlying diseases of elderly patients were mostly hypertension, diabetes, coronary heart disease, lacunar infarction, emphysema and that of young and middle-aged patients were mainly obsolete pulmonary tuberculosis. the shortest isolation time was 4 days, the longest was 38 days and the average isolation time was 13.8 ± 6.1 days ( table 2) . overall, 52.9% of the adult patients (36/68) had no obvious clinical symptoms, and the remaining patients had mild clinical symptoms (one patient was transferred to the icu because of worsening condition, and eventually improved after therapy), which had no specificity. adult patients' symptoms included cough in 18 cases (26.5%), diarrhoea in five cases (7.4%), chest distress in four cases (5.9%), expectoration in four cases (5.9%), sore throat in four cases (5.9%), nausea and vomiting in four cases (5.9%), fatigue in three cases (4.4%), eyes discomfort in three cases (4.4%), dizziness in two cases (2.9%), headache in two cases (2.9%) and there was only one case (1.5%) of fever (table 2 ). there was no significant relationship between symptoms and age (r = 0.131), rna detection results (r = 0.230) and the frequency of rp (r = 0.223) (all p > 0.05) ( table 5) . compared with the ⩽60-year-old group, the diarrhoea symptoms of patients in the >60-year-old group were significant (23.1% vs. 3.6%, χ 2 = 5.833, p < 0.05), and other symptoms were not significantly different between the two groups. in comparison with patients in ⩽60-year-old group, in addition to medication for chronic diseases, the application of drugs for digestive system (46.2% vs. 12.7%, χ 2 = 7.598, p < 0.01), sleep improvement (53.8% vs. 9.1%, χ 2 = 18.705, p < 0.001) and expectorants (15.4% vs. 0.0%, χ 2 = 8.718, p < 0.01) in >60-year-old group were significantly different, but antibiotics and antiviral drugs were not statistically significant between the two groups (p > 0.05). there was no statistical difference between the two groups in isolation time and frequency of rna detection, as well as the detection results of igm (15.4% vs. 16.4%, χ 2 = 0.007, p > 0.05) and igg (46.2% vs. 65.5%, χ 2 = 1.659, p > 0.05) antibodies. the results of rna detection of three covid-19 convalescent patients aged below 18 years did not turn positive. the symptoms, medication and isolation time (14.3 ± 2.9 days) for them were not significantly different from those of adult patients (p > 0.05) ( table 2) . in all adult patients, there were 25 patients (25/68, 36.8%) showing rp findings of sars-cov-2 rna in oropharyngeal swab specimens and 43 nrp patients (43/68, 63.2%). compared with nrp adult patients, the isolation time of rp patients increased (15.8 ± 6.0 vs. 12.6 ± 5.9 days, u = 344, p < 0.05) and frequency of rna detection increased (10.2 ± 4.2 vs. 5.6 ± 1.4, u = 78.5, p < 0.001). compared with rp patients, the cough symptom in the nrp adult patients was more the detection results of igm (16.0% vs. 16.3%, χ 2 = 0.001, p > 0.05) and igg (76.0% vs. 53.5%, χ 2 = 3.392, p > 0.05) antibodies were not statistically significant between the two groups. there was no significant difference in age and gender between the two groups (p > 0.05) ( table 2 ). compared with nrp adult patients, ct images showed significantly increased fibre streak shadow in rp patients (28.0% vs. 4.7%, χ 2 = 7.505, p < 0.01), but there was no significant difference in other imaging manifestations between the two groups such as ground-glass opacity (ggo), pleural thickening and pleural effusion (p > 0.05) ( table 3) . all patients used traditional chinese medicine recovery prescription (pinellinae rhizoma praeparatum 9 g, citri reticulatae pericarpium 10 g, codonopsis radix 15 g, astragali radix 30 g, poria 15 g, pogostemonis herba 10 g, amomifructus (added later) 6 g). compared with nrp adult patients, there was no significant difference in usage of antiviral drugs (40.0% vs. 25.6%, χ 2 = 1.540, p = 0.21), antibiotics (40.0% vs. 32.6%, χ 2 = 0.383, p = 0.54), respiratory system drugs (40.0% vs. 37.2%, χ 2 = 0.052, p = 0.82) or digestive system drugs (24.0% vs. 16.3%, χ 2 = 0.609, p = 0.43) in rp patients (p > 0.05) ( table 2) . of the 25 rp patients, 10 patients (40.0%) were male and 15 patients (60.0%) were female, including seven cases with rp findings >2 times and 18 cases with rp findings ⩽2 times, and one patient had eight times of rp findings. compared with patients with rp findings ⩽2 times, the isolation time of patients with rp findings >2 times was prolonged (20.9 ± 6.4 vs. 13.8 ± 4.7 days, u = 25, p < 0.05), the frequency of rna detection increased (15.4 ± 4.2 vs. 8.2 ± 1.6, u = 1.5, p < 0.001) and the age was older (59.6 ± 12.9 vs. 43.0 ± 16.9 yrs, u = 23.5, p < 0.05). there was no significant difference in gender between the two groups (p > 0.05). the detection results of igm (14.3% vs. 16.7%, χ 2 = 0.021, p > 0.05) and igg (85.7% vs. 72.2%, χ 2 = 0.503, p > 0.05) antibodies were not statistically significant between the two groups. in the 25 patients, none of them had fever, and 17 of them had no clinical symptoms. there was no significant difference between the respiratory and cardiovascular symptoms between the two groups (p > 0.05), but the symptoms of diarrhoea in patients with rp findings >2 times were worse than in patients with rp findings ⩽2 times (28.6% vs. 0.0%, χ 2 = 5.590), and the difference was statistically significant (p < 0.05). among the 25 rp patients, six showed normal lung imaging (24.0%), 10 patchy ggo (40.0%), seven fibre streak shadow (28.0%) and four patchy shadows (16.0%). there was no statistically significant difference in ct imaging between the two groups (p > 0.05). compared with patients with rp findings ⩽2 times, the application of drugs for sleep improvement (57.1% vs. 16.7%, χ 2 = 4.096, p < 0.05) and expectorants (28.6% vs. 0.0%, χ 2 = 5.590, p < 0.05) had statistical differences in patients with rp findings >2 times, but there was no statistical difference in other systemic drug applications between the two groups (p > 0.05) ( table 4 ). among the 68 adult patients, the medical staff accounted for 60.3%, of whom after infection, 32 cases (47.1%) entered the convalescence period in february and nine cases ( the clinical symptoms of covid-19 convalescent patients who reached the discharge criteria were mostly mild. asymptomatic patients accounted for more than 50%, and there was only one case of fever. the symptoms were mainly concentrated in the respiratory system and digestive system. it was reported that sars-cov-2 infected with respiratory tract host cells through cells expressing angiotensin-converting enzyme 2 (ace2) receptors [5] , but recent studies found that sars-cov-2 could also be detected in samples such as anal swabs, blood, urine and faeces [6, 7] . compared with the respiratory system, the virus remained in the gastrointestinal system for a longer time, and the clearance of viral rna in faeces was delayed [8] [9] [10] . nasopharyngeal, sputum and faeces were the major shedding routes for sars-cov-2, and virus shedding time in sputum was longer and more stable than nasopharyngeal and faeces. the median durations of virus shedding of them in turn were 12 , 19 (5-37) and 18 (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) days. the viral load in sputum was the highest among the three specimens, followed by nasopharyngeal and faeces were the lowest [11, 12] . another report on three cases of common covid-19 children showed that all patients had rp findings of sars-cov-2 rna detection in faeces specimens within 10 days after discharge, but no positive results were found in the two rna detection in throat swab specimens, and there were no clinical symptoms [13] . in this study, the >60-year-old group showed obvious symptoms of diarrhoea (p < 0.05), but there was no significant correlation between all symptoms with age or rna detection results (table 5 ). it is reasonable to speculate that the virus still persists in some convalescent patients, however, the pathogenicity of the virus is significantly weakened. the symptoms (including fever) cannot be used as the criteria for improvement in covid-19 patients. positive rna in throat swab is the diagnostic criterion for covid-19. the seventh edition of the guidelines issued by the national health commission of rpc recommended that patients required 14 days isolation and health monitoring after discharge [3] , suggesting that the infectivity of patients weakens or disappears. different companies or detection methods have different sensitivities and specificities for rna detection in throat swabs. the positive rate is 30−50%, and positive rna may indicate the persistent existence of the virus. this study showed that after 71 convalescent patients entered isolation wards, 25 patients (35.2%) had rp findings of sars-cov-2 rna in oropharyngeal swab specimens and the longest rna rp time was 7 days. some studies found that covid-19 patients who met the national discharge criteria had rp findings of rna detection during the follow-up observation outside hospital, and the time for rp findings of rna detection ranged from 5 to 13 days after discharge [9, [13] [14] [15] [16] . a case report of four mild-to-moderate covid-19 patients (all were medical staff) showed that all convalescent patients without clinical symptoms presented rp findings of rna detection in throat swab specimens at 5−13 days after discharge [14] . another study found that among 62 covid-19 convalescent medical staff, two patients without clinical symptoms showed rp findings of rna detection in throat swab specimens at 5−6 days after discharge [15] . a case report of seven common covid-19 patients also showed that three convalescent cases had rp findings of rna detection in sputum specimens after discharge, without clinical symptoms, and the time interval of rp findings of rna was 5−7 days [16] . these research studies indicated that rna detection and monitoring for asymptomatic discharged patients should be strengthened, such as collect multiple samples (including nasopharyngeal, sputum and faeces samples) from multiple parts of convalescent patients, and adjust isolation time according to the rna detection results. antibodies are the products of the humoral immune response after viral infection, and specific antibodies to sars-cov-2 were used to determine whether the patient has been recently infected with sars-cov-2. the human immune system was able to produce specific igm and igg antibodies against virus infection. igm is the earliest antibody that appears upon the first immune response. igg is produced later and lasts long. the detection of igm and igg antibodies against sars-cov-2 might be helpful in the diagnosis and epidemiological survey of covid-19, and could be used as an effective supplementary indicator for suspected cases of negative viral rna detection. a combination of rna and igm−igg antibodies detection could provide a more accurate sars-cov-2 infection diagnosis [17] . this study showed that there was no statistically significant difference in the detection results of igm and igg antibodies between different age groups, rp or nrp adult patients (all p > 0.05). it is speculated that for convalescent patients, rna antibody detection may have no practical significance. chest ct examination is an important indicator to judge the severity of covid-19 patients. according to this study, chest ct signs in 89.7% of adult patients showed a trend of gradual absorption and improvement, and in 10.3% of patients, the lesions were eventually absorbed and improved after short-term repeated progression. compared with nrp adult patients, there was no significant increase in ggo, pleural thickening and pleural effusion in rp adult patients, but fibre streak shadow increased significantly. among the 25 rp patients, those with rna positive findings ⩽2 times and >2 times manifested patchy ggo, fibre streak shadow, patchy shadow etc., but there was no statistically significant difference in ct imaging between the two groups. it had been reported that chest imaging of some covid-19 convalescent patients with rna rp findings manifested normal, ggo or fibre streak shadow after discharge [13, 18] . it is speculated that for convalescent patients, ct imaging changes are not affected by whether the rna turns positive or not, however, persistent virus carriers are likely to leave lung fibrosis. intermittent ct follow-up may be used to evaluate recovery of covid-19 patients. there is no clear standard to decide whether patients need medication or not during convalescence. this study showed that 21 adult patients (21/68, 30.9%) were treated with antiviral drugs such as arbidol and oseltamivir, 24 adult patients (24/68, .3%) were treated with antibiotics such as cephalosporins and moxifloxacin, and others were treated with symptomatic drugs. the application of drugs for digestive system, sleep improvement and expectorants in elderly patients increased significantly, however, there was no statistically significant difference in other drugs between different age groups, rp or nrp adult patients, and there was no statistically correlation between drug applications, isolation time and frequency of rp findings. previous studies showed that the virulence and infectivity of sars-cov-2 spontaneously weakened over time, and there were currently no specific antiviral drugs against sars-cov-2. in this study, the rna detection results of 35.2% of patients (25/ 71) in the convalescence period turned from negative to positive. rp patients were prone to present pulmonary fibre streak shadow in ct. the convalescent patients were mainly asymptomatic, and the clinical manifestations were not typical, most of which were cough, sputum, sore throat, gastrointestinal symptoms or other untypical manifestations. most patients' chest ct signs gradually improved, independent of rna rp or not. the drug treatment was mainly symptomatic support therapy, and antibiotics and antiviral drugs were ineffective. it is necessary to re-evaluate the isolation time and standard to terminate isolation for discharged covid-19 patients. molecular and serological investigation of 2019-ncov infected patients: implication of multiple shedding routes clinical characteristics of coronavirus disease 2019 in china national health commission of the prc and national administration of traditional chinese medicine (2020) diagnosis and treatment protocol for covid-19 (trial version 7) clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan, china sars-cov-2 cell entry depends on ace2 and tmprss2 and is blocked by a clinically proven protease inhibitor sars-cov-2 can be detected in urine, blood, anal swabs, and oropharyngeal swabs specimens detectable 2019-ncov viral rna in blood is a strong indicator for the further clinical severity persistence and clearance of viral rna in 2019 novel coronavirus disease rehabilitation patients prolonged viral shedding in feces of pediatric patients with coronavirus disease 2019 prolonged presence of sars-cov-2 viral rna in faecal samples viral kinetics and antibody responses in patients with covid-19. medrxiv quantitative detection and viral load analysis of sars-cov-2 in infected patients detectable sars-cov-2 viral rna in feces of three children during recovery period of covid-19 pneumonia positive rt-pcr test results in patients recovered from covid-19 post-discharge surveillance and positive virus detection in two medical staff recovered from coronavirus disease 2019 (covid-19), china follow-up testing of viral nucleic acid in discharged patients with moderate type of covid-19 evaluation of the auxiliary diagnosis value of antibodies assays for detection of novel coronavirus (sars-cov-2) causing an outbreak of pneumonia (covid-19). medrxiv time course of lung changes on chest ct during recovery from coronavirus disease 2019 (covid-19) acknowledgements. we thank all the patients and their families involved in this study, as well as all the medical staff who are working together fighting against covid-19 in hubei. xiaoyun si had the idea and designed the study and contributed to critical revision of the report. qingqing yang contributed to the statistical analysis. wen xie contributed to collect data. bingman liu, qingqing yang and liangyu zhao contributed to collect data and write the report. all authors contributed to data acquisition, data analysis or data interpretation, and reviewed and approved the final version. financial support. none.conflict of interest. none. key: cord-332097-mrsmwxvo authors: szilagyi, p. g.; blumkin, a.; treanor, j. j.; gallivan, s.; albertin, c.; lofthus, g. k.; schnabel, k. c.; donahue, j. g.; thompson, m. g.; shay, d. k. title: incidence and viral aetiologies of acute respiratory illnesses (aris) in the united states: a population-based study date: 2016-03-02 journal: epidemiol infect doi: 10.1017/s0950268816000315 sha: doc_id: 332097 cord_uid: mrsmwxvo we conducted prospective, community-wide surveillance for acute respiratory illnesses (aris) in rochester, ny and marshfield, wi during a 3-month period in winter 2011. we estimated the incidence of aris in each community, tested for viruses, and determined the proportion of aris associated with healthcare visits. we used a rolling cross-sectional design to sample participants, conducted telephone interviews to assess ari symptoms (defined as a current illness with feverishness or cough within the past 7 days), collected nasal/throat swabs to identify viruses, and extracted healthcare utilization from outpatient/inpatient records. of 6492 individuals, 321 reported an ari within 7 days (4·9% total, 5·7% in rochester, 4·4% in marshfield); swabs were collected from 208 subjects. the cumulative ari incidence for the entire 3-month period was 52% in rochester [95% confidence interval (ci) 42–63] and 35% in marshfield (95% ci 28–42). a specific virus was identified in 39% of specimens: human coronavirus (13% of samples), rhinovirus (12%), rsv (7%), influenza virus (4%), human metapneumovirus (4%), and adenovirus (1%). only 39/200 (20%) had a healthcare visit (2/9 individuals with influenza). ari incidence was ~5% per week during winter. seasonal influenza disease causes substantial morbidity in the united states -200 000 annual hospitalizations, thousands of annual deaths [1] , and many emergency department (ed) and outpatient visits [2] [3] [4] . most recent studies of the burden of influenza disease have focused on healthcare visits; few have examined the community-wide burden of influenza in the general population [5, 6] . because many individuals with influenza disease do not seek medical care but are nevertheless ill enough to miss work or school [7] [8] [9] [10] [11] [12] , the high morbidity documented in studies of influenza-related illness may underestimate the burden of influenza. with vaccination coverage hovering at only 50% (43% in the 2010-2011 season, http://www. cdc.gov/flu/fluvaxview/coverage_1011estimates.htm) [13] despite universal vaccination recommendations [14] , it is important to understand the full impact of this virus at the population level. other viruses also cause acute respiratory illnesses (aris) and it is important to document their population-wide impact. a healthy people 2020 goal is to prevent disease, disability, and death from infectious disease. population-level surveillance is an integral step in assessing morbidity from aris. also vaccines are in development for several viral pathogens including respiratory syncytial virus (rsv) [15] , parainfluenza viruses (pivs) [16] , and human metapneumovirus (hmpv, first identified in the early 2000s) which are significant causes of hospitalizations, ed visits and outpatient visits [2, [17] [18] [19] . longitudinal community-based surveillance studies conducted in the 1960s-1980s found that rsv and pivs accounted for many aris [6, [20] [21] [22] [23] [24] yet few recent studies have assessed their population-wide burden and little is known about the incidence of hmpv in the general population. finally, while rhinoviruses (rvs) [25] , coronaviruses (covs) [26] , and other viruses [2] are known to cause medically attended ari-related visits, studies on their population-wide burden are lacking. such knowledge could guide future vaccine development. we conducted a prospective, population-wide surveillance study in two distinct geographical areas to assess the incidence of aris during the winter months. our objectives were to (1) estimate, at the population level, the incidence of aris across the age spectrum; (2) identify the common viruses currently causing aris throughout the community and their relative frequency; (3) compare the types of viruses causing aris in child vs. adult populations; and (4) estimate the proportion of individuals in the community with aris who make a healthcare visit. the study was approved by institutional review boards at the university of rochester and marshfield clinic research foundation. our prospective, population-wide surveillance study used a previously described rolling cross-sectional (rcs) study design [27] in two communities (rochester, ny and marshfield, wi) in the 2011 winter season to identify individuals with aris (defined as a current illness with feverishness or cough within the past 7 days) and document viruses associated with these aris. the rcs design was first described and used by political scientists in the early 1980s for studies of voter preferences and election results; it consists of a series of cross-sectional samples in which each sample is representative of the source population [28] . a random sample of subjects was selected each week and then interviewed by telephone to identify those with aris and to obtain demographic and disease-specific information. consenting individuals provided nasal and throat swabs which were tested in a research laboratory by real-time polymerase chain reaction (rpcr) for a wide spectrum of viruses. we performed the study in two communities labelled 'rochester' and 'marshfield': (a) monroe county, ny surrounding the city of rochester (population ∼744 000); and (b) the marshfield, wisconsin area (population ∼49 000). the rochester sampling frame comprised 12 primary-care practices (six internal medicine, four paediatric, one family medicine, and one medicine-paediatrics practice) serving 90 245 patients whose age, race/ethnicity, and health insurance mirrored the demographics of monroe county. these practices were from the greater rochester practice-based research network [29] . the marshfield sampling frame consisted of 49 000 residents of the marshfield epidemiologic study area (mesa), a population-based cohort of residents living in 14 zip codes surrounding marshfield; >90% of mesa resident receive their healthcare at marshfield clinic [30] . the study focused on individuals (not households). participants were identified at random from the two source populations. in rochester, we created a denominator of all eligible individuals by merging the practice-level patient databases across 12 practices. we used random digit-dialling to call the primary contact telephone number (mobile or land-line) and called 490-825 people/week (based on previous power calculations), enrolling 118-247 subjects/week. in marshfield, we randomly called 500-800 individuals/week and enrolled 250-400 subjects/week. in both communities, a person was eligible if he/she was aged 56 months as of 1 january, had at least one healthcare encounter in the previous 24 months, and was a resident of monroe county (rochester) or a member of mesa (marshfield) for 512 continuous months prior to 1 january or since birth for those aged <1 year. rochester subjects received a letter from their primarycare physician explaining the study and that an interviewer would telephone them for consent to participate. using protocols developed by the cdc-funded new vaccine surveillance network [2] and influenza vaccine effectiveness (ve) network [31] , the two sites initiated telephone calls when the university laboratories (rochester) or the marshfield laboratories and wisconsin state laboratory of hygiene identified 52 positive influenza specimens within 1 week or one positive specimen/week for 2 consecutive weeks, and stopped calls when laboratories failed to meet this threshold. the surveillance periods were 11 january 2011 to 1 april 2011 (rochester) and 17 january 2011 to 8 april 2011 (marshfield), identical to surveillance in a parallel study of influenza vaccine effectiveness [31] . we used trained telephone interviewers to perform structured interviews. from the randomly selected list of individuals to call each week, interviewers called 6 days/week (monday-saturday, rochester) or 7 days/week (marshfield) during daytime and evening hours. on monday-friday in both communities, study enrollers were given a list of new patients to call and attempted to contact individuals on their assigned day for verbal consent and an interview. up to three calls (morning, afternoon, early evening) were made per day to telephone numbers listed on practice patient lists. if contact was unsuccessful on the assigned day, interviewers continued calls for 2 days, up to three calls per day, with times staggered. individuals from the source population could be eligible more than once in subsequent weeks (96% of those who agreed to participate were enrolled once). telephone interviewers screened subjects for ari symptoms in the identified person (subject or child) over the past 7 days, described the study, and obtained verbal consent for a survey that included a symptom assessment. subjects with ari symptoms in the past 7 days were defined as ari positive; the remainder were defined as ari negative. in rochester, if the sampled person was a child aged 417 years, the interviewer spoke with a parent or guardian. the interviewer conducted a brief computer-assisted interview to determine if the subject had new onset of feverishness or cough within 7 days. telephone procedures in marshfield were the same except: parents were interviewed if subjects were aged 412 years; subjects aged 13-18 years were interviewed but with parental consent. if a potential subject with an ari was identified, the telephone interviewer obtained the date of symptom onset and specific symptoms experienced (using a checklist of symptoms), and asked for verbal consent for an in-person visit to conduct an interview and collect nasal/throat specimens for viral testing; they were compensated $20 for a home visit or $30 for a clinic visit (or community site). subjects with no ari symptoms were compensated with a $5 gift card. in-person visits were conducted within 7 days of the onset of illness symptoms to (1) obtain written consent for specimens and access to subjects' medical records for ari-related information, (2) complete a brief interview to gather additional health-related information (ari symptoms 47 days, and in rochestersites for medical care for the ari), and (3) obtain nasal/throat swabs for testing for viruses. research staff met subjects at their home or a convenient location in rochester or at clinics in marshfield. enrollers collected one nasal and one throat swab for testing. trained abstractors reviewed medical records of ari-positive cases for: (1) ari-related healthcare visits (primary care, speciailty, ed, urgent care, or hospitalization) occurring within 7 days before or after the interview; (2) procedures obtained during visits (chest radiograph, bloodwork, nasopharyngeal or respiratory cultures, rapid antigen testing); and (3) primary diagnosis for outpatient and ed visits and all diagnoses for hospitalizations. nasal/throat specimens were combined in transport media, processed and stored at −80°c. laboratory testing was performed at the university of rochester. total nucleic acid was extracted according to the manufacturer's protocol using the qiaamp viral rna mini kit on a qiacube robotic instrument (qiagen, usa). specimens were tested by taqman array card (tac) methodology (life technologies, usa), which allows simultaneous, singleplex, rpcr to be performed in a 384-well microfluidic card format [32] . we tested specimens for influenza virues a and b, rsv, piv1-4, rv, adenovirus, hmvp, and covs 229e, nl63, oc43 and hku1 by tac assay. all primer and probe oligonucleotide sequences used for pcr assays were prepared and optimized by cdc. tac assays demonstrated high sensitivity (75-95%) for influenza, rsv, piv2-4, rv, and hmpv, and moderate sensitivity (54-56%) for piv1 and adenovirus compared with individual-virus pcrs [33] . the four cov rpcr assays were highly sensitive and specific for the detection of cov 229e, nl63, oc43 and hku1, with positive predictive values in the tac assay of 94%, 97%, 96% and 88%, respectively [34] . we assessed the frequency of each virus and also multiple viruses and tabulated results for all subjects, by age [child (6 months-18·9 years) vs. adult], setting (rochester vs. marshfield), and whether subjects had sought medical care for aris. we used tabular analyses to summarize diagnoses/procedures for subjects with ari-related medical visits. we used pearson's χ 2 tests to compare individual viruses between adult and child ari positive subjects. in instances where expected values were 1-5 we corrected the χ 2 test statistic with the n -1 correction, and used fisher's exact test when expected cell sizes were <1 [35] . we estimated population-wide ari and individual virus cumulative incidence during the 3-month study period by summing the percent of the population that had a new ari (symptoms for 47 days) each week (i.e. weekly incidence). for instance, if we had run a 3-week rcs, and weekly incidences were 2%, 3% and 4%, we would calculate the cumulative incidence as 9%. we used a stratified biased corrected and accelerated bootstrap procedure to calculate 95% confidence intervals (cis) for population-wide proportions, with 1000 replications [36] . incidence of aris during the influenza season ( fig. 1) in rochester, 17 485 telephone calls were made, 4130 calls were completed (i.e. direct contact with a household member), 2683 (65%) agreed to participate in symptom assessment and were enrolled, of whom 2263 (84%) consented for medical chart review for influenza vaccination dates. altogether, 98 nasal/ throat specimens were collected (64% of ari subjects); one specimen was invalid with no ribonucleoprotein (rnp) human cellular control detected. of individuals contacted by telephone, 153 had an ari with feverishness or cough for a weekly (i.e. ari within the previous 6 days) mean ari incidence of 5·7%. the cumulative incidence for ari during the 3-month study period was 52% (95% ci 42-63) as shown in figure 2 . it was 21% when restricted to subjects for whom a viral respiratory pathogen was detected (95% ci [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] . in marshfield, 20 530 telephone calls were made, 5538 calls were completed, 3809 (69%) subjects agreed to participate, 3261 (86%) consented for a medical chart review. altogether 115 nasal/throat specimens were collected (69% of ari subjects); five specimens were invalid with no rnp human cellular control detected. of individuals contacted by telephone, 167 had an ari for a weekly mean ari incidence of 4·4%. the cumulative ari incidence during the 3-month study period was 35% (95% ci 28-42) as shown in figure 2 ; it was 14% when restricted to viruspositive aris (95% ci [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] . across both sites and all ages, 43·1% of individuals had an ari during the 3-month study period (78·5% of children, 33·7% of adults). in rochester, none of the subjects qualified on the basis of fever alone. in marshfield, five of the 46 children aged <19 years and one of the 69 adults qualified on the basis of fever alone. the gender distribution of ari cases was slightly different than the source population (e.g. in rochester 42% of ari cases were female, vs. 44% from the source; in marshfield 61% were female vs. 51%). ari cases were slightly older (e.g. in rochester 75% of ari cases were adults aged >19 years, vs. 70% from the source population; in marshfield 60% of ari cases were adults vs. 77% from the source). viruses associated with aris during the influenza season ( table 1) in 61% of ari cases, no viruses were identified, in 35% a single virus was identified, and in 4% of cases 2-3 viruses were identified. the most common viruses were covs (14% of all ari cases), followed by rvs (12%), rsv (7%), influenza viruses (4%), hmpv (4%), and adenovirus (1%). the distribution of viruses variedsubjects from marshfield experienced more aris from covs and those from rochester had more from rvs. influenza viruses were identified in 1% of aris in children and 7% in adults. rsv was identified in a greater proportion of aris in children, while covs and rvs were identified in a high proportion of aris in both children and adults. at least one virus was identified in 45% of aris in children and 37% in adults; two viruses were identified in 8% of children with aris and 2% of adults. (table 3) out of the 200 subjects with aris and who also agreed to a medical record review (seven from rochester did not agree), 39 (19%) had a healthcare visit, five (2·5%) had an ed visit, and three (1·5%) had a hospitalization, all within 7 days of illness onset. of the 39 with a healthcare visit, 30 (79%) were to primary care. of 84 participating children with aris, healthcare utilization data were available from medical charts for 79 children, and 20 (25%) had a healthcare visit. of 123 adults with aris, healthcare utilization was available for 118 adults; 19 (16%) had a healthcare visit. given the small number of specific viral infections detected, it was not possible to compare healthcare utilization patterns by virus. the single child with an influenza infection did have a healthcare visit; 1/8 adults with influenza had a healthcare visit. this rcs study is one of the few recent epidemiological studies to assess the incidence and causes of aris in the general population. we performed the study during the respiratory season (january to early april), when influenza detections were occurring weekly. our study has several important findings. first, the overall incidence of ari with feverishness or cough per week was 4·9% (ranging from 4·4% in marshfield to 5·7% in rochester), for a cumulative [5, 10, [22] [23] [24] 37] . in these early studies, overall annual ari incidence rates per person-year ranged from 4·5 to 8·0 for children aged <5 years of age and from 1·3 to 6·2 for individuals aged 55 years. because of differences in ari symptom criteria, our focus on wintertime incidence only, and differences in study designs, findings from our study cannot be directly compared with many of these classic studies. however, we estimated that one-third of adults and nearly four-fifths of children in our two communities developed 51 ari during the 3-month respiratory period. a recent household cohort study from michigan [38] found that the mean number of aris per individual over a 6-month period that coincided with our 3-month period was 1·1 for children aged <5 years and 0·6 for adults aged 18-49 yearsthese results are similar to ours, despite the differences in study design. further studies over multiple years would be helpful to better understand the current patterns of ari incidence in the community setting. the weekly (or by extrapolation, 3-month) incidence of aris reported by our subjects in two us communities was in the range of the incidence of aris reported in a telephone survey of residents of australia during the 2008-2009 season, reported in this journal [39] . in that telephone-based survey, 20% of subjects selfreported ari symptoms (based on very similar definitions to ours) within the past 4 weeks. our study extends findings from the australian study by identifying viruses associated with aris and healthcare utilization patterns of individuals with aris. we found that during the 'flu season' of 2010-2011, influenza caused only 1% of aris in children and 7% in adults. these findings contrast with early longitudinal studies [5, 10, [22] [23] [24] 37] and the recent michigan cohort study [38] in which influenza accounted for a higher proportion of aris. data from previously published new vaccine surveillance network studies (which included the rochester site) found that during peak influenza season, influenzarelated aris accounted for up to one-quarter of ed visits in children aged <5 years [3] . the sum of the virus-specific percentages and the percentages of 'no virus identified' range between 104% and 106% because more than one virus present in some subjects. * valid acute respiratory illness specimens. ari, acute respiratory illness; ed, emergency department. * based on medical chart reviews, ari-related healthcare visit within 7 days prior to the telephone interview and 7 days post-interview. the low rates of influenza detection in subjects with ari in our community-based study were likely influenced by several factors. across the united states, the 2010-2011 influenza season was less severe than the 2009-2010 pandemic year or the 2007-2008 seasonal influenza season, but more severe than the 2008-2009 influenza season, with overall hospitalization rates around 19-20/100 000 individuals of all ages [40] . further, despite suboptimal influenza vaccination rates, influenza vaccination may have lowered the relative disease burden from this virus, although the tremendous variability in the timing, intensity, and duration of influenza circulation from year-to-year makes such assessments difficult. vaccination rates were 59% and 41% in the rochester and marshfield study populations, respectively, and it was estimated that vaccine effectiveness in preventing medically attended influenza visits in a four-community study that included both rochester and marshfield was 60% during the same influenza season as assessed in our study [31] . moreover, while our criteria for definition of aris comprised a list of symptoms used by other studies of medically attended ari-related visits, these criteria may have been too broad and some patients labelled with ari in this study may not have had a respiratory virus. for example, we included feverishness as one ari criteria. our data highlight several additional aspects of the current epidemiology of ari. while we confirm the high prevalence of rv, rsv, and piv as aetiological agents of ari, we also documented high rates of aris associated with covs and hmpv, a finding also highlighted by the recent michigan cohort study [38] . our previous studies of medically attended ari visits have noted that these two viruses are associated with ari-related medical visits [19, 26] . together, these findings highlight the potential value of future vaccines against these two viruses. we noted that the pattern of viruses was generally similar for children and adults, although there was a trend toward a greater proportion of aris being attributable to rsv in children than in adults. these mirror findings from cohort studies [5, 10, 22-24, 37, 38] . we found that only one-fifth of ari cases had a healthcare visit, and that four-fifths of these medical visits were to a primary-care physician. a similar pattern was noted in both children and adults, and in both communities. our study strongly suggests that the burden of aris is vastly greater than the burden demonstrated by studies limited to ari-associated healthcare visits. thus, efforts to develop vaccines to reduce aris have the potential for substantial benefit in reducing the disease burden from community-based aris as well as reducing medically attended healthcare visits. future studies should consider collecting additional data on the morbidity of aris that do not result in visits to healthcare providers. our study has several strengths, including rigorously defined protocols across both communities, large numbers of individuals contacted by random sampling, trained staff to obtain nasal/throat specimens and conduct chart reviews, case definitions that mirrored the definitions used in studies of medically attended aris, and detection of viral infections by advanced molecular techniques. however, we caution against over-interpretation of our findings due to several important limitations. first, we studied only two communities, during a single respiratory season over a 3-month period which missed the peak season for piv and rv. geographical and seasonal variability in the viral aetiologies of ari are substantial [2, 7, 10] , for example, it is difficult to draw inferences about ari incidence or virus burden from a single season of surveillance. also, the relative contribution of some viruses might be different had we included a longer study period. second, while our strict definition of aris (current illness with feverishness or cough) likely excluded other causes of stuffiness such as allergies, we may have missed some aris. third, we interviewed only two-thirds of persons contacted, and obtained nasal/throat specimens for only two-thirds of ari cases. we are aware of some selection bias in subjects agreeing to be interviewed. for example, rochester subjects who agreed were more likely than the original sampling frame to be male and adult. while others have noted lower ari reports in men [41] , our ari findings do not vary by gender; also we stratify findings by age group. it is possible that subjects from whom specimens were collected differed by viral aetiology or other characteristics. fourth, despite making >38 000 telephone calls, we were able to collect only 208 nasal/throat swabs from subject with aris. our sample sizes for individual viruses are small. fifth, we may have missed some subjects who would have tested positive for a viral respiratory pathogen because we performed nasal swabs and used tac detection methods rather than nasopharyngeal swabs and more sensitive conventional pcr assays. sixth, we selected subjects with symptoms for 47 days, but it is possible that some were no longer shedding virus, especially if they underestimated their days of symptoms prior to being contacted. finally, we simply summed estimated weekly ari incidence to derive a cumulative incidence, using a rcs design. this method has not yet been validated against more standard methods to calculate cumulative incidence in traditional cohort studies. we conclude that in the 2011 winter respiratory season, about one in 20 individuals within two communities had an ari during any single week, and one-third to one-half had an ari during the winter respiratory season. of those with aris, influenza infection accounted for only 4% of cases, while rsv, piv, rv, cov, and hmpv accounted for the bulk of aris. in individuals with aris, four-fifths did not make a visit to a healthcare provider because of their symptoms. aris due to both influenza and other viruses cause substantial morbidity undetected by the healthcare delivery system. in an era when actively following large cohorts of consenting subjects for disease surveillance purposes is expensive and resource intensive, the use of rcs designs similar to ours may offer a reasonable alternative strategy for conducting surveillance for more common illnesses. new vaccines for several common respiratory viruses may hold promise for further reducing the population-wide disease burden from aris. estimates of deaths associated with seasonal influenza population-based surveillance for hospitalizations associated with respiratory syncytial virus, influenza virus, and parainfluenza viruses among young children the underrecognized burden of influenza in young children influenza-associated hospitalizations in the united states studies of the community and family: acute respiratory illness and infection epidemiology of viral respiratory infections acute respiratory illness in the community. frequency of illness and the agents involved the annual impact of seasonal influenza in the us: measuring disease burden and costs estimates of the us health impact of influenza the seattle virus watch. iv. comparative epidemiologic observations of infections with influenza a and b viruses, 1965-1969, in families with young children patient knowledge and attitudes about antiviral medication and vaccination for influenza in an internal medicine clinic self-reported influenza-like illness during the 2009 h1n1 influenza pandemic -united states cdc. flu vaccination coverage: united states, 2012-13 influenza season prevention and control of influenza with vaccines: recommendations of the advisory committee on immunization practices (acip) understanding respiratory syncytial virus (rsv) vaccine-enhanced disease current status of vaccines for parainfluenza virus infections the burden of respiratory syncytial virus infection in young children parainfluenza virus infection of young children: estimates of the population-based burden of hospitalization the burden of human metapneumovirus infection in young children epidemiology of respiratory syncytial virus infection in washington, d.c. 3. composite analysis of eleven consecutive yearly epidemics influenza a and b virus infection in infants and young children during the years 1957-1976 the seattle virus watch. vi. observations of infections with and illness due to parainfluenza, mumps and respiratory syncytial viruses and mycoplasma pneumoniae risk of primary infection and reinfection with respiratory syncytial virus the virus watch program: a continuing surveillance of viral infections in metropolitan new york families. ix. a comparison of infections with several respiratory pathogens in new york and new orleans families rhinovirus-associated hospitalizations in young children coronavirus infection and hospitalizations for acute respiratory illness in young children can the rolling cross-sectional survey design be used to estimate the effectiveness of influenza vaccines? capturing campaign effects physician perspectives on incentives to participate in practice-based research: a greater rochester practice-based research network (gr-pbrn) study measuring disease frequency in the marshfield epidemiologic study area (mesa) effectiveness of seasonal influenza vaccines in the united states during a season with circulation of all three vaccine strains application of taqman low-density arrays for simultaneous detection of multiple respiratory pathogens field evaluation of taqman array card (tac) for the simultaneous detection of multiple respiratory viruses in children with acute respiratory infection human coronavirus infections in rural thailand: a comprehensive study using real-time reverse-transcription polymerase chain reaction assays chi-squared and fisher-irwin tests of two-by-two tables with small sample recommendations an introduction to the bootstrap a study of illness in a group of cleveland families. i. plan of study and certain general observations frequency of acute respiratory illnesses and circulation of respiratory viruses in households with children over 3 surveillance seasons incidence of acute respiratory infections in australia update: influenza activity -united states, 2010-11 season, and composition of the 2011-12 influenza vaccine subjective social status predicts wintertime febrile acute respiratory illness among women healthcare personnel this study was funded by the centers for disease control and prevention through cooperative agreements with the university of rochester (u01 ip000172) and marshfield clinic research foundation (u01 ip000183). the findings and conclusions in this report are those of the authors and do not necessarily represent the views of the centers for disease control and prevention.we thank dean erdman, md (cdc), for his assistance and invaluable advice on laboratory assays and edward belongia, md, for his overall guidance on the study methods. none. key: cord-337692-b89ow1mf authors: petti, s.; cowling, b. j. title: ecologic association between influenza and covid-19 mortality rates in european countries date: 2020-09-11 journal: epidemiol infect doi: 10.1017/s0950268820002125 sha: doc_id: 337692 cord_uid: b89ow1mf ecologic studies investigating covid-19 mortality determinants, used to make predictions and design public health control measures, generally focused on population-based variable counterparts of individual-based risk factors. influenza is not causally associated with covid-19, but shares population-based determinants, such as similar incidence/mortality trends, transmission patterns, efficacy of non-pharmaceutical interventions, comorbidities and underdiagnosis. we investigated the ecologic association between influenza mortality rates and covid-19 mortality rates in the european context. we considered the 3-year average influenza (2014–2016) and covid-19 (31 may 2020) crude mortality rates in 34 countries using eurostat and ecdc databases and performed correlation and regression analyses. the two variables – log transformed, showed significant spearman's correlation ρ = 0.439 (p = 0.01), and regression coefficients, b = 0.743 (95% confidence interval, 0.272–1.214; r(2) = 0.244; p = 0.003), b = 0.472 (95% confidence interval, 0.067–0.878; r(2) = 0.549; p = 0.02), unadjusted and adjusted for confounders (population size and cardiovascular disease mortality), respectively. common significant determinants of both covid-19 and influenza mortality rates were life expectancy, influenza vaccination in the elderly (direct associations), number of hospital beds per population unit and crude cardiovascular disease mortality rate (inverse associations). this analysis suggests that influenza mortality rates were independently associated with covid-19 mortality rates in europe, with implications for public health preparedness, and implies preliminary undetected sars-cov-2 spread in europe. the covid-19 outbreak, caused by the severe acute respiratory syndrome coronavirus 2, sars-cov-2, evolved in two distinct phases. the former was a local outbreak first detected in china in december 2019, and the latter the subsequent spread of the virus to the rest of the world. on 11 march 2020, the world health organization (who) made the assessment that the covid-19 outbreak could be characterised as a pandemic [1] . in the beginning of the second phase, the covid-19 pandemic has been particularly severe in europe and north america. indeed, by the end of may mortality rates were as high as 22.5 and 14.3 per 100 000, respectively, while in the remaining continents rates were lower than 1.0 per 100 000 [2] . covid-19 spread in europe has been uneven, with italy experiencing the highest death toll in february and march, followed by other western countries. by the end of may, covid-19 mortality rates were ranging between higher than 50 per 100 000 in belgium, spain, uk and italy, and lower than 2 per 100 000 in slovakia, greece and bulgaria. several factors probably explained these varying mortality rates, such as nature and timeliness of implementation of covid-19 control policy measures [3] , demographic variables [4] , healthcare system quality and ability to manage the rapid covid-19 spread [5] [6] [7] and the timing of sars-cov-2 introduction in the community [8] . as for the last issue, several studies suggest that sars-cov-2 could be circulating in europe before the detection of the early covid-19 cases. namely, although the first italian patient with covid-19 was identified on 20 february 2020 and the earlier containment measures were already implemented on 21 february [9] , on 21-29 february almost 3% of the residents of a small italian town in the area of the outbreak epicentre resulted infected with sars-cov-2 [10] . in addition, seroprevalence of sars-cov-2 (test accuracy, 100% sensitivity, 98.2% specificity) in blood samples collected in december 2019 from healthy donors living in milan was as high as 2.7%, according to a preprint survey [11] , while a nasopharyngeal swab collected in december 2019 from a french patient admitted to intensive care unit for severe influenza-like illness (ili) was reanalysed in april 2020 and resulted positive for sars-cov-2 [12] . phylogenetic analyses also support this hypothesis, showing that sars-cov-2 was circulating outside china since fall 2019, and there have been multiple sars-cov-2 introductions in europe [13] [14] [15] . uncontrolled virus circulation in humans before its discovery is typical of the human coronavirus (hcov) species [16] . the investigation of the dynamics of the covid-19 outbreak must, therefore, account for sars-cov-2 circulation that occurred before the implementation of nationwide public health measures and that could help explain why sars-cov-2 was ubiquitous in europe as early as in may, with seroprevalence estimates of 1-2% in blood donors and 4-5% in the general population, with peaks of 10-15% [17] [18] [19] , that were likely limited by the unprecedented public health measures implemented in most countries after the detection of covid-19 cases [3] . the design of specific anti-covid-19 control measures, the implementation of community-based control strategies and the proper allocation of resources, can benefit from the investigation of the country-based determinants associated with covid-19 mortality and severity. for this reason, during the sars-cov-2 pandemic many ecologic studies have been performed. the majority of them considered variables that reflected at population level the risk factors for covid-19 severity at an individual level, such as population ageing, prevalence of diseases associated with covid-19 death and severity, healthcare system capacity to face the public health emergency, etc. [4, [20] [21] [22] . the assessment of covid-19 mortality determinants could benefit from similarities between this and other respiratory infectious diseases, particularly influenza, as these diseases share several characteristics. indeed, during the 2019-2020 season, ili and covid-19-like illness (cli) followed similar weekly incidence rate trends, although absolute values were different, as shown by the national syndromic surveillance program in usa [23] , and the general practitioners' network in france [24] . in addition, the public health measures taken to constrain the sars-cov-2 outbreak in japan also limited the activity of seasonal influenza [25] . similarity between influenza and covid-19 incidence and mortality rates, however, does not result in the equivalence between these diseases, since influenza virus and sars-cov-2 are two distinct species of enveloped rna virus belonging to two different families. indeed, there are clinical, epidemiological and biological differences between the two diseases [26] which lead to differences in disease burden, case-fatality rates, proportion of asymptomatic individuals, etc. however, tolksdorf and colleagues found that community-based influenza determinants could somewhat predict covid-19 burden [27] . thus, in addition to the aforementioned aggregated variables, population-level counterparts of individual-level covid-19 severity risk factors, influenza-related variables could be eligible as determinant of covid-19 mortality in european countries, to build an accurate covid-19 mortality model. therefore, the aim of this study was to investigate the ecologic association between influenza and covid-19 mortality rates in the european countries. data on covid-19 deaths in 34 european countries were gathered from the covid-19 database of the european centre for disease prevention and control (ecdc) [2] . crude covid-19 mortality rates (number of deaths per 100 000) were assessed using the population on 1 january 2020 extracted from the eurostat database [28] . this study focused on the first major epidemic waves of covid-19 in europe up until 31 may 2020 [29] . crude influenza mortality rates for the last available years (i.e. 2014-2016), were assessed. the number of influenza deaths in a given year and the population on 1 january of that year were extracted from the eurostat database [28] and crude mortality rates calculated. then, for each country the 3-year average influenza mortality rates were assessed. some, but not all demographic, health and healthcare determinants, potentially associated with influenza and covid-19 mortality rates also were extracted through the same database and assessed. namely the association between covid-19 and 3-year average influenza mortality rates was explored using the nonparametric spearman's correlation coefficient ρ. then, the covid-19 and 3-year average influenza mortality rates were log transformed to normalise variances and simple and multiple regression analyses were performed with log covid-19 mortality rate as dependent variable. zero values that could not be log transformed were given the lowest detected value. the explanatory variables initially considered for the multiple regression analysis were, 3-year average population, life expectancy at birth, healthy life years at birth, 3-year average influenza vaccination coverage in population aged ≥65 years, 3-year average crude all-cause mortality rate, 3-year average crude pneumonia mortality rate, crude cvd mortality rate, number of hospital beds per population unit. mortality rates, average population and number of hospital beds were log transformed. correlation matrix was preliminarily performed to investigate collinearity that could inflate the coefficient estimates. only non-correlated variables, with pearson's correlation coefficients <0.4, were considered. influenza mortality was forced into the model. in order to control the regression model for overfitting, due to the inclusion of unneeded predictors, the regression was initially run with all the non-collinear variables and variables that yielded statistically non-significant coefficient estimates (p ≥ 0.05) were excluded, thus obtaining a limited set of meaningful variables. in order to study whether influenza and covid-19 mortality rates shared common determinants that may help justify the similarity between these population-based variables, a series of simple regression analyses was designed treating both covid-19 and influenza mortality rates as dependent variables, and using the same set of determinants considered for the multiple regression analysis. since the influenza mortality rates could be unreliable in small countries, the analysis was repeated considering only countries with population higher than 2 000 000 individuals. the agreement between influenza and covid-19 mortality severity also was studied. more specifically, countries were grouped in quartiles according to the two mortality rates, and the agreement between influenza and covid-19 quartiles was investigated. the absolute agreement (i.e. the proportion of countries located in the same influenza and covid-19 quartile), and 2 s. petti and b. j. cowling the intraclass correlation coefficient (icc), were assessed. two-way absolute agreement single measure icc was used, considering the country classification into covid-19 mortality quartiles as reference value [30] . in order to explore the potential of a multivariate ecologic study to predict covid-19 mortality (actually, this was not an aim of the current study), the agreement between the observed covid-19 mortality and the covid-19 mortality estimated by the multiple regression analysis, also was investigated using the same methodology. there were 34 countries included in the analysis (supplementary table s1 ). the overall covid-19 mortality rate was 27.76 per 100 000, the lowest and highest rates were reported in slovakia and belgium with 0.51 and 82.52 per 100 000, respectively. the overall 3-year average influenza mortality rate in the 34 countries was roughly 30 times lower, namely, 0.91 per 100 000, and the lowest and highest rates were reported in lichtenstein and finland, with 0.00 and 2.49 per 100 000, respectively. the two mortality rates were correlated (spearman's ρ = 0.439; p = 0.01). the simple regression coefficient was b = 0.743 (95% confidence interval, 0.272-1.214; p = 0.003), with r 2 = 0.244 ( table 1 ), suggesting that 3-year average influenza mortality rate could explain 24.4% of the between-country variations in covid-19 mortality rate. several investigated determinants were highly inter-correlated (supplementary table s2 ), and after the elimination of collinear variables, four variables were remaining that were considered for the initial multiple regression model (table 1) . after the removal of pneumonia mortality rate, the regression coefficient for influenza mortality resulting from the final model was b = 0.472 (95% confidence interval, 0.067-0.878; p = 0.02), with final model r 2 = 0.549 (table 1) , that confirmed the robustness of the association between the two mortality rates. life expectancy at birth, influenza vaccination coverage in the elderly (direct associations), number of hospital beds and cvd mortality rates (inverse associations) were significantly associated with both influenza and covid-19 mortality rates, while population size was directly associated with covid-19 mortality ( table 2) . the countries with population lower than 2 000 000 were lichtenstein, iceland, malta, luxembourg, cyprus, latvia and estonia. the analyses repeated considering only the remaining 27 countries improved the association between influenza and covid-19 mortality rates, and confirmed the previous results. namely, spearman correlation ρ = 0.476 (p = 0.01), simple regression coefficient b = 0.837 (95% confidence interval, 0.326-1.349; p = 0.002; r 2 = 0.313), multiple regression coefficient b = 0.887 (95% confidence interval, 0.438-1.336; p = 0.0004; r 2 = 0.496) (data not shown in table). ten countries were classified in the same covid-19 and influenza mortality quartiles, with a fair absolute agreement of 29.4%, that was higher for countries in the first and the highest quartiles. namely, bulgaria, cyprus, lichtenstein and slovakia were in the first quartiles, and belgium, france, netherlands, sweden were in the fourth quartiles (supplementary table s3) . seventeen countries showed a discrepancy of only one quartile, while two-quartile discrepancies were reported for seven countries and no third-quartile discrepancy was found. the icc resulted 0.44 (95% confidence interval, 0.12-0.68). the highest covid-19 mortality rates estimated through multiple regression were provided for france, germany, spain, italy and uk, while the lowest were provided for lichtenstein, slovakia, hungary, bulgaria and cyprus (supplementary table s4 ). as expected, the multiple regression model provided higher agreement between quartile distributions. indeed, the absolute agreement was 55.9% (19 countries) and icc = 0.723 (95% confidence interval, 0.512-0.852) (supplementary table s3 ). this analysis showed that 3-year average influenza mortality rate was associated with covid-19 mortality rate in the european context, although influenza mortality alone could explain only part of the covid-19 mortality variability. the discrepancy between the two mortality rates was likely due to the aforementioned differences between the two diseases at population and individual levels [26, 31] . an apparently perplexing characteristic of the reported association between the two mortality rates was that while influenza virus circulation during the seasons considered in the present analysis was uncontrolled, sars-cov-2 circulation was probably limited by the widespread exceptional public health measures implemented in europe [32] . therefore, assuming that the reported association between the two rates was not spurious, the most likely explanation of the present results was that sars-cov-2 circulation also was partly uncontrolled. actually, surveys and phylogenetic analyses support the idea of multiple introductions of the virus in europe since 2019 [10] [11] [12] [13] [14] [15] . such an undetected virus circulation is not surprising, since patients with covid-19/cli have been frequently misclassified as patients with ili [23, 24] , and is corroborated by sars-cov-2 seroprevalence surveys [1719, 33] . the implementation of country-based control policies likely prevented further severe sars-cov-2 outbreak propagation, thus explaining the covid-19 incidence rate of 3-5% in may 2020, lower than influenza rate that is usually 10% or higher [34, 35] . the history of other hcovs corroborates this hypothesis. for example, the first patient infected with hcov-nl63, a child with atypical bronchiolitis, was detected in january 2003 in amsterdam. soon after, hcov-nl63 positive patients from all over the world with upper and/or lower respiratory tract infections were detected, and seroprevalence values as high as 2-9% were reported. such an apparently rapidly spreading pandemic was explained by the long undetected virus circulation confirmed by the analysis of a specimen collected from a child with pneumonia that was stored on kidney simian cells since 1988. thus, hcov-nl63 was already circulating fifteen years before its detection [16, 36] . relatively free sars-cov-2 circulation in europe also was promoted by inefficient and untimely crisis coordination at central level [29] , and by delays and contradictions of some international public health organisations in acknowledging community transmission, typical of pandemics, that must lead to public health control measures. indeed, on 19 april 2020, community transmission was not acknowledged yet in france, spain, uk, italy, where 15 000-25 000 covid-19 deaths were already reported, and in belgium and the netherlands, with 3500-5000 deaths, but was confirmed in small countries such as san marino, andorra, bosnia and kosovo [37] . the question remains unanswered, on whether earlier community transmission acknowledgement in europe, and consequent timely implementation of coordinated covid-19 control measures would have limited the high burden of covid-19. the current study corroborated the assumption that influenza and covid-19 mortality rates share similar determinants. indeed, both diseases were significantly associated with similar demographic, health and healthcare determinants, excluding population size that was associated only with covid-19 ( table 2 ). this is also the reason why crude mortality rates were used instead of standardised rates, as the standardisation process would have reduced the impact of population age structure on mortality rates, while the rationale of the current study was that influenza and covid-19 share similar population-based determinants, and population age structure was among them. influenza and covid-19 share another important population-level characteristic. namely, the problem of misclassifications and disagreements in classification that lead to inconsistent burden of disease estimates. although influenza has been recognised as an important cause of mortality, particularly in the elderly, mortality rates are generally low, because much of related mortality is not attributed to primary influenza infection, but to complications and secondary infections. this problem generated incongruences in classifying influenza as underlying or contributing cause of death [38] . as for covid-19, differences in mortality between countries and even within countries were partly attributable to the use of different criteria to classify covid-19 deaths [39] . to overcome the problem of misclassification the us national center for health statistics coined an aggregated variable called 'pic', that considered all deaths attributed to pneumonia, influenza and covid-19, updating another variable called 'p&i', based on influenza and pneumonia [40] . in the current study, however, pneumonia mortality did not result associated with influenza and covid-19 (table 2) , and unreported analyses using 3-year average 'p&i' mortality rate provided non-significant results. influenza and covid-19 mortality rates resulted associated with population age structure, as shown in table 2 , and corroborated by the eurostat report showing that between 2012 and 2016, as many as 70% influenza deaths occurred in the elderly aged ≥65 years, and the european standardised influenza mortality rates in this age group were between ten and twenty times higher than in subjects younger than 65 years [41] . the covid-19 burden in the elderly was even higher. indeed, the elderly aged ≥65 years accounted for 90-95% of deaths in european countries and their risk of dying was up to 80 times higher than in younger individuals [42] . another characteristic shared by influenza and covid-19 deaths was the impact of comorbidities on mortality. indeed, three-fourth influenza deaths occur in patients with comorbidities [43] , while for covid-19 such a proportion is higher than 90% [42, 44] . this study reported an inverse association between number of hospital beds and mortality rates (table 2) , thus showing that high influenza and covid-19 mortality was also due to inefficiencies of the healthcare systems, and corroborated by data from several european countries [45] . similarly, the inadequateness of the healthcare system has been responsible for the high covid-19-related death toll reported in many countries, such as uk [5] , italy [46] and spain [47] . the direct association between influenza vaccination coverage among the elderly and influenza and covid-19 mortality rates reported in this study ( table 2) was corroborated by populationtable 2 . associations between demographic, health and healthcare determinants and 3-year average crude influenza mortality rate and crude covid-19 mortality rate (log transformed), assessed through simple regression analyses (regression coefficients; 95% confidence intervals in brackets) influenza mortality rate covid-19 mortality rate [48] [49] [50] . this paradoxical effect of influenza vaccine is due to the fact that vaccine uptake is more likely in the categories who need it least, that is, women, elderly younger than 80 years and subjects without comorbidities [51] , an effect called inverse care law by julian tudor hart in 1971, who stated that 'the availability of good medical care tends to vary inversely with the need for it in the population served' [52] . unfortunately, the inverse care law also applies to preventive medicine including cancer screening [53, 54] , and influenza vaccination [55] , and explains the apparently puzzling direct association between influenza vaccination coverage and covid-19 mortality, since covid-19 mortality risk was twofold higher in men than in females, 13-fold higher in individuals older than 80 years than in those aged 65-79 years, and 5-to-15-fold higher in patients with comorbidities, than in those without [56] . in other words, individuals at higher influenza and covid-19 mortality risk are those who are less likely to get vaccinated. the reported association between high influenza vaccine coverage and high influenza and covid-19 mortality has nothing to do with intrinsic vaccine efficacy, since ecologic studies are subject to ecologic fallacy that prevents from inferring associations at an individual level. the multiple regression analysis showed that cvd mortality was inversely associated with covid-19 mortality (table 1) . cvd, particularly ischaemic heart disease and stroke, is the leading cause of death in europe, accounting for 40% and 49% of all deaths in males and females, respectively, and is also the leading cause of premature death, accounting for more than 35% of all deaths under 75 years. differences in cvd mortality are, therefore, the main responsible of differences in life expectancy at birth, country distribution for age, and potential years of life lost. these differences are particularly evident between eastern and western european countries [57] . cvd and older age are also the main risk factors associated with covid-19 death at an individual level [58] [59] [60] . these considerations help explain why western european countries showed generally high influenza and covid-19 mortality rates, while eastern european countries showed the reverse. indeed, covid-19 and influenza are particularly lethal in elderly individuals and, thus, influenza and covid-19 mortality rates are particularly high in countries where the proportion of elderly is higher. since cvd deaths are responsible for premature mortality, countries where cvd mortality is higher, also are those with the lowest proportion of elderly and, consequently, with the lowest proportion of susceptible individuals at higher risk of dying from both covid-19 and influenza. present research is an ecologic study with all the corresponding advantages and disadvantages of this approach. indeed, the use of aggregated data prevented the identification of associations at an individual level, a problem known as ecologic fallacy. on the other hand, since these studies are relatively simple and reproducible they provide useful information in emergency situations like the covid-19 pandemic. such information, however, must be considered carefully and implementing public health control measures on the basis of ecologic studies alone could be problematic [61] . during this pandemic several ecologic studies have been published, reporting associations between covid-19 incidence and mortality rates and bcg vaccine coverage [62] , malaria prevalence [63] , environmental and meteorological factors, pollutants (reviewed in [64] ), etc. although these associations were robust enough, they could not be considered to design covid-19 control policies, due to ecologic fallacy. in the same way, the current study did not show that influenza prevention at an individual level leads to covid-19 prevention, but only that the two mortality rates were associated at the population level. the second important limit of this study was the reported problem of the reliability of both influenza and covid-19 death counts [38, 39] , that could lead to uncertainties in the true mortality rates in the countries under investigation in this study. an ideal, yet unfeasible, approach would be that every dying individual with ili, cli, acute respiratory illness and pneumonia was tested for all the circulating influenza strains and for sars-cov-2. in the case of influenza, this uncertainty regarding the death counts, led to varying estimates of the global number of attributable deaths, ranging from the lowest limit provided by the global burden of disease study of 99 000, to the highest limit provided by the cdc of 650 000 [65] . the problem of consistency of aggregated data, however, is shared by almost all diseases and conditions. for example, the estimated global number of deaths from breast cancer was 630 000 according to globocan 2018 [66] , and 535 000 according to the global burden of disease study [67] , with important differences within each country. the last limitation of this study was that influenza mortality rate alone could not be considered an optimal covid-19 mortality rate predictor, since the multiple regression analysis showed that there were other important population-based confounders associated with covid-19 mortality. they could be variables related to age structure and prevalence of comorbidities associated with covid-19 mortality. for example, age structure explained part of the between-country differences in covid-19 mortality and case-fatality rates [4, 20] ; median prevalence of the five conditions most frequently associated with severe covid-19 in usa allowed to identify the areas at highest risk for covid-19 death [21] ; age-specific prevalence of comorbidities explained the differences in mortality between nigeria, brazil and italy [22] . economic and healthcare associated variables are other aggregated data potentially useful to predict covid-19 severity and spread [68] [69] [70] , as well as inequalities within the general population [71] . unlike these studies, however, the present analysis considered the mortality rate from an infectious disease that was not somewhat causally associated with covid-19 mortality and death and was based on a different assumption, namely, that the two diseases shared a set of determinants, ranging from the characteristics of the population at highest risk, to transmission routes, from case and death misclassifications, to the efficiency of the healthcare systems. in conclusion, influenza and covid-19 mortality rates were significantly associated and influenza mortality could be an eligible predictor for the design of more accurate multivariable covid-19 mortality assessment and prediction models. data availability statement. the data that support the findings of this study can be downloaded from the ecdc and eurostat databases and are, in part, displayed in supplementary material. world health organization (2020) coronavirus disease 2019 (covid-19) situation report -51 european centre for disease prevention and control. covid-19 database whose coronavirus strategy worked best? scientists hunt most effective policies demographic science aids in understanding the spread and fatality rates of covid-19 covid-19: 'staggering number' of extra deaths in community is not explained by covid-19 covid-19: implications for prehospital, emergency and hospital care in patients with acute coronary syndromes indications for healthcare surge capacity in european countries facing an exponential increase in coronavirus disease (covid-19) cases geographic differences in covid-19 cases, deaths, and incidence -united states critical care utilization for the covid-19 outbreak in lombardy, italy: early experience and forecast during an emergency response suppression of a sars-cov-2 outbreak in the italian municipality of vo sars-cov-2 seroprevalence trends in healthy blood donors during the covid-19 milan outbreak sars-cov-2 was already spreading in france in late whole genome and phylogenetic analysis of two sars-cov-2 strains isolated in italy in emergence of genomic diversity and recurrent mutations in sars-cov-2 genomic characterization and phylogenetic analysis of sars-cov-2 in italy human coronavirus nl63, a new respiratory virus institudo de salud carlos iii database the infection fatality rate of covid-19 inferred from seroprevalence data the contribution of the age distribution of cases to covid-19 case fatality across countries: a 9-country demographic study estimated county-level prevalence of selected underlying medical conditions associated with increased risk for severe covid-19 illness -united states besides population age structure, health and other demographic factors can contribute to understanding the covid-19 burden excess cases of influenza-like illnesses synchronous with coronavirus disease (covid-19) epidemic, france seasonal influenza activity during the sars-cov-2 outbreak in japan assessment of deaths from covid-19 and from seasonal influenza influenza-associated pneumonia as reference to assess seriousness of coronavirus disease (covid-19) eurostat database coronavirus response. public health. crisis coordination forming inferences about some intraclass correlation coefficients coronavirus disease 2019 vs. the flu. johns hopkins medicine the effect of social distance measures on covid-19 epidemics in europe: an interrupted time series analysis prevalence of sars-cov-2 specific neutralising antibodies in blood donors from the lodi red zone in lombardy, italy, as at 06 natural attack rate of influenza in unvaccinated children and adults: a meta-regression analysis seasonal incidence of symptomatic influenza in the united states a previously undescribed coronavirus associated with respiratory disease in humans world health organization the burden of influenza: a complex problem how comparable is covid-19 mortality across countries? 4 centers for disease control and prevention. national center for health statistics. technical notes. provisional death counts for coronavirus disease how many eu citizens die from influenza? population-level covid-19 mortality risk for non-elderly individuals overall and for non-elderly individuals without underlying diseases in pandemic epicenters the burden of influenza in england by age and clinical risk group: a statistical analysis to inform vaccine policy european centre for disease prevention and control (2020) coronavirus disease 2019 (covid-19) in the eu/eea and the uk -ninth update health care costs of influenza-related episodes in high income countries: a systematic review the italian health system and the covid-19 challenge the resilience of the spanish health system against the covid-19 pandemic the effect of influenza vaccination for the elderly on hospitalization and mortality: an observational study with a regression discontinuity design vaccines for preventing influenza in the elderly influenza-related mortality in the italian elderly: no decline associated with increasing vaccination coverage influenza vaccination and mortality: differentiating vaccine effects from bias the inverse care law how many individuals must be screened to reduce oral cancer mortality rate in the western context? a challenge access to lung cancer screening programs in the united states: perpetuating the inverse care law influenza vaccination, inverse care and homelessness: cross-sectional survey of eligibility and uptake during the 2011/12 season in london opensafely: factors associated with covid-19-related hospital death in the linked electronic health records of 17 million adult nhs patients european cardiovascular disease statistics risk factors associated with mortality among patients with covid-19 in intensive care units in factors associated with death in critically ill patients with coronavirus disease 2019 in the us clinical features of fatalities in patients with covid-19. disaster medicine and public health preparedness ecologic studies revisited bcg vaccine protection from severe coronavirus disease 2019 (covid-19) global spread of coronavirus disease 2019 and malaria: an epidemiological paradox in the early stage of a pandemic covid-19 and the environment: a critical review and research agenda global mortality associated with seasonal influenza epidemics: new burden estimates and predictors from the glamor project global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 29 cancer groups, 1990 to 2016: a systematic analysis for the global burden of disease study potential association between covid-19 mortality and health-care resource availability effect of income on the cumulative incidence of covid-19: an ecological study the death rate for covid-19 is positively associated with gross domestic products association between state-level income inequality and covid-19 cases and mortality in the usa key: cord-253208-wknht58z authors: wang, xue; li, xincheng; shang, yu; wang, junwei; zhang, xiaona; su, dongju; zhao, shuai; wang, qin; liu, lei; li, yupeng; chen, hong title: ratios of neutrophil-to-lymphocyte and platelet-to-lymphocyte predict all-cause mortality in inpatients with coronavirus disease 2019 (covid-19): a retrospective cohort study in a single medical centre date: 2020-09-09 journal: epidemiol infect doi: 10.1017/s0950268820002071 sha: doc_id: 253208 cord_uid: wknht58z the coronavirus disease 2019 (covid-19) caused by severe acute respiratory syndrome coronavirus 2 (sars-cov-2) is a public health emergency of international concern. the current study aims to explore whether the neutrophil-to-lymphocyte ratio (nlr) and platelet-to-lymphocyte ratio (plr) are associated with the development of death in patients with covid-19. a total of 131 patients diagnosed with covid-19 from 13 february 2020 to 14 march 2020 in a hospital in wuhan designated for treating covid-19 were enrolled in the current study. these 131 patients had a median age of 64 years old (interquartile range: 56–71 years old). furthermore, among these patients, 111 (91.8%) patients were discharged and 12 (9.2%) patients died in the hospital. the pooled analysis revealed that the nlr at admission was significantly elevated for non-survivors, when compared to survivors (p < 0.001). the nlr of 3.338 was associated with all-cause mortality, with a sensitivity of 100.0% and a specificity of 84.0% (area under the curve (auc): 0.963, 95% confidence interval (ci) 0.911–1.000; p < 0.001). in view of the small number of deaths (n = 12) in the current study, nlr of 2.306 might have potential value for helping clinicians to identify patients with severe covid-19, with a sensitivity of 100.0% and a specificity of 56.7% (auc: 0.729, 95% ci 0.563–0.892; p = 0.063). the nlr was significantly associated with the development of death in patients with covid-19. hence, nlr is a useful biomarker to predict the all-cause mortality of covid-19. since december 2019, a cluster of unexplained pneumonia cases has been reported in wuhan, hubei province. this was subsequently identified as severe acute respiratory syndrome coronavirus 2 (sars-cov-2), and the disease caused by this was named, coronavirus disease 2019 , by the world health organization (who) [1] . as of 13 july 2020, 12 552 765 cases have been infected by sars-cov-2, and 561 617 patients have died worldwide. furthermore, 216 countries have been affected, posing a major threat to human public health [2] . to date, a specific treatment for sars-cov-2 has not been recommended, except for meticulous supportive care. therefore, it is crucial to identify risk factors that are associated with the poor prognosis of patients with covid-19. some studies suggested that older age, comorbidities (hypertension and cardiovascular diseases), sequential organ failure assessment scores and some laboratory indices, such as neutrophil, d-dimer, lymphocyte, interleukin (il)-6 and c-reactive protein, are associated with the development of poor prognosis [3] [4] [5] [6] [7] [8] . however, some of these risk factors are cumbersome and costly. the neutrophil-tolymphocyte ratio (nlr) and platelet-to-lymphocyte ratio (plr) are novel biomarkers that provide important information about the systemic inflammation status, and these are easily available from routine laboratory studies. the elevated nlr and plr are significantly associated with the mortality of patients with infectious diseases [9] [10] [11] [12] . therefore, recent studies have suggested that nlr is an effective predictor for the mortality of patients with covid-19 [4, 13, 14] . the dynamic increase of plr during the hospitalisation might suggest the severity and prognosis of the disease [15] . however, to date, no studies have simultaneously explored the values of nlr and plr in predicting the mortality in covid19 . in the current study, the investigators aimed to determine whether the plr can serve as a valuable predictor of in-hospital mortality, and the value of nlr for predicting the all-cause mortality in patients with covid-19. the current retrospective cohort study included 151 patients with covid-19 in wuhan no. 1 hospital from 13 february 2020 to 14 march 2020. covid-19 was diagnosed according to the seventh edition of the interim guidance of the national health commission of the people's republic of china [16] . the patients were excluded from further analyses when they had an active condition at the time of covid-19, which could significantly influence the blood cell count, including chronic obstructive pulmonary disease (copd). furthermore, patients that did not have a record of their blood cell count were also excluded ( fig. 1) . the epidemiology, demography, clinical manifestations, laboratory examination and outcome were extracted from the electronic medical records by the harbin aid hubei medical team. most of the clinical data were collected from the first day of admission, unless otherwise noted in the current study. for severe pneumonia (meeting any of the following): (1) dyspnoea, respiratory rate of ≥30 breaths/min; (2) peripheral oxygen saturation ≤93% at rest and (3) oxygen partial pressure/oxygen uptake fraction of ≤300 mmhg (1 mmhg = 0.133 kpa). for non-severe pneumonia: the above criteria were not met. the definition of non-effective antibiotic treatment (meeting any of the following): (1) no decrease in temperature after 48-72 h of antibiotic therapy and (2) no improvement in symptoms after 48-72 h of antibiotic therapy. all patients included in the current study had a definite outcome (death or discharge). the current study was approved by the medical ethics committee of the second affiliated hospital of harbin medical university (ky2020-011). spss statistics 25 (ibm spss) was used for the statistical analyses. the categorical variables were described as the number/total number (%), and continuous variables were described using the mean, median and interquartile range (iqr) values. the kolmogorov-smirnov test was used to evaluate the data for the normality of the distribution. the means for continuous variables were compared according to the independent group t-tests, when the data were normally distributed. otherwise, the mann-whitney test was used. the categorical data were compared by χ 2 test or fisher's exact test. the receiver operating characteristic (roc) curve was used to state the sensitivity and specificity of nlr and plr for all-cause mortality and disease severity. the youden index was calculated to determine the optimal cut-off values. in order to explore the risk factors associated with in-hospital mortality, the logistic regression model was used. bilateral test (the test level α = 0.05) was used, and p < 0.05 was considered statistically significant. a total of 131 patients with covid-19 were included in the current study. the demographic characteristics of these patients are shown in table 1 . among the 131 patients with a median age of 64 years old (iqr: 56-71), 12 (9.2%) patients died. there were no significant differences in gender and comorbidities between these two groups. in the non-survivor group, more patients had severe pneumonia (33.3% vs. 13.4%, p = 0.087). approximately 87% of patients received varying degrees of antibiotic therapy during their hospitalisation, and more nonsurvivors were non-effective to antibiotic treatment (70% vs. 15.4%, p < 0.001). compared to the survivors, the non-survivors were older (median age 80 (iqr: 70-85) vs. 64 (iqr: 56-69), p < 0.01), and were more likely to present with initial symptoms of dyspnoea (12 (100%) vs. 15 (12.6%), p < 0.01). however, more survivors presented with initial symptoms of cough (85 (71.4%) vs. 4 (33.3%), p = 0.018), when compared with the non-survivors. the initial laboratory indicators of surviving and dead patients are presented in table 2 . the leucocytes (6.03 × 10 9 /l vs. 11.66 × 10 9 /l, p < 0.001) and neutrophils (3.73 × 10 9 /l vs. 10.07 × 10 9 /l, p < 0.001) were significantly higher in non-survivors. the lymphocyte counts (1.53 ± 0.61 × 10 9 /l vs. 0.74 ± 0.38 × 10 9 /l, p < 0.001) and platelet counts (240 × 10 9 /l vs. 158.27 × 10 9 /l, p = 0.007) were significantly lower in non-survivors. the value of aspartate aminotransferase (ast, 23 u/l vs. 48 u/l, p < 0.01) was higher in non-survivors. at the same time, the albumin level in the nonsurvivor group was lower (33.81 g/l vs. 29.16 g/l, p = 0.008), while the globulin level was higher (27.6 ± 3.14 g/l vs. 33.27 ± 6.97 g/l, p = 0.032). these results revealed that the albumin/globulin ratio for non-survivors was lower (a/g, 1.27 ± 0.25 vs. 0.90 ± 0.17, p < 0.001). furthermore, the levels of lactate dehydrogenase (ldh, 200.5 u/l vs. 721.0 u/l, p < 0.01) and blood urea nitrogen (bun, 4.10 mmol/l vs. 9.70 mmol/l, p < 0.01) were higher in nonsurvivors. the results revealed that the values of alanine aminotransferase (24 u/l vs. 35 u/l, p < 0.01), creatine kinase (51.0 u/l vs. 107.0 u/l, p = 0.017) and serum creatinine (61.5 μmol/l vs. 95.0 μmol/l, p = 0.04) were all within the normal range in these two groups. there were no statistical differences in the other laboratory indices between these two groups. as shown in table 3 , nlr was significantly elevated in nonsurvivors, when compared to survivors (13.87 (7.50-24.82) vs. 1.95 (1.43-2.58), p < 0.001). however, there were no significant differences in plr for non-survivors, when compared to survivors (p = 0.251). in addition, it was found that nlr was higher in the severe group (6.88 (3.54-11.18) vs. 2.21 (1.51-9.85), p = 0.065) after grouping the patients according to their severity, and plr had similar results (195.97 (157.75-246.05) vs. 165.89 (112.90-227.96)), but the difference was not significant (p = 0.104). as shown in table 4 , the univariate analysis revealed that age (odds ratio (or) 1.116, 95% confidence interval (ci) 1.050-1.187, p < 0.01), white blood cell count (or 1.419, 95% ci 1.117-1.711, p < 0.001), lymphocytes (or 0.028, 95% ci 0.004-0.400, p < 0.001), neutrophils (or 2.265, 95% ci 1.284-3.997, p < 0.001), ast (or 1.034, 95% ci 1.012-1.057, p = 0.002), albumin (or 0.855, 95% ci 0.756-0.966, p = 0.012), a/g ratio (or 0.000, 95% ci 0.000-0.085, p = 0.005), ldh (or 1.008, 95% ci 1.003-1.012, p < 0.001), creatine kinase (or 1.004, 95% ci 1.001-1.007, p = 0.020), serum creatinine (or 1.025, 95% ci 1.001-1.048, p = 0.037) and bun (or 1.329, 95% ci 1.050xue wang et al. 1.682; p = 0.018) were significantly correlated with death induced by covid-19. considering the results of the univariate analysis and the problem of sample size, five serological indices (creatine kinase, albumin, ast, serum creatinine and nlr) were finally included in the multivariate analysis to determine the close relationship between nlr and death (table 4 ). after excluding the influences of the other four factors, nlr was still closely correlated with death (adjusted or 1.513, 95% ci 1.001-2.263, p = 0.044), suggesting that nlr may be a valuable biomarker in response to mortality in covid-19. based on the roc curve analysis (fig. 2 the current study suggests that the elevated nlr is associated with all-cause mortality in patients with covid-19, while plr was not associated with this. the nlr was significantly higher in non-survivors, when compared to survivors, which is consistent with the reports of other studies [4, 14] . the logistic regression analysis revealed that the nlr is associated with the mortality of covid-19 (crude or 1.860, 95% ci 1.385-2.498). after adjusting the other confounding factors, the nlr remained as a risk factor for covid-19 (adjusted or 1.513, 95% ci 1.101-2.263). it was also demonstrated that the nlr of 3.328 has a good predictive value of all-cause mortality in patients with covid-19, with a sensitivity of 100.0% and a specificity of 84.0%. in the current study, the elevated nlr may serve as a diagnostic indicator for severe covid-19, and this has been shown in other studies [4, 17] . in addition, it was found that there are many biochemical indicators closely correlated with death, such as ast, creatine kinase and serum creatinine, which are consistent with the reports of other studies [3, 6, 8, 18, 19] . these risk factors suggest that these dead patients might have had multiple organ damage at the beginning of the hospitalisation. similar to other studies [3, 8, 19] , age was also closely correlated with death in the current study. comorbidities (hypertension, diabetes and cardiovascular disease) that have been shown to be associated with death did not differ between survivors and non-survivors [3, 8, 20] , which may be due to the population heterogeneity. as it is known, the human immune system plays a major role in putting out viral infections. the nlr reflects that the high systemic inflammatory response is associated with the poor prognosis of infectious diseases [9] [10] [11] [12] . several studies have manifested that severe cases (including dead patients) of covid-19 were more likely to have higher neutrophil counts and lower lymphocyte counts, when compared with non-severe cases. thus, the elevated nlr tends to predict the severity of covid-19 [4] . through a retrospective analysis of 452 patients, qin et al. [4] reported that severe cases are likely to have higher nlrs caused by the higher neutrophil counts, but these cases would have lower lymphocyte counts, when compared to non-severe patients, indicating that the surveillance of nlrs might be helpful for the early screening of the critical illness of covid-19. furthermore, both helper t cells and suppressor t cells were below the normal levels, and the decline in helper t cells was more pronounced in severe cases that have been proven to be a key point in the weakening or suppressing overactive immune responses of sars [21] . diao et al. [22] reported that t cells significantly decreased in patients in the intensive care unit (icu), when compared to patients who were not in the icu. furthermore, the total t cells, and cd4 + and cd8 + t cells in severe and critical patients were significantly lower, when compared to the mild/moderate diseases in patients who were not in the icu [22] . in a report from china, minimally invasive autopsies were performed on three patients [23] . in addition to the known severe pulmonary lesions, there was a marked disturbance in the lymphatic haematopoietic system. splenic lymphocytosis, cellular degeneration and necrosis presented in the lymph nodes with reduced lymphocytes and focal necrosis. as a retrospective study, due to the limitations of the isolation ward at the beginning of the epidemic and the urgency of containing the covid-19 epidemic, the study did not document some data in detail, such as the dynamics of the laboratory indicators. hence, the investigators referred to the data obtained by other researchers. in the study conducted by tan et al. [24] , lymphopaenia was found to be a predictor of prognosis in patients with covid-19 pneumonia. in their time-to-lymphocyte percentage (lym%) curve model, at 10-12 days after symptom onset, patients with lym% <20% were initially classified as severe. at 17-19 days after onset, patients with lym% <20% were still at risk, and required monitoring. patients with lym% <5% were critical, had a high mortality rate, and required intensive care. the nlr was higher for identifying patients with severe pneumonia, when compared to that for identifying those who have died, suggesting that a higher nlr might predict the progression during hospitalisation [25] . in the study conducted by ding et al. [26] , the nlr was significantly higher in severe patients, when compared to non-severe patients, at all time points after hospital admission, and the nlr was positively correlated with hospitalisation time from day 5 after admission. these findings might be explained by the following reasons: (1) angiotensin-converting enzyme 2 (ace2) has been considered as the main receptor for sars-cov-2, which may be expressed in lymphocytes, cause the sars-cov-2 to directly infect these cells, and ultimately bring about lymphopaenia [27] , leaving the body vulnerable to bacterial invasion and inducing an increase in neutrophils [4] . (2) the cytokine storm has been considered as an important contributor to respiratory viral infections [28] , while cytokines, such as il-10, il-6 and tumour necrosis factor-α, might activate the neutrophils and damage the lymphocytes [25, 29] . thus, the nlr may reflect the level of severity of covid-19. it was also found that the antibiotic inefficiency was higher in the non-survival group, which might also suggest that inflammatory storms are involved in the progression of covid-19. however, due to the number of deaths (n = 12), further studies, particularly prospective studies, are needed to ascertain the value of the nlr in mortality. in addition, the current study explored the association between plr and all-cause mortality in covid-19. platelets are important immune cells in the human body, which are produced by mature megakaryocytes in the bone marrow. this plays an important role in blood coagulation, angiogenesis, immunity and inflammation [30] . the mechanism of thrombocytopaenia may be the joint action of many factors. there are many similarities between the outbreak of covid-19 and the outbreak of sars in 2003. previous studies on sars suggested that the reasons might be, as follows: (1) the viral infection and mechanical ventilation led to endothelial injury, platelet aggregation, pulmonary thrombogenesis [31] and megakaryocyte reduction, and as a result, the platelet production decreased and the consumption increased; (2) the coronavirus directly invades haematopoietic cells or xue wang et al. bone marrow stromal cells, resulting in haematopoiesis inhibition [32] . indeed, there are marked differences in the physical and chemical properties between sars and covid-19. however, it remains to be determined whether this phenomenon can be explained by the same mechanism. as a new inflammatory index, the plr can reflect the infection and factor aggregation [33] , which is more valuable, when compared to the simple platelet or lymphocyte count. by comparing the dynamic changes of the plr during hospitalisation, qu r et al. [15] reported that the increase in plr was correlated with the poor prognosis on covid-19. unfortunately, due to the low sample size and mortality rates, no value was identified in the current study. first, patients with copd were excluded, which reduced the impact of long-term inhaled glucocorticoids and airway inflammation on the outcomes. second, covid-19 imposes a significant global medical and economic burden. compared with factors, such as il-6, which respond to inflammatory markers, nlrs are simple, fast and inexpensive to obtain directly from the blood, and this can help clinicians identify the serious illness and prognosis of covid-19, thereby allowing for the aggressive adjustment of treatment plans to reduce patient death. frist, due to the retrospective study design and limited sample size, the real value of the plr might be underestimated in predicting the in-hospital death. hence, further studies are needed to ascertain the real value of the nlr and plr in predicting the mortality for covid-19. second, the investigators did not compare the nlr and plr changes in the progress of covid-19 with the values at baseline. this may lead to the misunderstanding of the predictions of biomarkers in all-cause mortality in covid-19 [15] . nlr is a simple biomarker that reflects the presence of systemic inflammation, and is associated with all-cause mortality in covid-19. the elevation of nlr was a useful biomarker to predict the mortality in covid-19. further studies are needed to ascertain the dynamic values of the nlr in predicting all-cause mortality in covid-19, and explore more useful markers to timely detect critical patients. albumin/globulin ration; bun, blood urea nitrogen; covid-19, coronavirus disease 2019; ldh, lactate dehydrogenase. data are shown as n (median, iqr), n (mean ± standard deviation) student's t-test or mann-whitney u test was used for continuous data, χ 2 test or fisher's exact test for categorical variables. references 1. world health organization (2020) naming the coronavirus disease (covid-19) and the virus that causes it world health organization (2020) coronavirus disease (covid-19) pandemic. numbers at a glance clinical course and risk factors for mortality of adult inpatients with covid-19 in wuhan, china: a retrospective cohort study dysregulation of immune response in patients with coronavirus 2019 (covid-19) in wuhan, china impact of cerebrovascular and cardiovascular diseases on mortality and severity of covid-19-systematic review, meta-analysis, and meta-regression clinical characteristics of coronavirus disease 2019 (covid-19) in china: a systematic review and meta-analysis clinical characteristics of 113 deceased patients with coronavirus disease 2019: retrospective study clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan, china inflammation biomarkers in blood as mortality predictors in community-acquired pneumonia admitted patients: importance of comparison with neutrophil count percentage or neutrophillymphocyte ratio neutrophil count percentage and neutrophillymphocyte ratio as prognostic markers in patients hospitalized for community-acquired pneumonia diagnostic value of blood parameters for community-acquired pneumonia prognostic value of an inflammatory biomarkerbased clinical algorithm in septic patients in the emergency department: an observational study neutrophil-to-lymphocyte ratio as an independent risk factor for mortality in hospitalized patients with covid-19 neutrophil-to-lymphocyte ratio and lymphocyte-to-c-reactive protein ratio in patients with severe coronavirus disease 2019 (covid-19): a meta-analysis platelet-to-lymphocyte ratio is associated with prognosis in patients with coronavirus disease-19 national health commission of the people's republic of china (2020) update on novel coronavirus-infected pneumonia situation on value of neutrophil-to-lymphocyte ratio in the classification diagnosis of coronavirus disease 2019 cardiovascular implications of fatal outcomes of patients with coronavirus disease 2019 (covid-19) risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in wuhan, china role of data registries in the time of covid-19 t cell responses to whole sars coronavirus in humans reduction and functional exhaustion of t cells in patients with coronavirus disease 2019 (covid-19) a pathological report of three covid-19 cases by minimal invasive autopsies lymphopenia predicts disease severity of covid-19: a descriptive and predictive study short-term outcomes of covid-19 and risk factors for progression dynamic profile and clinical implications of hematological parameters in hospitalized patients with coronavirus disease 2019 sars-cov-2 cell entry depends on ace2 and tmprss2 and is blocked by a clinically proven protease inhibitor an interferon-gamma-related cytokine storm in sars patients inflammatory response cells during acute respiratory distress syndrome in patients with coronavirus disease 2019 (covid-19) platelets as key players in inflammation and infection effects of oxygen-induced lung damage on megakaryocytopoiesis and platelet homeostasis in a rat model inactivation of three emerging viruses -severe acute respiratory syndrome coronavirus, crimean-congo haemorrhagic fever virus and nipah virus -in platelet concentrates by ultraviolet c light and in plasma by methylene blue plus visible light the platelet-to-lymphocyte ratio as an inflammatory marker in rheumatic diseases acknowledgements. we acknowledge the first hospital of wuhan for providing the medical records.author contributions. d.j.s., l.l., q.w., s.z., j.w.w. and x.n.z. performed the data collection. x.c.l., x.w. and y.s. prepared the first draft of the manuscript, validated the data collection, refined the research idea, performed the data analysis and edited the manuscript. h.c. and y.p.l. developed the research idea, refined the research idea, validated the data collection and edited the manuscript. h.c. and y.p.l. are the guarantors of the manuscript. ethical standards. the ethics committee of the second affiliated hospital of harbin medical university approved this study, and granted a waiver of informed consent from the study participants (ky2020-011), which was authorised by the ethics committee of wuhan no. 1 hospital.data available statement. the data that support the findings of this study are available from the corresponding author, hc, upon reasonable request. key: cord-340938-mk01k235 authors: xu, kandi; zhou, min; yang, dexiang; ling, yun; liu, kui; bai, tao; cheng, zenghui; li, jian title: application of ordinal logistic regression analysis to identify the determinants of illness severity of covid-19 in china date: 2020-07-07 journal: epidemiol infect doi: 10.1017/s0950268820001533 sha: doc_id: 340938 cord_uid: mk01k235 corona virus disease 2019 (covid-19) has presented an unprecedented challenge to the health-care system across the world. the current study aims to identify the determinants of illness severity of covid-19 based on ordinal responses. a retrospective cohort of covid-19 patients from four hospitals in three provinces in china was established, and 598 patients were included from 1 january to 8 march 2020, and divided into moderate, severe and critical illness group. relative variables were retrieved from electronic medical records. the univariate and multivariate ordinal logistic regression models were fitted to identify the independent predictors of illness severity. the cohort included 400 (66.89%) moderate cases, 85 (14.21%) severe and 113 (18.90%) critical cases, of whom 79 died during hospitalisation as of 28 april. patients in the age group of 70+ years (or = 3.419, 95% ci: 1.596–7.323), age of 40–69 years (or = 1.586, 95% ci: 0.824–3.053), hypertension (or = 3.372, 95% ci: 2.185–5.202), alt >50 μ/l (or = 3.304, 95% ci: 2.107–5.180), ctni >0.04 ng/ml (or = 7.464, 95% ci: 4.292–12.980), myohaemoglobin>48.8 ng/ml (or = 2.214, 95% ci: 1.42–3.453) had greater risk of developing worse severity of illness. the interval between illness onset and diagnosis (or = 1.056, 95% ci: 1.012–1.101) and interval between illness onset and admission (or = 1.048, 95% ci: 1.009–1.087) were independent significant predictors of illness severity. patients of critical illness suffered from inferior survival, as compared with patients in the severe group (hr = 14.309, 95% ci: 5.585–36.659) and in the moderate group (hr = 41.021, 95% ci: 17.588–95.678). our findings highlight that the identified determinants may help to predict the risk of developing more severe illness among covid-19 patients and contribute to optimising arrangement of health resources. the pandemic of the novel corona virus disease 2019 (covid19) , which originally emerged in wuhan, china in december 2019 has spread around the world [1, 2] . as of 5 june 2020, the who has reported a total of 6 535 354 covid-19 cases and 387 155 deaths globally, with an average mortality of 5.92% and the person-to-person transmission is still continuing [3] . the clinical spectrum of covid-19 appears to be wide, ranging from asymptomatic infection to mildly, severely, critically ill cases. mild patients present only upper respiratory tract symptoms like cough and fever, however, respiratory failure, acute respiratory distress syndrome, heart failure, septic shock and even death can be observed in patients with critical conditions [4] . although most confirmed patients (81%) were classified as mild or moderate, 14% were severe and 5% were critical according to the largest investigation of 72 314 cases to date [5] . accumulated evidences have indicated that older age, male, smoking, comorbidity, neutrophilia, coagulopathy, elevated d-dimer level and organ dysfunction were associated with increased risk of death from covid-19 [5] [6] [7] [8] [9] [10] . however, investigations of determinants of severity of covid-19 are scarce. early detecting cases with the potential deterioration of disease will contribute to optimising the use of limited health resources and allocating the proper care. liang et al. developed a clinical risk score to predict the occurrence of critical covid-19 based on severe or non-severe [11] . to our knowledge, no previous studies have been conducted to investigate the risk factors of severity of covid-19 based on ordinal response, namely moderate, severe and critical illness. the estimation of risk factors of disease severity is therefore not very robust. here, we conducted a retrospective study based on covid-19 patients from four designated hospitals in shanghai, hubei and anhui provinces to describe the clinical features of covid-19, and aimed to identify the predictors of multi-level response of severity from moderate, severe to critical illness. this multi-centre retrospective study encompassed covid-19 patients classified as being moderately, severely and critically ill. the illness severity of covid-19 was defined according to the guideline on the diagnosis and treatment of covid-19 by the national health commission (v.5) as described previously [12] . patients were admitted to shanghai public health clinical center, wuhan jinyintan hospital and tongji hospital of tongji medical college hust in hubei province and tongling municipal people's hospital in anhui province from 1 january 2020 to 8 march 2020. all patients recruited in this study were laboratory-confirmed covid-19. the study was approved by the ethics committees of these four hospitals, respectively. written informed consent was waived owing to the need of rapid emergency response to this infectious disease. medical records of covid-19 patients were reviewed by the research team, and demographic, epidemiological, clinical, laboratory, treatment and outcome data were retrieved from electronic medical records using a standardised case report form. all data were cross-checked by two experienced doctors. to ascertain the medical histories or epidemiological data, which were unavailable from electronic medical records, the patients or their close relatives were interviewed by researchers. data from the medical records were adopted if there was a discrepancy between the subjective description and the medical records. method of laboratory confirmation of severe acute respiratory syndrome coronavirus 2 (sars-cov-2) has been described elsewhere [1] . simply, the chinese center for disease control and prevention (cdc) and local cdc were in charge of detecting sars-cov-2 in throat-swab specimens from the upper respiratory tract by real-time reverse transcription polymerase chain reaction assay (rt-pcr). the criteria of discharge included absence of fever for at least 3 days, remission of respiratory symptoms, complete improvement in bilateral lungs in chest ct, together with negative for 2 times in throat-swab samples for sars-cov-2 rna at least 24 h apart. initial clinical laboratory examinations involved complete blood count, serum biochemical tests (including liver and kidney functions, creatine kinase, lactate dehydrogenase (ldh) and electrolytes), myocardial enzymes, d-dimer and procalcitonin (pct). frequency of examinations was under the discretion of treating physicians. chest computed tomographic (ct) scans were carried out for all covid-19 patients. two radiologists were invited to interpret chest ct scans independently and were blinded to the severity of the patient. when disagreement arose, a third radiologist was consulted to reach a final decision. continuous variables were presented as median with interquartile range (iqr) and the analysis of variance (anova) or kruskal −wallis h test were used to compare the difference among three groups as appropriate. categorical variables were expressed as frequency with percentages, and were analysed by pearson's χ 2 test or fisher's exact test. bonferroni's correction was used for pairwise comparison. all patients were divided into moderate, severe and critical illness groups. potential predictive variables included the following case characteristics on admission: demographic and epidemiological features, comorbidity, clinical signs and symptoms, laboratory findings and chest imaging results. to explore the risk factors associated with illness severity of covid-19, namely moderately, severely and critically ill, which means the response variable was ordinally scaled, a cumulative logit model was used to investigate the effect of predictors of covid-19 severity. imputation for missing variables of some patients at hospital admission was considered if missing values were less than 20%, and imputation based on the expectation −maximisation algorithm method was used to replace missing values. before ordinal logistic regression model was fitted, continuous variables of laboratory findings were transformed into categorical variables according to their reference values. the univariate and multivariate cumulative logit models were fitted with moderate illness as the reference level. potential predictors of severity were investigated using univariate ordinal logistic regression firstly. we further conducted a backward stepwise multivariate ordinal logistic regression analysis excluding variables which were not significant in univariate cumulative logit model. since missing rate of 34.6% occurred in the lung imaging results and over 40% existed in urine protein and urine glucose, these variables were excluded from multivariate ordinal logistic model. the overall survival (os) was estimated using the method of kaplan−meier and the log-rank test was applied to compare the survival difference among different severity illness groups. the hazard ratio with 95% confidence interval (ci) was estimated with cox proportional hazard model. a two-sided α of less than 0.05 was considered statistically significant. all statistical analyses were conducted using sas software (v. 9.4) (sas institute inc., usa). as of 28 april 2020, data from 598 covid-19 cases admitted to these four hospitals, including 400 (66.89%) moderate cases, 85 (14.21%) severe cases and 113 (18.90%) critical cases, had been collected to be incorporated into this study, of whom 79 cases had died during hospitalisation, with an average mortality of 13.21%, and 457 cases had recovered and been discharged. the remaining 62 cases were still in hospitals. the median age of the 598 patients was 57 years (iqr 42-66), ranging from 11 to 89 years, and 58.03% patients were male (table 1 ). at least one comorbidity was present in 51.54% of patients, with hypertension being the most frequent comorbidity (33.90%), followed by diabetes (13.18%) and cardiovascular disease (6.51%). few cases had a current (7.53%) or former (2.57%) smoking habit. the most common symptoms on admission were fever (81.22%) and dry cough (33.63%), followed by sputum production (30.77%) and shortness of breath (26.65%). overall, the median interval between illness onset and confirmed diagnosis was 4 days (iqr 2-7), whereas the median interval between illness onset and admission was 7 days (iqr 3-11). the substantial differences of laboratory findings on admission among three groups of patients were observed ( table 2 ). the results indicated that significantly higher proportion of patients showing abnormal alanine aminotransferase (alt) and aspartate aminotransferase (ast), and significantly higher levels of c-reactive protein, neutrophil count, serum potassium, cardiac troponin i (ctni), myohaemoglobin, pct, brain natriuretic peptide (bnp) and d-dimer were observed in the critically ill group and the severely ill group than in the moderately ill group, whereas, the levels of haemoglobin and serum albumin were significantly lower in critically ill group and severely ill group as compared with moderately ill group (p < 0.05). the critical group showed a significantly higher level of platelet count, fibrinogen and serum calcium as compared with severe group and moderate group. the severe group showed a significantly lower level of lymphocyte count than the moderate group. the normal ranges of laboratory indicators are shown in supplementary table s1 . abnormalities on chest radiographs on admission were seen in most patients (table 2) . overall, typical findings on chest ct images were ground-glass opacity (94.37%), followed by pleural thickening (47.06%) and consolidation (35.74%). chest ct scans showed significantly higher percentage of bilateral lungs involvement in critical group (92.45%) and severe group (95.24%) as compared with moderate group (80.07%). the median lung lobes involved in critical group (5, iqr 5-5) and severe group (5, iqr 2.75-5) were greater than those in moderate group (4, iqr 2-5). totally, 320 (61.07%) patients were given antivirals within 2 days after admission including lopinavir/ritonavir, arbidol, darunavir and chloroquine. in all, 369 (70.83%) patients received antibiotics and 117 (23.08%) received corticosteroids. more patients received corticosteroids in critical group and severe group as compared with the moderate group (p < 0.05). the proportions of patients accepting high-flow nasal cannula oxygen therapy and non-invasive mechanical ventilation, respectively, in critical and severe groups were significantly higher than in mode rate group days from illness onset to admission 7 (3−11) 6 (3−10)* † 10 (6−14) 10 (0−14) <0.0001 copd, chronic obstructive pulmonary disease. data are median (iqr) or n/ total (%). p value denotes the comparison among moderate, severe and critical illness group. * and †signify p < 0.05 for pos-hoc comparison. *refers to comparison between the critical group and the severe group or the moderate group. †refers to comparison between the severe group and the moderate group. kandi xu et al. (p < 0.05). compared with the moderate group, the critical group and the severe group had a significantly lower rate of discharge and a higher mortality rate (p < 0.05), as shown in table 3 . the comparison of demographic and baseline characteristics, symptoms, laboratory parameters, lung image features, treatment and prognosis among moderately, severely and critically ill patients were shown in tables 1-3 . fifty-three variables on admission were successively included in the univariate ordinal logistic regression, and 35 variables were found to be associated with illness severity, including age, gender, hypertension, diabetes, interval between illness onset and diagnosis, interval between illness onset and admission, pharyngodynia, shortness of breath, early administration of antiviral, c-reactive protein, white blood cell (wbc) count, neutrophil count, lymphocyte count, haemoglobin, platelet count, alt, ast, albumin, blood urea nitrogen, creatinine, potassium, ldh, creatine kinase, myohaemoglobin, troponin i (ctni), pct, erythrocyte sedimentation rate (esr), bnp, fibrinogen, d-dimer, bilateral lungs involved, consolidation, linear opacity, pleural thickening and lung lobes involved (table 4) . except five lung image variables, 30 significant predictors of severity in univariable analysis were included in a multivariable stepwise cumulative logit model, and seven variables retained in the final model which were statistically significant independent determinants of covid-19 illness severity ( table 5 ). the results of multivariate model revealed that the risks of having more severe illness were 1.586 (95% ci: 0.824-3.053) and 3.419 (95% ci: 1.596-7.323) times higher among patients belonging to the age group 40-69 and 70+ years, respectively, when compared with patients of less than 40 years. patients with hypertension had 3.372 (95% ci: 2.185-5.202) times greater risk of having worse severity of illness compared with patients without hypertension. the risk of having worse severity of illness was found higher for patients with alt>50 μ/l (or = 3.304; 95% ci: 2.107-5.180) when compared with those having alt⩽50 μ/l. the risk of having worse severity of illness was found significantly higher for patients with higher ctni (>0.04 ng/ml) than those with normal ctni (⩽0.04 ng/ml) with or being 7.464 (95% ci: 4.292-12.980). covid-19 patients with myohaemoglobin>48.8 ng/ml at admission had a 2.214 (95% ci: 1.42-3.453)-fold greater risk of having worse severity of illness when comparison was made with patients having normal myohaemoglobin level. table 5 also shows that interval between illness onset and diagnosis and interval between illness onset and admission were independent significant predictors of illness severity with or being 1.056 (95% ci: 1.012-1.101) and 1.048 (95% ci: 1.009-1.087), respectively. in the critical illness group, 68 of 113 patients (60.18%) died, as compared with five of 85 (5.88%) in the severe illness group and six out of 400 (1.50%) in the moderate illness group. the median os was 29 days in critical group, as compared with not attainable in severe group and moderate group. the 30-day os rates were 97.7% (95% ci: 95.5-100%), 95.3% (95% ci: 90.2-100%) and 46.4% (95% ci: 37.4-57.6%) in moderate, severe and critical groups, respectively (p < 0.001), as shown in data are median (iqr) or n/ total (%). p value denotes the comparison among moderate, severe and critical illness group. * and †signify p < 0.05 for post-hoc comparison. *refers to comparison between the critical group and the severe group or the moderate group. †refers to comparison between the severe group and the moderate group. figs s1-s5 ) demonstrated that the differences of os with respect to stratification by these risk factors were all statistically significant (p < 0.05). covid-19 has presented an unprecedented challenge to the health-care system across the world. with the increasingly scarce health resources, mortality is the most important issue when dealing with epidemics. early identifying potential of severe and critical patients becomes the priority in minimising the mortality and contributes to allocation of limited critical care. previously, a risk forecasting model to predict the occurrence of critical illness among hospitalised covid-19 patients in china has been reported by liang et al. [11] . the response variable in liang's model had just two levels of severe and non-severe. actually, the mortalities of covid-19 patients with different severity are variant. this study demonstrated the mortality of 60.18% in critical cases, followed by 5.88% in severe cases and 1.50% in alt, elevated ctni, elevated myohaemoglobin, together with prolonged interval between illness onset and diagnosis and interval between illness onset and admission were independent determinants of severity of covid-19 and represented higher odds of worse severity of illness. evidence is gradually accumulating with regard to the risk factors associated with severity of covid-19. as described previously, older age has been reported as an important independent predictor of severity in covid-19 patients [5, 6, 11] , which is proved by this study. comorbidity of hypertension was found to be associated with an increased risk of death. a meta-analysis including 46 248 patients with confirmed covid-19 indicated that those with the most severe illness were more likely to have hypertension with or of 2.36 (95% ci: 1.46-3.83). hypertension was reported to increase the or for death by 3.05 (95% ci: 1.57-5.92) in patients with covid-19 [4] . similarly, patients with hypertension had 3.372 (95% ci: 2.185-5.202) times greater risk of developing worse severity of illness in comparison with those without hypertension in our study. the relationship between hypertension and covid-19 may relate to the role of angiotensin converting enzyme (ace2) [13] . as a key element in the renin-angiotensin-aldosterone system (raas), ace2 is critically involved in the pathophysiology of hypertension. studies demonstrated that inhibition of the raas with ace inhibitors (aceis) or angiotensin ii receptor blockers (arbs) may result in a compensatory increase in tissue levels of ace2 and poorer clinical course and prognosis, leading to suggestions that these drugs may be detrimental to covid-19 patients [14] . it has been reported that covid-19 had significant impact on the liver function [15, 16] . a meta-analysis including 20 retrospective studies with 3428 covid-19 patients revealed that higher serum levels of ast (mean difference = 8.84 u/l, 95% ci: 5.97-11.71, p < 0.001) and alt (mean difference = 7.35 u/l, 95% ci: 4.77-9.93, p < 0.001) and lower serum levels of albumin (mean difference = −4.24 g/l, 95% ci: −6.20 to −2.28, p < 0.001) were associated with a significant increase in the severity of covid-19 [17] . our univariate logistic regression results also showed patients with elevated ast, alt and decreased albumin had 3.410-, 3.783-and 5.734-fold greater risk of worse severity, which was consistent with this meta-analysis. particularly, our multivariate ordinal regression demonstrated that alt is an independent predictor of severity of covid-19 patients. since elevated liver injury indicators are strongly associated with the severity risk and subsequent death risk, the liver function should be monitored during hospitalisation. acute myocardial injury is the most commonly described cardiovascular complication in covid-19 [18] . the overall incidence of acute myocardial injury has been variable but roughly 8-12% of covid-19 patients are found to develop significant elevation of ctni [19] . the patients admitted to icu or having severe/fatal illness have several-fold higher likelihood of ctni elevation. several recent studies indicated that higher concentration of ctni and myohaemoglobin were associated with the severity and case fatality rate of covid-19 [20] [21] [22] . chen et al. reported that elevated ctni (or = 26.909, 95%ci: 4.086-177.226, p = 0.001) were the independent risk factors of critical disease status [20] . han et al., found there were statistically significant differences in the level and positive rate of ctni and myohaemoglobin among the mild, severe and critical covid-19 case groups [23] . our results are in agreement with the previous studies that the elevated myocardial injury markers such as ctni and myohaemoglobin are independent determinants of illness severity in covid-19 patients representing negative clinical course and potentially life-threatening prognosis. in particular, ctni is the strongest predictor of worse severity with or being 7.464 (95% ci: 4.292-12.980). direct myocardial injury due to viral myocarditis or the effect of systemic inflammation appears to be the most common mechanisms of acute cardiac injury. our study also found that interval between illness onset and diagnosis and interval between illness onset and admission were independently associated with illness severity. the risk of developing worse illness severity increased by 1.056-fold and 1.048-fold for each day delay of interval between illness onset and diagnosis and interval between illness onset and admission, respectively. chen et al. [24] found the median time from symptom onset to admission was 10 days among deceased covid-19 patients, one day longer than that of discharged patients. considering the limited health resources and high case volume in some regions currently, adjusted tactics and strategies should be taken to maximise the availability of and accessibility to medical service to shorten the diagnosis delay or admission delay. our study had several limitations. first, this is a retrospective study design, which could be subject to recall bias and selection bias. second, not all laboratory parameters were tested in all patients, including ldh and d-dimer. although imputation technique was used to replace missing values, their role might be underestimated in predicting illness severity. third, due to massive loss of chest ct results, the predicting role of chest ct abnormalities could not be evaluated in this study. its role of predicting critical illness of covid-19 has been demonstrated by other studies. last but not least, generalisability of our findings might be limited by the sample size, and the results need to be validated based on a much larger patient population. in this study, we identified older age, presence of hypertension, elevated alt, ctni and myohaemoglobin, prolonged interval between illness onset and diagnosis and admission as the independent determinants to predict the risk of developing more severe illness among covid-19 patients. given the ongoing global pandemic of covid-19, this study will contribute to early identifications of patients with high risk of developing critical illness and optimising the arrangement of health resources. supplementary material. the supplementary material for this article can be found at https://doi.org/10.1017/s0950268820001533. clinical features of patients infected with 2019 novel coronavirus in wuhan epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in wuhan, china: a descriptive study world health organization (2020) coronavirus disease (covid-19) situation report -137 clinical course and risk factors for mortality of adult inpatients with covid-19 in wuhan, china: a retrospective cohort study characteristics of and important lessons from the coronavirus disease 2019 (covid-19) outbreak in china: summary of a report of 72314 cases from the chinese center for disease control and prevention risk factors associated with acute respiratory distress syndrome and death in patients with corona virus disease coronavirus disease 2019 (covid-19) in italy comorbidity and its impact on 1590 patients with covid-19 in china: a nationwide analysis clinical findings of patients with coronavirus disease clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan, china development and validation of a clinical risk score to predict the occurrence of critical illness in hospitalized patients with covid-19 covid-19 with different severity: a multi-center study of clinical features hypertension, the renin-angiotensin system, and the risk of lower respiratory tract infections and lung injury: implications for covid-19. european society of hypertension covid-19 task force review of evidence renin-angiotensin-aldosterone system inhibitors in patients with covid-19 liver impairment in covid-19 patients: a retrospective analysis of 115 cases from a single centre in wuhan city longitudinal association between markers of liver injury and mortality in covid-19 in china liver injury is associated with severe coronavirus disease 2019 (covid-19) infection: a systematic review and meta-analysis of retrospective studies 2020) cardiovascular disease and covid-19 laboratory abnormalities in patients with covid-2019 infection analysis of myocardial injury in patients with covid-19 and association between concomitant cardiovascular diseases and severity of covid-19 characteristics and outcomes of patients hospitalized for covid-19 and cardiac disease in northern italy clinical utility of cardiac troponin measurement in covid-19 infection analysis of heart injury laboratory parameters in 273 covid-19 patients in one hospital in wuhan clinical characteristics of 113 deceased patients with coronavirus disease 2019: retrospective study acknowledgements. we acknowledge all health-care workers involved in the diagnosis and treatment of patients in four sites and we thank dr qiqi cao and dr weixia li for assistance in the analysis and interpretation of the lung image data. dr jian li had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. conflict of interests. we declare no competing interests. kandi xu et al.data availability statements. the data that support the findings of this study are available from clinical research center, ruijin hospital, shanghai jiao tong university school of medicine. restrictions apply to the availability of these data, which were used under license for this study. key: cord-307834-shmpfnrj authors: massad, eduardo; amaku, marcos; wilder-smith, annelies; costa dos santos, paulo cesar; struchiner, claudio jose; coutinho, francisco antonio bezerra title: two complementary model-based methods for calculating the risk of international spreading of a novel virus from the outbreak epicentre. the case of covid-19 date: 2020-06-09 journal: epidemiol infect doi: 10.1017/s0950268820001223 sha: doc_id: 307834 cord_uid: shmpfnrj we present two complementary model-based methods for calculating the risk of international spread of the novel coronavirus sars-cov-2 from the outbreak epicentre. one model aims to calculate the number of cases that would be exported from an endemic country to disease-free regions by travellers. the second model calculates the probability that an infected traveller will generate at least one secondary autochthonous case in the visited country. although this paper focuses on the data from china, our methods can be adapted to calculate the risk of importation and subsequent outbreaks. we found an average r(0) = 5.31 (ranging from 4.08 to 7.91) and a risk of spreading of 0.75 latent individuals per 1000 travellers. in addition, one infective traveller would be able to generate at least one secondary autochthonous case in the visited country with a probability of 23%. given the extent of global travel patterns [1] [2] [3] , newly emerging diseases can rapidly spread globally. in general, respiratory pathogens spread faster [4] than vector-borne viruses [5, 6] or those that require very close contact such as ebola [7] or lassa [8] . in late 2019, a novel coronavirus, severe acute respiratory syndrome coronavirus 2 (sars-cov-2) emerged in wuhan, china, which rapidly spread globally [9] and caused epicentres of covid-19 disease in multiple countries. sars-cov-2 has a high reproduction rate and is easily transmitted via respiratory droplets among humans [10] . population flow data between wuhan and other major cities in mainland china were clearly correlated with the number of cases exported from wuhan to other city clusters in mainland china before the lock-down [11] . the potential for rapid international spread via air travel was enormous, and indeed first exportations followed high travel volumes to thailand, hong kong and singapore [9] . as of 1 june 2020, more than 6 million cases and more than 360 000 deaths due to covid-19 have been reported in more than 200 countries. the speed of spread depends on the air passenger volumes, the basic reproduction rate as a measure of transmissibility and the incubation time [12, 13] . in this paper we present two complementary methods for calculating the risk of international spread of a new virus from an epicentre. the first method aims to calculate the number of cases that would be exported from an endemic country to disease-free regions by travellers. the second method calculates the probability that one of the infected travellers will generate at least one secondary autochthonous case in the visited country. the calculation for disease exportation is simpler than the calculation for infection importation. one difference is that in the case of disease importation travellers to endemic areas return infective to their home country, whereas in the case of disease exportation travellers depart from their endemic home country in a latent state. this latter assumption is based on the conjecture that symptomatic individuals do not travel. for asymptomatic travellers, their disease will manifest itself either during the flight or after arrival in the visited disease-free country dependent on the time of infection and incubation time. in the case of disease importation, the key parameter is the force of infection of the disease in the visited endemic country. in the case of disease exportation, the key parameter is the latency duration of the disease in the travellers' home country. in the case of disease importation latency is not too important and the model considers only susceptible, infected and removed individuals. on the other hand, in the case of disease exportation latency is important because it is assumed that infected and symptomatic individuals are either so sick that they do not manage to travel or are not allowed to board the plane due to exit screening. the models model 1. calculating the number of exported cases from an endemic country in this section we consider the case of infective travellers from an endemic country visiting a disease-free country, and so exporting the infection to the visited country. once arriving in the visited disease-free country those infective visitors may trigger an outbreak that can establish itself depending on the value of the basic reproduction number r 0 of the infection in the disease-free country. if r 0 is greater than one, the disease will spread. we will approach the problem with a deterministic formulation. the model is a classic susceptible-exposed-infected-removed (seir) model given by the following set of equations: where s h (t) is the number of susceptible individuals, e h (t) is the number of incubating and asymptomatic individuals, who have the disease but do not transmit it, i h (t) is the number of infectious individuals, r h (t) is the number of individuals recovered from infection and n h (t) = s h (t) + e h (t) + i h (t) + r h (t) is the total population. the parameters are β, the potentially infective contact rate, δ h , the inverse of the incubation (or latency) period, γ h , the duration of infectiousness and μ h and α h are the natural and disease-induced mortality rates, respectively. hence, the number of new infections per unit of time corresponds to the infection incidence, denotedλ(t) = βs h (i h (t)/n h (t)). the basic reproduction number, r 0 , that is, the number of secondary infection produced by an infectious individual in an entirely susceptible population along his/her infectiousness period, associated with system (1) is deduced in appendix a: for exportations, our interest is the prevalence of latent infections in the local population, from which, some individuals will travel already infected but not yet symptomatic. we estimated the disease prevalence in the population, that is, the number of infected individuals at each instant of time, i h (t), by integrating the third equation of equation (1) to obtain [3] : dividing i h (t) by the size of the local population, n h , we obtain the prevalence, that is, the proportion of infectious individuals, p i (t), in the endemic country as follows: on the other hand, multiplying the number of visitors to a given disease-free country by the prevalence of latent (infected but not infectious individuals), p e (t), generates the number of infected visitors or exportations of infections. integrating the second equation of (1) yields the following quantity e h (t), exposed or latent individuals: dividing e h (t) the total population n h , yields the prevalence of infected but not yet infectious individuals in the home country as follows: to obtain this prevalence, the force of infection of the disease, that is, the number of new cases of infection per time unit, β (i h (t)/n h ), in this endemic region is a necessary input variable. the best information normally available is the notification rate of infectious individuals, δ h e h (this term is the number of individuals that evolve from the latent to the infectious state), provided by disease surveillance systems. equation (6) will be used later in the paper. in this section we calculate the probability that an infected traveller (index case) from an endemic country arriving infective in a disease-free country generates a secondary autochthonous case. on arriving in the disease-free country each infected visitor will trigger an outbreak that will establish itself depending if the value of the basic reproduction number r 0 of the infection is greater than one. since we are dealing with a low number of travellers, we need to approach the problem with a stochastic formulation. this model assumes that a density of one infected individual, i h (t 0 ), arrives at t = t 0 and remains infective for a period of (μ h + γ h + α h ) −1 days, that is where μ h , γ h and α h are the natural mortality rate, the recovery rate from infection and the disease-induced mortality rate, respectively. if the region where these infected travellers arrive had an area a the number of them is i h (t 0 )a. eduardo massad et al. the total number of new cases infected by these travellers, δ weeks after its introduction, new cases, is given by where β is the potentially infective contact per unit time between infected and one susceptible individuals, and s h and n h are the susceptible and total population, respectively. the risk of new cases invasion of a previously unaffected country, risk new cases , can be defined as the probability that at least one autochthonous case be produced by the arrival of one single infected individual at the area during his/her infectiousness period. for calculating this risk, we assumed a non-homogeneous simple birth process [5] , which describes the propagation of the disease. let p n (t) be the probability of n cases. the probability generating function of such process is p(x, t) = n p n x n . after some calculation, we obtain the probability of x cases at time t [5, 6] : we have assumed that the region to be studied has an area a, the number of infected travellers that arrived at t = t 0 is a = i h (t 0 ) a. we set a = 1 from now on, that is, a single index case arrives at the non-affected area. expanding (9) in powers of x we find that the risk, that is, the probability of having n infected individuals at time t, denoted risk new cases (n, t) as the risk (probability) of having no infected individuals is in equations (9a) and (9b) the probability of at least one autochthonous case in a previously unaffected region can be calculated as the tail probability, that is, the probability of the infection invading the previously non-affected area: for m = 1, equation (5) to illustrate the models' performance, we consider the case of the outbreak of covid-19 in the province of hubei, china. at the time of writing, this province was responsible for approximately 82% of the total number of covid-19 cases in the world. we used data from the who website [7] . as we use the case of hubei outbreak only to illustrate the models, we calculated the incidence of cases in that province by multiplying the total world number of daily cases of infections by 0.84. as china modified the diagnostic criteria along the course of the outbreak, we used incidence data only until 11 february 2020. we begin by fitting a continuous function to the daily number of reported cases, that is, the incidence of new cases per time unit, δ h e h (t). we assume here that new cases are symptomatic cases. this assumption is probably very reasonable for the hubei epidemic because tests were developed only at the end of the hubei outbreak. the function has the bell-shaped form: where c i , (i = 1, …, 3) are the fitting parameters and t is the time. the fitting of reported new cases per time unit to equation (11) is shown in figure 1 . note that equation (12) fits the hubei covid-19 cases reasonably well. if equation (12) is inserted into equation (4), the covid-19 prevalence at each instant of time i h (t) is obtained. on the other hand, we can fit the initial exponential phase (e ϕt ) of the prevalence curve obtained in figure 1 . it is then possible to estimate the value of r 0 according to (see appendices a and b): where w is the rate of the exponential growth and m h , a h , g h and δ h as in equation (1). the parameter values used to calculate r 0 are shown in table 1 . the value of the estimated r 0 resulted in an average of 5.31 (ranging from 4.08 to 7.91) for the outbreak in the province of hubei. figure 2 shows the fitting of data from hubei province to an exponential function. it is possible, in principle, to fit the parameters of system (1) in order to retrieve the prevalence curve. the parameters then can be used to estimate the number and the prevalence of latent individuals (equations (5) and (6)). alternatively, taken the covid-19 average latency period of 3 days (i.e. δ h = (1/3) per day), e h (t) can be calculated by simply dividing equation (11) by δ h , that is, from which it is possible to estimate the prevalence of asymptomatic latents in the population: the result for the case of covid-19 in the province of hubei is shown in figure 3 . as an example in a cohort of travellers that depart from hubei at week 15 the relative number of latent individuals carrying the covid-19 virus is of 0.75 individuals per 1000 travellers, which is much higher than ebola, for instance [7] . in other words, out of 1333 travellers from that region, 1 would be infected. next, we assume that one individual traveller from an endemic country (index case) visits a disease-free country and remains infective for a period of (μ h + γ h + α h ) −1 days. to calculate the probability of a secondary autochthonous case generated by each infected traveller, we used the incidence curve described above to calculate the probability generating function according to equation (9) where the incidence is represented by the parameter λ(t). next, we calculate the values of parameters π and σ from equations (9a) and (9b) to estimate the probability that the infected traveller who imported the virus to his/her home country would generate at least one secondary case, according to equation (11) . the result is 23%, that is, one single infective traveller would be able to generate at least one secondary case along his/her infectiousness period, with probability of 23%. note that the expected number of secondary case is the average value of the basic reproductive rate, that is, 5.31. in this paper we propose two complementary models for calculating the risk of international spreading of the novel coronavirus sars-cov-2 from the initial epicentre of covid-19 in wuhan, china. one model addresses the case of disease exportation from the epidemic outbreak and considers a certain number of travellers leaving the epidemic region during the incubation period, thereby importing the virus into another country. the model is deterministic and was illustrated with the data from the initial outbreak in the province of hubei in china. the first model's simulation resulted in an average r 0 = 5.31 (ranging from 4.08 to 7.91) and a risk of spreading of 0.75 latent individuals per 1000 travellers. if we consider the monthly number of travellers from the city of wuhan described by wu et al. [14] to other asian countries of around 86 000, we should expect almost 65 cases of the infection to these countries. the second model addresses the case of the probability of disease introduction in a disease-free country by an index case from the epidemic epicentre. the model considers the situation in which a single infected traveller from an epidemic region, acquires the infection and travels to a disease-free country where he/she can trigger a local outbreak. as we consider a single traveller we approach the case with a stochastic formulation. we simulated the model with the same case of the province of hubei in china and the results show that one single infective traveller would be able to generate at least one secondary autochthonous case in the visited country, along his/her infectiousness period, with a probability of 23%. this probability should be contrasted with the average number of secondary cases the traveller would generate at his/ her home country of 5.31. the latter is the average basic reproduction number of covid-19 in the community of hubei and should not be essentially different elsewhere when the population is immunologically naïve and there is a homogenously mixing pattern of contact. in a stochastic context, even when r 0 is greater than 1, there is a probability of extinction of the infection. moreover, the 23% risk of exportation means the probability that one traveller when arriving in the infectious condition would generate at least one secondary autochthonous case of covid-19. some important limitations are worthwhile mentioning about our approach: our model assumes that only latent individuals travel. however, it is possible that some mildly symptomatic cases can escape from the screening measures at the moment of the travel. moreover, for example, a number of the earliest known exported cases travelled when sick. there are reports of travellers taking antipyretics to mask their fever, and then board the plane. however, the number of patients who travel with mild symptoms is likely to be very small when compared to the non-symptomatic latent individuals and this should not interfere with our results. furthermore, our results will depend on the incidence of covid-19 in the departing country. for example, during the peak of the covid-19 outbreak in europe, about 3−6% of air passengers were sars-cov-2 positive on repatriation flights [15] . from the modelling perspective an important limitation is the homogeneously mixing assumption. we are well aware of the many heterogeneities involved in transmission of a directly transmitted pathogen like sars-cov-2. in addition, the deterministic approach of the exportation model is an approximation of the real dynamics involved in transmission. however, both limitations above do not invalidate the qualitative results of the models. considering the large number of people involved in the current epidemic the deterministic approach and the homogeneously mixing assumption can be considered as a good first approximation of the problem. however, heterogeneities can be introduced in the model using the techniques described in [16] , and these heterogeneities could have significant influences in the quantitative results of our model. for instance, if variation in infectiousness would be included, the risk of spread could be lower on average, and the speed of the infection spread could be affected as well. equation (1) assumes that only infected individuals, i h , are infectious. in fact, at least a fraction f of the exposed individuals, e h , may be infectious. so i h in the first and the second equations of system (1) should be replaced with i h + fe h . equation (a6) in appendix a shows that e h (0) and i h (0) are related. this artificial feature can be removed by adding to equations (a1) and (a2) the initial infection terms e h (0)e −(m h +dh)t and i h (0)e −(m h +ah+g h )t , respectively, and solving them [10] . another limitation of our approach concerns the data upon which we exemplify our application. we have access only to the global number of cases from the who website on covid-19 and in order to apply the model for the province of hubei we assumed that that region represents 84% of the global cases. hence, we assumed a direct proportionality of the cases to simulate the model on these data. in addition, we simulate the model until 11 february because china modified the diagnostic criteria along the course of the outbreak. the incidence curve, however, had already started to wane at that time. moreover, we used the incidence data for the province of hubei only to exemplify the models, which could be applied, in principle, to other situations related to the spread of pathogens from outbreak epicentres. finally, it should be commented that the model assumes that the potentially infective contact rate β in the receptive country of the index cases is the same as in the province of hubei. in fact, β has a remarkable seasonality, being higher during winter time, declining throughout spring time into summer time. therefore, the probability of autochthonous cases is superestimated for receptive countries during the summer season. we believe that the models presented here may present a significant step forward in estimating the risk of importation of the novel coronavirus sars-cov-2. financial support. this work was partially supported by the project zikaplan, funded by the european union's horizon 2020 research and innovation programme under grant agreement no. 734584, by lim01-hfmusp, cnpq andfapesp and fundacao butantan. global travel patterns: an overview global trends in air travel: implications for connectivity and resilience to infectious disease threats mapping global variation in human mobility pandemic preparedness and response − lessons from the h1n1 influenza of imported dengue in spain: a nationwide analysis with predictive time series analyses severe dengue in travellers: pathogenesis, risk and clinical management ebola virus outbreak in north kivu and ituri provinces, democratic republic of congo, and the potential for further transmission through commercial air travel fifty years of imported lassa fever − a systematic review of primary and secondary cases potential for global spread of a novel coronavirus from china the reproductive number of covid-19 is higher compared to sars coronavirus the positive impact of lockdown in wuhan on containing the covid-19 outbreak in china estimating the probability of dengue virus introduction and secondary autochthonous cases in europe modeling importations and exportations of infectious diseases via travelers nowcasting and forecasting the potential domestic and international spread of the 2019-ncov outbreak originating in wuhan, china: a modelling study high prevalence of sars-cov-2 infection in repatriation flights to greece from three european countries modelling heterogeneities in individual frailties in epidemic models ethical standards. not applicable (this is a purely theoretical work with no human subject involved). finding the basic reproduction number given by equation (2) linearising the second and the third equations of the system (1) assumingthe above system has non-trivial solution if the determinant of the unknown c 1 and c 2 are different from zero. we then obtain is greater than 1. thus, r 0 is the basic reproduction number. also, now we relate the basic reproduction number (equation (a5)) to the rate w of new cases n c (t) given by equationthe rate w is obtained by fitting equation (b1) to the initial exponential growth shown in figure 2 .using equation (a4) we have that the number of new infected cases at t = 0 is given bysince e h (t) and i h (t) are given by equations (a3) key: cord-354474-hbl2ywix authors: temsah, m. h.; alhuzaimi, a. n.; alamro, n.; alrabiaah, a.; al-sohime, f.; alhasan, k.; kari, j. a.; almaghlouth, i.; aljamaan, f.; al-eyadhy, a.; jamal, a.; al amri, m.; barry, m.; al-subaie, s.; somily, a. m.; al-zamil, f. title: knowledge, attitudes and practices of healthcare workers during the early covid-19 pandemic in a main, academic tertiary care centre in saudi arabia date: 2020-08-28 journal: epidemiol infect doi: 10.1017/s0950268820001958 sha: doc_id: 354474 cord_uid: hbl2ywix as the middle east respiratory syndrome coronavirus (mers-cov) continues to occur in small outbreaks in saudi arabia, we aimed to assess the knowledge, attitudes and intended practices of healthcare workers (hcws) during the early stage of the covid-19 pandemic and compare worry levels with previous findings during the mers-cov outbreak in 2015. we sent an adapted version of our previously published mers-cov questionnaire to the same cohort of hcws at a tertiary hospital in saudi arabia. about 40% of our sample had previous experience with confirmed or suspected mers-cov patients, and those had a significantly higher knowledge score (13.16 ± 2.02 vs. 12.58 ± 2.27, p = 0.002) and higher adherence to protective hygienic practices (2.95 ± 0.80 vs. 2.74 ± 0.92, p = 0.003). the knowledge scores on covid-19 were higher in the current cohort than the previous mers-cov outbreak cohort (68% vs. 79.7%, p < 0.001). hcws from the current cohort who felt greater anxiety from covid-19 compared to mers-cov were less likely to have been exposed to mers-cov infected/suspected cases (odds ratio (or) = 0.646, p = 0.042) and were less likely to have attended the hospital awareness campaign on covid-19 (or = 0.654, p = 0.035). we concluded that previous experience with mers-cov was associated with increased knowledge and adherence to protective hygienic practices, and reduction of anxiety towards covid-19. since the world health organization (who) declared coronavirus disease 2019 (covid19) as a pandemic, it has become a major challenging public health problem worldwide [1, 2] . this pandemic has affected all aspects of people's life in almost all nations and among all socioeconomic groups. healthcare workers (hcws) of all types are facing an unprecedented crisis with the rapid spread of covid-19 and severity of the disease in many infected individuals. as such many healthcare systems have been overwhelmed and hcws presented with depression and anxiety [3, 4] . there is a potential shortage of physical resources, such as ventilators and intensive care unit beds, needed to care for surges of critically ill patients [5] ; however, additional medical supplies and beds will be of limited help unless there is an adequate medical workforce. the mental impact of the covid-19 epidemic on general population [6] , psychiatric patients [7] , workers [8] patients [9] , children [10] , older adults [11] and medical students [12] has been reported. however, little attention has been paid to the psychological wellbeing and fatigue levels among hcws [13, 14] . to further understand the knowledge, attitudes and intended practices of hcws during the early stage of the covid-19 pandemic, it is particularly beneficial to obtain their input, especially in an area of the world where other respiratory viral illnesses are either endemic, such as mers-cov, or seasonal, such as influenza. this study was the baseline for a serial, cross-sectional survey among hcws in a tertiary care hospital with a 1000-bed capacity in riyadh, saudi arabia. the hcws had multinational backgrounds; in addition to saudi nationals, they were mostly from the indian subcontinent and the philippines. data were collected between 5 and 16 february 2020, which was just before the presentation of the first case of covid-19 in saudi arabia. the survey was a pilot-validated, self-administered questionnaire that was sent to hcws online. the questionnaire was similar to that used in our previous mers-cov study [15] , with modification and additional questions related to the current covid-19 pandemic. the questions queried the demographic characteristics of the respondents ( job category, age, sex, work area and years of clinical experience) and previous exposure to mers-cov patients, either suspected or confirmed cases. we assessed the following domains for every participant: kap scores [16] , knowledge about covid-19 [17] , hcw attitudes toward infection control measures [18] and hygienic practice change scores. in addition, we assessed [19] the perceived adequacy of covid-19 information and [15] perceived high fear/stress from the covid-19 pandemic as compared to the previous mers-cov outbreak. the hcws knowledge of covid-19 disease was tested using eight true/false questions (supplementary table s1 ). the perceived adequacy of knowledge, hygienic practice changes and hcw attitudes toward infection control measures were assessed using a series of likert-based questions (supplementary tables s2-s4 ). the practice score is measuring the degree of improvement in protective practices after covid-19, from 1 to 4, with 1 representing no change, 2 little change, 3 moderate change and 4 significant change. the attitude score represents the level of agreement with protective hcws' attitudes toward infection control. the score ranges from 1 to 5, with 1 representing 'strongly disagreeing' and 5 representing 'strongly agreeing' with attitudes. additionally, the sources of information on the outbreak and attendance at the covid-19 awareness campaign (educational day conference) that was conducted at the hospital in the first week of february 2020 were evaluated. the participants were asked about their worry level regarding the current covid-19 compared to their worry level from the mers-cov outbreaks. we analysed the data using spss ibm v20 (spss, inc., chicago, il, usa). for all tests, statistical significance was set at p < 0.05. for the five domains we used a summative score to summarise the results from continuous likert's scale-based questions for each participant. an unpaired t test and analysis of variance (anova) analysis followed various post hoc tests to compare the means of different groups. the model was significant based on a model goodness-of-fit hosmer-lemeshow test [χ 2 (8) = 7.4, p = 0.490, model auc roc = 74%, χ 2 (16) = 103.8, p < 0.001]. the pearson's (r) test was used to assess the bivariate associations between the measured hcw scores. fisher's exact tests were used to establish the differences between hcw groups (physicians vs. non-physicians) for nominal variables. we compared this current analysis with data from our previous study conducted during the mers-cov outbreak in 2015 in the same institution [15] . the study was approved by the institutional review board at the college of medicine and king saud university medical city (approval no. 20/0065/irb). the questionnaire was sent to 800 hcws with a 72.8% response rate; 582 hcws completed the questionnaire. the majority of participating hcws were female (75%) and the most common age group was 31-39 years (38.3%, mean 38.6). nurses constituted 62% of our study population. the majority of respondents were hcws working in critical care units (44.8%) followed by outpatient clinics (28%), and inpatients wards (19.4%), see table 1 . hcws used multiple sources of information about covid-19, as shown in figure 1 . the mean knowledge score among the whole sample was 12.75 ± 2.2 of a total score of 16. there was no difference in the knowledge scores for clinical role groups, gender or hospital working areas (table 1 ). nearly 40% of the whole sample attended the hospital's educational day conference. a higher perceived adequacy of covid-19 information (mean: 3.97 ± 1.00 vs. 3.7 ± 0.92, p = 0.001) and better hygienic practices (3.10 ± 0.84 vs. 2.65 ± 0.87, p < 0.001) were observed among hcws who attended the educational day conference (fig. 2) . however, there was no difference in knowledge scores and attitudes toward selected preventive measures between those who attended and those who did not attend the educational day conference. the mean hygienic practice score was 2.76 ± 0.91 for the whole sample. there was no difference in knowledge scores between males and females. however, female hcws scored higher in terms of adherence to hygienic practices (2.98 ± 0.82 vs. 2.36 ± 0.90, p < 0.001), attitudes toward infection control measures (3.85 ± 0.93 vs. 4.15 ± 1.10, p = 0.003), and perception of adequacy of knowledge (3.88 ± 0.93 vs. 3.58 ± 1.03, p = 0.002; fig. 3 ). there was no significant difference in knowledge scores among different hcws across their different clinical roles. however, nurses had significantly higher hygienic practice scores (3.20 ± 0.68, p < 0.001) compared to all levels of physicians, in addition to significantly higher attitude scores toward infection control practices compared to resident physicians (p = 0.041). taking clinical area assignment into consideration, staff working in critical care units had significantly higher perceived information regarding covid-19, which was reflected by significant higher hygienic behaviour compliance. the younger hcws (⩽30 years) had significantly less adherence to hygienic practices compared to older hcws (31-39 years; p < 0.001) but both the young age group and the group working in critical care area did not show any significant difference in terms of attitude toward infection control practices. fig. 4 ). however, there was no significant difference in the protective attitude toward domestic hygiene among hcws who had previous experience with mers-cov cases compared with the hcws who did not. almost one-third (29.4%) of hcws did not adhere to flu vaccinations. kap scores analysed based on adherence to influenza vaccinations showed higher knowledge mean scores (13.00 ± 2.10 vs. 12.23 ± 2.25, p < 0.001) and hygienic practice scores (3.01 ± 0.81 vs. 2.38 ± 0.89, p < 0.001; fig. 5 ). there was no difference in attitude scores. there was a moderate positive correlation between hcw infection control attitudes and their perceived adequacy of information about covid-19 (r = 0.53, p value <0.01) ( table 2 ). in addition, there was a weak but significant correlation between hygienic practice scores and perceived adequacy of information (r = 0.32, p < 0.01) and with hcw knowledge scores (r = 0.22, p < 0.01). these findings suggested that education plays an important role in improving hcw attitudes and practices to prevent covid-19. hcw's perceived higher fear/stress from covid-19 than from previous mers-cov outbreaks the hcws were asked about their worry level from current covid-19 compared to their worry level from previous mers-cov outbreaks. multivariate logistic regression analysis of the predictors of the hcw's higher stress from covid-19 compared to mers-cov was performed and the results are displayed in table 3 . hcws previously exposed to mers-cov infected/suspected cases were significantly less likely to have higher stress from the covid-19 outbreak (odds ratio (or) = 0.646, p = 0.042). we found that hcws who had attended their hospital educational day on covid-19 were less likely to have stress related to covid-19 compared to stress related to mers-cov (or = 0.654, p = 0.035), accounting for other confounders. hcw worry levels based on contacting covid-19 themselves was associated with higher stress from covid-19 disease compared to mers-cov (or = 1.933 times higher, p < 0.001). hcws' hygienic practice scores converged significantly on higher odds of having high stress from covid-19 compared to mers-cov (or = 1.303, p = 0.046). nonetheless, the hcws' attitude scores correlated positively and significantly with higher odds of being highly stressed from covid-19 disease (or = 1.27, p = 0.032), accounting for the other predictors. we compared our current analysis with data from the previous study conducted in the same institution during the mers-cov outbreak in 2015 [15] . the resulting data analysis (table 4 ) suggested that the hcws' worry levels for contracting covid-19 and passing it on to their families and friends were significantly lower than those measured during the mers-cov (p < 0.001). also, the hcws' worry levels for contracting covid-19 was significantly lower than those measured during the previous mers-cov outbreak in 2015 (p < 0.001). this might be due to in part because this study was done before who declared covid-19 a global pandemic. interestingly, the proportion of hcws who underwent annual influenza vaccination at the current time significantly exceeded those measured during the 2015 mers-cov outbreak; the difference in percentages of hcws who immunised annually indicated an 18.86% rise in the proportion of hcws who immunised against seasonal influenza during this current point of time in 2020 compared to the same period in 2015 (p < 0.001). however, comparing the hcws' perceived hygienic changes and intents to be absent from their work did not differ significantly between the current study and the previous mers-cov (2015) study (p > 0.05 for each). the world has experienced several epidemics with novel coronaviruses; namely, sars-cov-1, which emerged in china in 2003 followed by middle east respiratory syndrome coronavirus (mers-cov) in the middle east in 2012, and the current severe acute respiratory syndrome corona virus-2 (sars-cov-2) pandemic [16] . mers-cov continues to be endemic in saudi arabia with weekly reported cases. with the ongoing circulation of mers-cov and continuing zoonotic spillover with 70% of the cases resulting from hospital outbreak, the emergence of covid-19 within the same setting will be overwhelming to healthcare facilities and workers [17] . therefore, it is of great importance to know the impact of such epidemics on hcws. this is an expected finding since there are established guidelines on the treatment of mers-cov and seasonal influenza and lack of comprehensive knowledge and experience with sars-cov-2. as the understanding of the epidemiology of sars-cov-2 evolved, human-to-human transmission was confirmed with the potential for asymptomatic transmission as well [18] . sars-cov-2 has also demonstrated a very rapid transmission rate with a reported r 0 = 2.5; i.e. each patient can spread the virus to two other patients [20] . hospital transmission was also reported and was estimated to account for 41.3% of cases [19] . this highlights the importance of strict infection control measures and continuous hcw education and competency, not only to decrease transmission but to limit hcw anxiety, which will result in better compliance, performance and patient care. these hcws reported that their main concern was the risk of transmitting the infection to their families (2.71/5) or acquiring it themselves (2.57/5) [21] . in response to the global crisis, as part as hospital preparedness [22] , king saud university medical city (ksumc) arranged an educational day for all hospital staff in an attempt to increase awareness and improve the preparedness of hcws. a survey was distributed afterward and completed by 582 hcws working at king khalid university hospital at ksumc to ensure the adequacy of education. the majority (62.3%) of the responding hcws were nurses, who were approximately half of the attendees (46.2%), and they reported adequacy of the information given regarding symptoms, treatment, prognosis and prevention of covid-19 disease. we observed a better hygienic practices and higher perceived adequacy of covid-19 information among those who attended the hospital's educational day conference. however, only 40% of the whole sample attended this activity. therefore, a different approach is required to improve the attendance of educational activities. suggested approaches include making the covid-19 educational activities periodic, making the attendance mandatory for all staff and utilising alternative or additional online modules. these results showed that a greater proportion (53.8%) still did not get a proper education and, hence, there was a need for more awareness activities and campaigns, including for example but not limited to lectures, departmental educational activity, concise educational leaflets, in addition to newly innovated methods including virtual lectures, smart phone applications (apps) and hotline access during the current pandemic to answer questions and queries. covid-19 pandemic has been accompanied by an overabundance of information that makes it difficult to obtain what is true from false, as most individuals get their information from social media [17] , and this might contribute to higher levels of misconception or anxiety. in this survey, the top source accessed by hcws was hospital announcements (77.8%), which were similar to the main sources used in the previous mers-cov outbreak [15] . this finding again highlights the importance of having a dedicated team to provide accurate information from trusted sources as most hcws rely on it. social network news remained a source of information among 61.7% of the hcws, which might elevate anxiety and should be discouraged in future awareness campaigns. the current study showed that hcws who had previous experience with mers had higher knowledge scores and more adherence to protective hygienic practices. these results could be explained by the fact that previous hospital educational campaigns and managing previous mers-cov cases may have enhanced their knowledge and intentions to be in compliance with infection control practices [23] . another speculation is that the occurrence of mers in saudi arabia is ongoing and there is more awareness about it among hcws [24, 25] . similarly, hcws who were more adherent to receiving the annual influenza vaccine had higher knowledge mean score and higher compliance with hygienic practices. the association between the effect of knowledge, among other multimodal interventions, and compliance with influenza vaccination, has been demonstrated in previous studies, including those conducted in saudi arabia [26] [27] [28] . influenza vaccine utilisation may indicate general awareness and initiative for self-healthcare. most hcws had strong perceptions of the importance of change in hygienic behaviours, including compliance to hand hygiene, infection control measures, avoiding contact with people who have flu-like symptoms, and decreasing handshaking and social visits. the level of knowledge of hcws toward viral infection outbreaks during the current covid-19 pandemic are much higher compared to the previous study conducted in the same institution during mers-cov a few years ago [15] . this finding is promising as it is expected that hcws would be more compliant with infection control measures. for different subgroups we observed higher scores in hygiene practices for hcw attending educational day conference, especially among females and nurses. this is directly reflective on the higher number of nurses attending such activities and the predominance of female nurses in our institution. hcw contribution to proper management of covid-19 infected patients is substantial, as the more knowledgeable they become, the more likely their management to be more appropriate. hcw not wearing masks have a 30% chance of developing infection [29] , meaning at least one in every three hcw not adherent to ppe will get infected. this highlights the importance of sending correct information to enhance their knowledge which would reflect on their attitudes and practices. as of 17 august 2020 within the eastern mediterranean region, the kingdom of saudi arabia ranks second only to the islamic republic of iran in total number of infected and active cases, follow by pakistan then iraq, yet the case fatality rate (cfr) is 1.1 in saudi arabia compare to 5.75, 2.14 and 3.31 in iran, pakistan and iraq, respectively [30] . this lower cfr in comparison to high attack rate can be attributed to the healthcare system in saudi arabia that can be anticipated in a country that dealt with several outbreaks of middle east respiratory syndrome coronavirus (mers-cov). moreover, the strict social distancing and lockdown strategies that were implemented in saudi arabia since the early stages of the pandemic contributed in slowing down the spread of covid-19 and decreased the cfr via preventing the healthcare system from overwhelming covid-19 patients, and thus allowing better and lifesaving care. evaluations of the related covid-19 cfr have suggested that cfr was dependent on the efficacy of local response efforts [31, 32] . a study from china, for example, found that the timely supply of adequate medical resources lowered the cfr from around 4.5-0.5% [33] . we suggest that high cfrs in the abovementioned countries were in part determined by hospital and healthcare surge capacity being exceeded and, as a result, patients potentially receiving suboptimal care during the crisis. recently, nguyen et al. reported that front-line hcws had at least a threefold increased risk of covid-19 infections [34] . given the above-mentioned initial high attack rates in saudi arabia and low cfr, future research that defines the associated risks and outcomes for hcws who contracted covid-19 in saudi arabia is advised. there are several limitations to this study. first, it was done in a single centre and therefore it cannot be generalised to the entire population and this also contributes to selection bias. second, a self-administered electronic questionnaire was used, which increases the chances of recall bias; however, this was balanced with previous data that was collected using the same method from the same cohort. third, although there were statically significant differences in knowledge, attitude and practice scores between different groups, but a small difference should be interpreted with caution. finally, the study was done early in the pandemic and not much information was available and thus it corresponded to a variable level of anxiety and stress. despite these limitations, the study has highlighted the importance of addressing hcw stress levels and ensuring providing information from trustable sources, which will all contribute to better compliance with infection control measures and limiting disease spread. the hcws' worry levels regarding contracting or transmitting mers-cov were higher than for covid-19 during the early stage of the covid-19 pandemic. the current study showed that previous experience with mers-cov and subsequent awareness campaigns that were conducted were associated with increased knowledge, adherence to protective hygienic practices and reduction of anxiety toward the covid-19 pandemic. supplementary material. the supplementary material for this article can be found at https://doi.org/10.1017/s0950268820001958. the novel coronavirus originating in wuhan, china: challenges for global health governance world health organization. situation report -51. who website psychological impact of the covid-19 pandemic on health care workers in singapore a multinational, multicentre study on the psychological outcomes and associated physical symptoms amongst healthcare workers during covid-19 outbreak ihme covid-19 health service utilization forecasting team, murray cjl (2020) forecasting covid-19 impact on hospital bed-days, icu-days, ventilator-days and deaths by us state in the next 4 months immediate psychological responses and associated factors during the initial stage of the 2019 coronavirus disease (covid-19) epidemic among the general population in china do psychiatric patients experience more psychiatric symptoms during covid-19 pandemic and lockdown? a case-control study with service and research implications for immunopsychiatry is returning to work during the covid-19 pandemic stressful? a study on immediate mental health status and psychoneuroimmunity prevention measures of chinese workforce mental health and the covid-19 pandemic covid-19 and its impact on child and adolescent psychiatry -a german and personal perspective mental health services for older adults in china during the covid-19 outbreak. the lancet the psychological impact of the covid-19 epidemic on college students in china how essential is to focus on physician's health and burnout in coronavirus (covid-19) pandemic? cureus 12 dentistry and coronavirus (covid-19) -moral decision-making middle east respiratory syndrome coronavirus epidemic impact on healthcare workers' risk perceptions, work and personal lives isolation of a novel coronavirus from a man with pneumonia in saudi arabia covid-19 in the shadows of mers-cov in the kingdom of saudi arabia severe acute respiratory syndrome coronavirus 2 (sars-cov-2) and coronavirus disease-2019 (covid-19): the epidemic and the challenges clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in wuhan early transmission dynamics in wuhan, china, of novel coronavirus-infected pneumonia the psychological impact of covid-19 pandemic on health care workers in a mers-cov endemic country coronavirus disease 2019 pandemic in the kingdom of saudi arabia: mitigation measures and hospital preparedness knowledge and attitudes towards middle east respiratory syndrome-coronavirus (mers-cov) among health care workers in south-western saudi arabia command and control centre, ministry of health, saudi arabia knowledge and attitude of healthcare workers about middle east respiratory syndrome in multispecialty hospitals of qassim, saudi arabia attitudes, believes, determinants and organisational barriers behind the low seasonal influenza vaccination uptake in healthcare workers -a cross-sectional survey healthcare professionals' knowledge, attitude and acceptance of influenza vaccination in saudi arabia: a multicenter cross-sectional study influenza vaccine uptake, determinants, motivators, and barriers of the vaccine receipt among healthcare workers in a tertiary care hospital in saudi arabia association between 2019-ncov transmission and n95 respirator use world health organization (2020) coronavirus disease (covid-19) pandemic how will country-based mitigation measures influence the course of the covid-19 epidemic? flattening-the-curve associated with reduced covid-19 case fatality rates -an ecological analysis of 65 countries wuhan and hubei covid-19 mortality analysis reveals the critical role of timely supply of medical resources risk of covid-19 among front-line health-care workers and the general community: a prospective cohort study acknowledgement. the authors are thankful to all the healthcare workers who contributed to this research, especially those in the frontline during any infectious disease outbreak. data availability. the datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. key: cord-289907-wzctqkd7 authors: elimian, k. o.; ochu, c. l.; ilori, e.; oladejo, j.; igumbor, e.; steinhardt, l.; wagai, j.; arinze, c.; ukponu, w.; obiekea, c.; aderinola, o.; crawford, e.; olayinka, a.; dan-nwafor, c.; okwor, t.; disu, y.; yinka-ogunleye, a.; kanu, n. e.; olawepo, o. a.; aruna, o.; michael, c. a.; dunkwu, l.; ipadeola, o.; naidoo, d.; umeokonkwo, c. d.; matthias, a.; okunromade, o.; badaru, s.; jinadu, a.; ogunbode, o.; egwuenu, a.; jafiya, a.; dalhat, m.; saleh, f.; ebhodaghe, g. b.; ahumibe, a.; yashe, r. u.; atteh, r.; nwachukwu, w. e.; ezeokafor, c.; olaleye, d.; habib, z.; abdus-salam, i.; pembi, e.; john, d.; okhuarobo, u. j.; assad, h.; gandi, y.; muhammad, b.; nwagwogu, c.; nwadiuto, i.; sulaiman, k.; iwuji, i.; okeji, a.; thliza, s.; fagbemi, s.; usman, r.; mohammed, a. a.; adeola-musa, o.; ishaka, m.; aketemo, u.; kamaldeen, k.; obagha, c. e.; akinyode, a. o.; nguku, p.; mba, n.; ihekweazu, c. title: descriptive epidemiology of coronavirus disease 2019 in nigeria, 27 february–6 june 2020 date: 2020-09-11 journal: epidemiol infect doi: 10.1017/s095026882000206x sha: doc_id: 289907 cord_uid: wzctqkd7 the objective of this study was to describe the epidemiology of covid-19 in nigeria with a view of generating evidence to enhance planning and response strategies. a national surveillance dataset between 27 february and 6 june 2020 was retrospectively analysed, with confirmatory testing for covid-19 done by real-time polymerase chain reaction (rt-pcr). the primary outcomes were cumulative incidence (ci) and case fatality (cf). a total of 40 926 persons (67% of total 60 839) had complete records of rt-pcr test across 35 states and the federal capital territory, 12 289 (30.0%) of whom were confirmed covid-19 cases. of those confirmed cases, 3467 (28.2%) had complete records of clinical outcome (alive or dead), 342 (9.9%) of which died. the overall ci and cf were 5.6 per 100 000 population and 2.8%, respectively. the highest proportion of covid-19 cases and deaths were recorded in persons aged 31–40 years (25.5%) and 61–70 years (26.6%), respectively; and males accounted for a higher proportion of confirmed cases (65.8%) and deaths (79.0%). sixty-six per cent of confirmed covid-19 cases were asymptomatic at diagnosis. in conclusion, this paper has provided an insight into the early epidemiology of covid-19 in nigeria, which could be useful for contextualising public health planning. on 31 december 2019, a cluster of cases of pneumonia of unknown aetiology was detected in wuhan city, hubei province, china [1] . on 7 january 2020, the chinese authorities identified and announced a novel type of coronavirus as the cause of the disease [2] . on 30 january 2020, the world health organization (who) declared the 2019-ncov outbreak a public health emergency of international concern [3] and a few days later announced the official name of the virus as severe acute respiratory syndrome coronavirus 2 (sars-cov-2) and the disease as coronavirus disease 2019 (covid-19) [4] . covid-19 was declared a pandemic on 11 march 2020 by the who. the first case of covid-19 in nigeria was confirmed on 27 february 2020. the case was a 44-year old italian citizen who arrived nigeria through the murtala mohammed international airport, lagos, on a flight via milan, italy [5] . this index case led to the activation of covid-19 public health emergency operation centers (pheoc) at national and sub-national levels, with associated active case finding via contact tracing. by 9 march 2020, 217 contacts were linked to this index case [5] , out of which 136 (63.0%) were under follow-up, with one contact confirmed positive [6] . the 14-day follow-up for contacts of the index case ended on 12 march 2020. during this period, two additional unlinked cases were reported in nigeria. in addition, 42 suspected cases were identified across seven states in nigeria namely the federal capital territory (fct), edo, kano, lagos, ogun, rivers and yobe [5] . since the confirmation of the first covid-19 case in nigeria, cases and deaths have risen steadily in the country, although the government has implemented public health interventionse.g. advocacy for physical distancing, complete and partial lockdown, and ban on large public gatherings including at churches and mosquesto contain or mitigate spread. as of 6 june 2020, 35 (out of 36) states, plus the fct, have reported at least one confirmed covid-19 case. a descriptive analysis of the clinical characteristics, treatment modalities and outcomes of the first 32 covid-19 patients admitted to mainland hospital in lagos state, nigeria, found that two-thirds of patients were male, and the mean age was 38.1 years [7] . this early analysis however is insufficient to provide a national overview of covid-19 epidemiology in nigeria. the nigeria centre for disease control (ncdc) coordinates the public health response to covid-19 in the country. through ncdc's surveillance and laboratory network as well as coordination of state pheocs, epidemiological information on covid-19 cases are captured into a real-time networked platform called surveillance outbreak response management and analysis system (sormas). this forms the basis for the release of daily situation reports for covid-19 on ncdc covid-19 microsite [8] . by 6 june, thousands of individual records with laboratory diagnosis contained on sormas offered opportunities to expand and explore country-specific epidemiologic and clinical characteristics of covid-19 from the onset of the outbreak. this study aims to provide the initial descriptive epidemiology of covid-19 in nigeria, with emphasis on the disease magnitude and patterns in terms of person, place and time. we conducted a retrospective analysis of nigeria surveillance data between 27 february and 6 june 2020. nigeria is administratively divided into 36 states plus the fct, which are zoned across six geopolitical areas: south-south; south-west; south-east; north-east; north-west and north-central. during the study period, 36 states plus fct had reported confirmed covid-19 cases; all states were actively monitoring for cases through the integrated disease surveillance and response system (idsr) system [9] . sormas, an open-source real-time electronic health surveillance database, was the primary data source for this study. in 2017, ncdc adopted sormas as its primary digital surveillance platform for implementing the idsr system [9] , and customised it for the surveillance of priority diseases of public health importance in nigeria. as part of the country's preparedness activities, a covid-19 module was developed and added to sormas in january 2020. all the surveillance data generated through sormas is owned by ncdc, processed and stored in a central server at the ncdc headquarters in abuja, nigeria. the study population was persons investigated for sars-cov-2 infection and captured on sormas during the study period. samples were collected from suspect cases in line with the ncdc case definitions (which were in turn derived from who case definitions) in table 1 [10] . however, these guidelines were not strictly adhered to as samples were also collected from some asymptomatic cases and contacts of cases. trained healthcare personnel (and rapid response team members) investigated suspected covid-19 cases, completed a detailed case investigation form (cif) and collected a minimum of one nasopharyngeal or nasal swab, and one oropharyngeal swab using synthetic fibre swabs with plastic shafts. collected specimens were triplepackaged and aseptically transported in viral transport media, under appropriate temperature conditions (2-4°c) to a designated ncdc-certified laboratory in the country, usually based on proximity. laboratory diagnosis of covid-19 was done by residential setting c residential setting of each person tested for covid-19 was based on the population size and administrative/ legal criteria for the reporting local government areas (lga) as recorded by field staff, in line with common classification of urban and rural classification in nigeria [12] . for example, an lga was classified as urban if 'any one' of the following criteria was met: (1) state capital; (2) an estimated population size of ⩾20 000; (3) >75% of its population is engaged in non-agricultural occupations; (4) availability of infrastructure, good transportation system and a broad array of economic, social and recreational activities. health facility health facility refers to the type of facility each person tested for covid-19 visited prior to diagnosis or was identified for diagnosis. because it was listed on sormas without specific categorisation into health facility type, we utilised the nigeria health facility registry (hfr) of the federal ministry of health [13] to identify each health facility type so as to minimise misclassification errors. the hfr has details of all the registered health facilities in nigeria including the state, lga, facility level (primary, secondary and tertiary) and ownership type (private and public). overall, each health facility was defined either as primary, secondary, or tertiary facility; health facilities that could not be identified in the registry were treated as unknown. education completed classified as a categorical variable in line with the nigerian educational system: no formal education; nursery/ primary; secondary and tertiary/post-secondary. however, given the peculiar nature of the almajiranci/ quranic educational system in nigeria, they were classified under a separate category termed 'alternative' education. classified as a categorical variable as follows: pupil/student; child; housewife; trader/business; health professional (e.g. nurse, clinician, laboratorian etc.); animal-related work (e.g. butcher and hunter); farmer; religious/traditional leaders; transporter and other. travel history classified as local, international and no travel in the last 14 days prior to diagnosis. clinical signs and symptoms defined relative to 14 days before sample collection and classified as binary: yes/no. examples of clinical variables include fever (defined as an axillary temperature of 37.5°c or higher), cough, difficulty breathing, diarrhoea, headache among others. quarantine location defined as a binary variable: formal institution (e.g. health facility) and informal institution (e.g. home). time from symptom onset to diagnosis defined as the time difference between the dates of sample collection and self-reported symptom onset among symptomatic covid-19 cases only. (continued ) real-time polymerase chain reaction (rt-pcr) in accordance with the who interim guidelines [11] . in addition to clinical samples, information on patients' sociodemographic characteristics, signs and symptoms in the 14 days prior to diagnosis, laboratory findings and clinical outcome as detailed in the national cif was captured on sormas. surveillance and laboratory data were submitted by trained data collectors (i.e. healthcare personnel) in real time to the ncdc through the sormas platform (configured on mobile devices (e.g. tablets and smartphones) and laptops) by each reporting state epidemiologist and testing laboratory, respectively. all laboratory-confirmed covid-19 cases were managed according to the ncdc case management protocol [10] , while adherence to infection prevention and control measures for both health workers and patient was ensured. testing for covid-19 during this study period is free of charge in nigeria. de-identified data were retrieved from sormas. covid-19 classifications (suspect, probable and confirmed case) were entered by trained data collectors as per the ncdc case definitions [10] . data management and definitions of key study variables are presented in table 1 . the missing indicator approach was used to address missing data. the primary outcome variables for this study were cumulative incidence (ci) and case fatality (cf). ci was defined as the ratio of covid-19 cases in a defined area to the estimated population of that area. based on a national average growth rate of 3.2%, ci for each reporting state was calculated using the projected nigerian population of 2020 from the 2006 national census and was multiplied by 100 000 for ease of interpretation. cf was defined as the proportion of persons diagnosed with covid-19 who died during the study period, expressed as a percentage (%). both ci and cf were calculated for nigeria and for each state separately. binary/categorical variables were described using frequencies and percentages (%), normally distributed continuous variables by means and standard deviations (s.d.), and non-normally distributed continuous variables by medians and interquartile ranges (iqr). pearson χ 2 test was used to assess how the sociodemographic and clinical characteristics between covid-19 cases (confirmed cases vs. non-cases) and clinical outcome (alive vs. dead). a p-value of <0.05 was considered statistically significant. all statistical analyses were carried out in stata version 13 (stata corp. lp, college station, tx, united states of america). the report of this study was structured in accordance with the strobe statement. the study protocol was approved by the nigeria national health research ethics committee (nhrec/01/01/2007-22/06/2020). between 27 february and 6 june 2020, 60 839 records were entered in the covid-19 sormas database in nigeria, these were classified as follows: 18 790 suspected cases (30.9%), 73 probable cases (0.1%), 12 289 confirmed cases (20.2%), 28 637 non-cases (47.1%) and 1050 non-classified cases (1.7%). this study focuses on individuals with definitive diagnostic classification (40 926): confirmed cases (n = 12 289) and non-cases (n = 28 637). the daily incidence of cases is shown in the epicurve in figure 1 . males (65.8%) constituted a higher proportion of confirmed covid-19 cases than females (31.6%) (fig. 2) . the mean (s.d.) age of confirmed covid-19 cases was 37.1 (15.7) years, with the highest proportion of these cases recorded among persons aged 31-40 years (25.5%) and 21-30 years (21.0%) ( table 2) . despite the high proportion of confirmed cases with missing information on education (53.1%), 30.6% reported completing tertiary time from sample collection to arrival in the laboratory defined as the time difference between the dates of sample arrival in the laboratory and sample collection; it was treated as a continuous variable. defined as the time difference between sample collection and the date diagnostic test was ready (including sample collection, transportation, collection and diagnosis at the laboratory); it was also treated as a continuous variable. a initially, some of the returnees from abroad were tested for covid-19 even in the absence of symptoms. b for confirmed asymptomatic cases, period of contact was measured as the 2 days before, through the 14 days after the date on which the sample was taken which led to confirmation; for symptomatic cases, it was presumably 2 days before symptom onset through 14 days after. c for more information on the criteria for urban/rural classification in nigeria, see [12] . d all negative values following the subtraction of date variables were dropped. education, followed by secondary school certificate holders at 8.6%. for confirmed cases with occupation information available, 9.3% were healthcare workers, while pupil/students and traders accounted for 6.7% each. the proportion of confirmed cases who reported history of travel 14 days prior to diagnosis was generally low, with local and international travels at 4.3% and 1.6%, respectively. sixty-six per cent (8150/12 289) of confirmed covid-19 cases were asymptomatic in the 14 days prior to diagnosis. among confirmed covid-19 cases with symptoms (n = 4139; 33.7%), fever (56.4%) and cough (55.5%) were the most common signs and symptoms reported. other symptoms commonly reported among confirmed covid-19 cases were runny nose (23.8%), sore throat (19.8%), difficulty in breathing (18.6%), headache (14.1%), diarrhoea (7.9%), nausea (7.5%), vomiting (5.5%) table 3 ). of these, cough (72.6%), fever (64.6%) and difficulty in breathing (51.4%) were the most commonly recorded signs and symptoms. other common symptoms recorded at diagnosis were sore throat (16.5%), runny nose (15.1) and vomiting (12.3%). cumulative incidence of covid-19 and case fatality in nigeria, 27 february-6 june 2020 the overall ci of covid-19 infection and cf in nigeria during the study period was 5.6 per 100 000 population and 2.8%, respectively (table 5 ). lagos state (39.9 per 100 000), followed by the fct (19.4 per 100 000), recorded the highest ci in nigeria during this study period. other states with ci higher than the national figure include edo (8.6 per 100 000), kano (6.8 per 100 000), ogun (5.9 per 100 000) and gombe (5.7 per 100 000). regarding cf across the various figure 3 . b percentages in some instances may be greater than 100.0% due to rounding up. c only for symptomatic confirmed covid-19 cases with records of clinical outcome: survivor (n = 1366), dead (n = 212), and total (n = 1578). d 692 total records were used for the assessment of temperature. †p-value <0.05; ‡ p-value <0.001; ns = p-value not statistically significant (i.e. >0.05). φ: p-value from t-test was <0.0001; mean difference was 19.9 years. we have provided a description of the first national epidemiology of covid-19 cases and associated clinical features and outcomes for nigeria. there were 12 289 confirmed covid-19 cases and 28 637 non-cases in 35 states plus the fct in nigeria between 27 february and 6 june 2020. during this period, there were 342 deaths, a ci of 5.6 per 100 000 and a cf of 2.8% overall. after south africa, nigeria is the second most-affected african country in terms of recorded confirmed covid-19 cases and death as of 7 june 2020 [14] . however, the ci of covid-19 in nigeria during the study period, at 5.6 per 100 000 population, is substantially lower than in some non-african countries at a similar stage in their epidemic. for example, about three months after the first confirmed case in the united states, ci was 119⋅6 per 100 000 population, far more than that of nigeria's; with minnesota, the state with the lowest ci, having a ci of 20.6 per 100 000 population [15] . additionally, many european countries reached a ci of at least 4.0 confirmed cases per 100 000 population over a period of less than 1 month [16] . a possible reason for lower ci in nigeria could be due to a relatively low testing capacity in the country as compared to the us and european countries. there was substantial variability in covid-19 incidence among the states in nigeria. the heterogeneity in cis within nigeria could be attributable, in part, to international travels as indicated by the figures recorded by lagos state (39.9 per 100 000) and the fct (19.4 per 100 000) with the two major international airports in the country. another possible explanation might be due to variations in the estimated population of states in nigeria, with smaller population recording a higher ci and vice-versa. for example, ekiti state (3 655 663 population) and enugu state (4 926 955 population) each recorded 29 confirmed covid-19 cases during this study period; but the latter recorded a lower ci (0.6 per 100 000 population) than the former (0.8 per 100 000). moreover, all the nigerian states did not have a similar testing capacity during the study period, and this might have contributed to the observed findings in terms of the numerator figures for calculating cis. similarly, the cf of 2.8% in this study is lower than several other countries which have been hard hit by the covid-19 pandemic. there is a wide range of cfs among non-african countries (from 0.1% in singapore to 16 .2% in belgium [17] ) and in african countries (from 0.0% in uganda to 8.2% in chad) during this study period [14] . nigeria's observed cf of 2.8% is on the lower end of the range reported outside and within africa, but higher than the 2.4% (3210 deaths/133 119 confirmed cases) recorded for the entire africa as of 7 june 2020 [14] . the variation in cf in nigeria could be an indication of varying health system capacity and preparedness across the country. an unpublished study indicates that lagos statewith the highest ci but a cf of 1.3%invested substantially in case management of covid-19 patients as part of its preparedness activities. the overall cf in nigeria could be partly due to its much younger population compared to the united states and most countries in europe [18] ; similar trends in deaths by age from covid-19 have been reported in china [19] . just as cases are potentially underestimated due to inadequate testing, it is likely that deaths from covid-19 are also underestimated, especially in places like kano, which reported significant increases in deaths in april [20] . in contrast to deaths from covid-19, a higher proportion of covid-19 cases was recorded among economically active age groups, suggesting potential role of socio-economic or workrelated activities rather than immunological capacity. children under 5 years of age and those aged 5-13 years, respectively, accounted for 1.7% and 3.9% of confirmed covid-19 cases in this study. these findings are comparable to those from a recent global systematic review [21] . although it remains unclear why children are less affected by covid-19 than older individuals, evidence suggests differences in immune system function [22] . the higher infection rate among males in this study corresponds to evidence reported in the who african region, where males in the 31-39 and 40-49 age groups accounted for 62% of 5178 recorded cases [14] . outside africa, early findings of the clinical characteristics of 41 confirmed covid-19 cases in wuhan, china, reported males to have accounted for 30 (73.0%) of the cases [23] . a study in italy also reported male preponderance [24] . a combination of genetic and physiological factors has been hypothesised as possible explanations for the potential male bias. for example, the wider distribution of sars-cov-2 cellular receptor, angiotensin-converting enzyme 2 (ace-2), in male over females has been postulated [25] . in a patriarchal system such as seen in nigeria, men are more likely to engage in economic activities outside of the household and potentially become more exposed to sars-cov-2 infection than women. while this may be more feasible during a controlled economy, such as that seen during the suspension of non-essential economic activities in the early phase of covid-19 outbreak in nigeria, it may not be applicable when socio-economic activities are functional. this is because women are increasingly partaking in the workforce in nigeria, such that the traditional trends of 'male breadwinner and female family support' are fast eroding [26] . the median length of stay of 111 patients with covid-19 in hospital in this study was 19 days, which is within the range outside of china (4-21 days), but comparatively lower than that from china (4-53 days) [27] . in general, differences in the length of hospital stay may be attributable to variations in criteria for admission and discharge across different countries as well as timing within the pandemic [27] . early diagnosis is fundamental for effective management of covid-19 cases; thus, a median turnaround time of 2 (1-4) days for laboratory diagnosis as noted in the current study seems impressive, and possibly an indication of ongoing measures being championed by the ncdc to strengthen molecular diagnostic capacity in nigeria. however, we lacked information on when laboratory test was received by a covid-19 suspected case, as turnaround time only included the time from sample collection to availability of result. the symptomatic status of confirmed covid-19 cases in this analysis is noteworthy, as over half of them were asymptomatic at testing. a scoping review of the literature found that between 5% and 80% of people testing positive for sars-cov-2 may be asymptomatic [28] , placing the 66% in the current study closer to the maximum range. it is possible that the case investigation approach adopted during testing might have underestimated symptoms: patients were initially asked whether they were symptomatic and probed about individual symptoms only if they answered in the affirmative. stigma associated with covid-19 in nigeria might contribute to people not reporting symptoms when they get tested [29] . furthermore, it is possible for asymptomatic status at diagnosis to change in the course of an illness, in which case such persons could be better classified as presymptomatic cases, so the proportion of truly asymptomatic cases cannot be described by these data. nevertheless, this scenario could pose a challenge to community surveillance activities and implementation of public health interventions (e.g. quarantine and isolation). thus, the possibility of covid-19 transmission by asymptomatic cases in nigeria needs to be explored and addressed, both in terms of research and community risk communication activities. the most common signs and symptoms among symptomatic confirmed covid-19 cases in the 14 days prior to diagnosis were fever (56.4%) and cough (55.5%). this trend is similar to that recorded in a recent systematic review of the literature for china [30] ; however, while fatigue was the third most frequently recorded symptom in china, its frequency was low in our study at 5.2%. similarly, cough, fever and difficulty in breathing, in that order, were the most commonly recorded symptoms at diagnosis among persons who died from covid-19 infection. the common occurrence of difficulty in breathing in deceased patients has been identified as a major driver of adverse clinical outcomes among covid-19 patients [31] . although relatively small in proportion due to late recording during the study period, loss of smell and loss of taste among confirmed covid-19 cases in this study are consistent with available evidence [32] . however, being a descriptive study, these data do not have the capacity to establish a causal association between observed clinical symptoms and covid-19 infection or death. thus, a follow-up study aimed at exploring these associations is recommended. it is also worth noting that the fever which is one of the common symptoms noted in this study is often common in endemic febrile illnesses in nigeria including malaria, lassa fever and yellow fever. as such, in the case of a co-infection, misclassification of illnesses is likely if symptoms alone are used for covid-19 case definitions [33] . the symptomatic and geographic convergence of covid-19 and common febrile diseases in nigeria therefore requires continuous strengthening of definitive diagnostic approaches in the country. about 9% of covid-19 infections occurred in healthcare workers during this study period. covid-19 infection among health workers is of prominent public health importance as it could potentially enhance disease transmission [34] and further weaken a health system that already struggles with insufficient human resources for health. this study has provided the first national epidemiological evidence on covid-19 in nigeria, necessary for public health planning and health system strengthening. however, this study is limited by the substantial proportion of missing data within some of the sociodemographic (e.g. residential setting and health facility) and clinical (e.g. malaise, pharyngeal exudate, rapid breathing, loss of smell and taste) variables studied. the late addition of loss of smell and taste to the cif in nigeria may partly explain why data recorders were not accustomed to capturing them. the high proportion of missing data on some key indicators has prompted a systemic effort to improve the quality of sormas data, and a dedicated data quality improvement project (dqip) was initiated in april to improve completeness of key variables to above 90%. in conclusion, this study has provided an early insight into the epidemiology of covid-19 in nigeria. evidence from this study, such as the high proportion of cases among the active age group and high proportion of asymptomatic cases at diagnosis, will be useful for policymakers and stakeholders in the health and other sectors in contextualising public health planning and response as well as for scientific activities in the country. such measures could include intensifying npis at work and commercial places where this age group is mostly found, and adapting case finding protocols to include routine testing of asymptomatic contacts of confirmed cases. return of the coronavirus: 2019-ncov world health organization (2020) pneumonia of unknown cause-china. who world health organization declares global emergency: a review of the 2019 novel coronavirus (covid-19) preliminary estimation of the basic reproduction number of novel coronavirus (2019-ncov) in china, from 2019 to 2020: a data-driven analysis in the early phase of the outbreak nigeria centre for disease control (2020) covid-19 outbreak in nigeria: situation report-009 can nigeria contain the covid-19 outbreak using lessons from recent epidemics? the lancet global health 8, e770 clinical presentation, case management and outcomes for the first 32 covid-19 patients in nigeria nigeria centre for disease control (2020) an update of covid-19 outbreak in nigeria integrated disease surveillance and response (idsr) strategy: current status, challenges and perspectives for the future in africa nigeria centre for disease control (2020) national interim guidelines for clinical management of covid-19 world health organization (2020) laboratory testing for coronavirus disease (covid-19) in suspected human cases: interim guidance a review of the criteria for defining urban areas in nigeria federal ministry of health (2020) nigeria health facility registry. nigeria health facility registry geographic differences in covid-19 cases, deaths, and incidence -united states rapidly increasing cumulative incidence of coronavirus disease (covid-19) in the european union/european economic area and the united kingdom johns hopkins coronavirus resource center (2020) mortality analyses. maps and trends demographic science aids in understanding the spread and fatality rates of covid-19 clinical features of covid-19 in elderly patients: a comparison with young and middle-aged patients centre for democracy & development (2020) unpacking falsehoods: covid-19 and responses in kano state children are unlikely to be the main drivers of the covid-19 pandemic -a systematic review coronavirus disease (covid-19) in children -what we know so far and what we do not? clinical features of patients infected with 2019 novel coronavirus in wuhan baseline characteristics and outcomes of 1591 patients infected with sars-cov-2 admitted to icus of the lombardy region emerging markers in cardiovascular disease: where does angiotensin-converting enzyme 2 fit in? turbulent but i must endure in silence: female breadwinners and survival in southwestern nigeria covid-19 length of hospital stay: a systematic review and data synthesis covid-19: what proportion are asymptomatic? cebm world health organization (2020) social stigma threatens covid-19 response but patients heal faster with everyone's support. world health organization regional office for africa clinical characteristics of coronavirus disease 2019 (covid-19) in china: a systematic review and meta-analysis clinical characteristics of coronavirus disease 2019 in china the role of self-reported smell and taste disorders in suspected covid-19 covid-19 and malaria: a symptom screening challenge for malaria endemic countries covid-19 and infection in healthcare workers: an emerging problem acknowledgement. we wish to thank all the nigerian frontline health personnel for the contribution to the collection of the data used for this study. the leadership and coordination of all the state commissioners for health, epidemiologists and disease notification and surveillance officers are very much appreciated. last but not least, we are grateful to colleagues who provided technical and administrative support for this manuscript. key: cord-341795-zbqfs77n authors: sikkema, r. s.; farag, e. a. b. a.; islam, mazharul; atta, muzzamil; reusken, c. b. e. m.; al-hajri, mohd m.; koopmans, m. p. g. title: global status of middle east respiratory syndrome coronavirus in dromedary camels: a systematic review date: 2019-02-21 journal: epidemiol infect doi: 10.1017/s095026881800345x sha: doc_id: 341795 cord_uid: zbqfs77n dromedary camels have been shown to be the main reservoir for human middle east respiratory syndrome (mers) infections. this systematic review aims to compile and analyse all published data on mers-coronavirus (cov) in the global camel population to provide an overview of current knowledge on the distribution, spread and risk factors of infections in dromedary camels. we included original research articles containing laboratory evidence of mers-cov infections in dromedary camels in the field from 2013 to april 2018. in general, camels only show minor clinical signs of disease after being infected with mers-cov. serological evidence of mers-cov in camels has been found in 20 countries, with molecular evidence for virus circulation in 13 countries. the seroprevalence of mers-cov antibodies increases with age in camels, while the prevalence of viral shedding as determined by mers-cov rna detection in nasal swabs decreases. in several studies, camels that were sampled at animal markets or quarantine facilities were seropositive more often than camels at farms as well as imported camels vs. locally bred camels. some studies show a relatively higher seroprevalence and viral detection during the cooler winter months. knowledge of the animal reservoir of mers-cov is essential to develop intervention and control measures to prevent human infections. middle east respiratory syndrome (mers) is a highly fatal respiratory tract disease in humans that was first detected in 2012 in the kingdom of saudi arabia (ksa) [1] . after its first detection, mers-coronavirus (mers-cov) was being reported in human patients across the arabian peninsula, with occasional travel-related cases in other continents. as of the end of march 2018, a total of 2189 human laboratory-confirmed cases from 27 countries have been reported to the world health organisation (who), including 782 associated deaths [2] . dromedary camels (camelus dromedaries) have been shown to be the natural reservoir from where spill-over to humans can occur [3, 4] . human-to-human infection is also reported frequently, especially in healthcare settings [5] . sustained human-to-human transmission outside of hospital settings has not been shown yet [6] . direct or indirect human contact with camels has resulted in repeated introductions of mers-cov into the human population [7] . it has been suggested that camels may have acquired mers-cov from a spill-over event from a bat reservoir, but evidence for that remains inconclusive [8] . infections with mers-cov generally are thought to be mild or inapparent in camels [9] , and are therefore of low economical or animal welfare significance. this systematic review was done to compile and analyse all published data on mers-cov in the global camel population to provide an overview of current knowledge on the distribution, spread and risk factors of mers-cov infections in dromedary camels as a basis for the design of intervention and control measures to prevent human infections. on 2 may 2018, a literature search on pubmed was performed, using the terms 'middle east respiratory syndrome coronavirus' and 'mers-cov'. using the term 'mers' did not result in any additional articles that fit the scope of this review. only articles published in english were included. two reviewers individually selected all original research articles containing laboratory evidence of mers-cov infections in dromedary camels in the field. articles that were mentioned in food and agriculture organization (fao) updates [10] or in the references of included publications, but did not appear in the pubmed search were added. subsequently, abstracts, follow-up studies of mers-cov-positive camels and genome studies without prevalence data were excluded from the analysis. data on variables such as year of sampling, country, region, age, sex and animal origin were extracted and analysed. for each variable, the number of positive camels, total number of camels tested and the median percentage positivity was calculated. data from experimental infection studies were not included in this analysis, but they were included in the review to provide additional information and context to the field studies. additional information on the distribution and trade of dromedary camels was collected from references in the publications on mers-cov in camels and extracted from official fao and world organisation for animal health (oie) databases [11, 12] . the additional literature on camel trade was collected in a less systematic way from pubmed. the literature search resulted in a total of 53 papers (fig. 1 ). forty-three research papers described the results of crosssectional studies in dromedary camel populations, six papers described outbreak investigations, including an analysis of camel samples, and four papers described longitudinal studies. in total, 33 papers describe camel studies in the middle east, 13 studies investigated camels from africa and the remaining seven surveys were from spain, australia, japan, bangladesh and pakistan (table 1) . most recent fao statistics estimate the world population of camel to be around 29 million [11] , of which approximately 95% are dromedary camels [13] . however, it is believed that the true population size is even larger due to inaccurate statistics and feral camels, such as the feral dromedary camel population in australia that is estimated to be around 1 million [14] . over 80% of the camel population lives in africa. the main camel countries are chad (6 400 000), ethiopia (1 200 000), kenya (2 986 057), mali (1 028 700), mauritania (1 379 417), niger (1 698 110), sudan (4 830 000), somalia (7 100 000) and pakistan (1 000 000) [12] (table 2) . a large number of camels are being transported from the horn of africa to the middle east each year. these are mainly meat camels coming from the east of africa going to egypt, libya and the gulf states, and sudanese camels that are being imported into the middle east to participate in camel racing competitions [15] . for example, the fao reported that somalia exported 77 000 camels in 2014 [16] . the largest camel market in africa is the birqash market near cairo (egypt), where camels from sudan and ethiopia are most common, but trade routes include animals from chad, somalia, eritrea and kenya [17] . imported camels are usually quarantined for 2-3 days at the border before they are allowed to enter egypt [17] . most somali and sudanese camels that are exported to the ksa are shipped from the ports of berbera and bosaso in north somalia to the ksa ports of jizan and jeddah [15] . in general, only minor clinical signs of disease have been observed in animals infected with mers-cov and most mers-cov infections do not appear to cause any symptoms [9] . disease symptoms that have been described after experimental and field infections are coughing and sneezing, respiratory discharge, fever and loss of appetite [18] [19] [20] . although mers-cov rna can be detected in several organs after experimental infection, in studies of natural infectious virus it has only been detected in the tissues of the upper and lower respiratory tract and regional lymph nodes of the respiratory system in part of the infected camels. histologically, a mild-to-moderate inflammation and necrosis could also be seen on the upper and lower respiratory tract. no viral antigen or lesions were detected in the alveoli. histopathological examination showed that the nasal respiratory epithelium is the principal site of mers-cov replication in camels [18, 21] . in one study investigating experimental infection of camels, mers-cov shedding started 1-2 days post-infection (dpi). in that study, infectious virus could be detected until 7 dpi, and viral rna until 35 dpi in nasal swab samples and, in lower amounts, in oral swab samples [18] . no infectious virus or viral rna was detected in faecal or urine samples [18] . viral rna detection in nasal, but also rectal swabs of camels after experimental infection until day 14, has been confirmed in a recent vaccine study [21] . in the field surveys included in this review, mers-cov rna has been described in rectal swab samples, although other field studies report negative results [3, [22] [23] [24] and when viral rna can be detected, the positivity rate of rectal swabs is lower compared with nasal swab samples [19, [25] [26] [27] . oral swabs are usually negative or show a lower positivity rate even when nasal swabs test positive for mers-cov rna [3, 19, 26] . some studies have reported mers-cov rna in milk samples [27, 28] . longitudinal studies of camel herds show that pcr results of nasal swabs can remain positive after 2 weeks [27, 29] . when an interval of sampling of 1 or 2 months was maintained, nasal swabs become negative for viral rna in the next sampling round [24, 30] . mers-cov infections have also been detected in camels with mers-cov antibodies, both in calves with maternal antibodies as well as older camels that had already acquired antibodies from a previous infection. however, virus replication and thus the virus load is generally lower in infected seropositive animals compared with seronegative camels [19, 21, 23, 24, 30, 31] . little is known about the longevity of antibody titres after infection from longitudinal studies. a study following camels on a closed farm found that neutralizing antibodies remained consistent during a year [30] , while other studies found that antibody titres rapidly drop by 1-4-fold within a period often as short as 2 weeks [24, 27] . the first evidence of mers-cov in camels described so far is the detection of antibodies to mers-cov in camel sera from somalia and sudan from 1983 of which 81% tested positive [32] . additional serological evidence of the widespread presence of mers-cov infection in camels, included in this review, has been found in 18 additional countries: bangladesh, burkina faso, egypt, ethiopia, iraq, israel, jordan, kenya, ksa, mali, morocco, nigeria, oman, pakistan, qatar, spain, tunisia and the uae (fig. 2 ). in addition, promed mail reported that virus-positive camels had been found in kuwait and iran, the latter reportedly in imported animals (archive number 20140612.2534919 and 20141029.2912385). in 11 countries, serological findings were complemented with the finding of viral rna in dromedary camels: burkina faso, egypt, ethiopia, iraq, jordan, ksa, morocco, nigeria, oman, qatar and the uae. investigations of mers-cov circulation amongst dromedary camels in australia, japan, kazakhstan, usa and canada did not find any proof of mers-cov circulation. all countries where mers-cov circulates in the camel population, with the exception of spain (canary islands), pakistan and bangladesh, are located in the middle east or africa [4, 33] . one out of 17 camels that had mers-cov antibodies in bangladesh was born in bangladesh, 16 others were imported from india [34] . however, there have not been any additional reports of mers-cov in camels in india. there is no record of foreign origin of the seropositive camels from pakistan [35] . moreover, in previous studies there had already been evidence of seropositive camels that originate from pakistan [37, 58] . when combining serology data from all papers included in this review, the overall median seroprevalence of camels in africa is 81% (6106/8526; range 28-98%), compared with a median seroprevalence of 93% (3230/3846; range 53-100%) in camels from the middle east. based on viral shedding studies from african countries, the median rate of viral shedding was 5% (1108/6318; range 1-15%), compared with 12% in camels from the middle east (1191/14902; range 0-100%). the seroprevalence of mers-cov antibodies increases with age in camels, while the fraction of camels that test positive for mers-cov rna in their nasal swabs decreases with age [17, 31, 36, 38, 39] . when all serological results of papers that included sufficient age information is combined, the median seroprevalence of camels aged under 2 years is 52% (992/1972; range 0-100%), while the age groups 2-5 years (702/924; range 30-100%) and over 5 years old (1226/1370; range 0-100%) had a combined median seroprevalence of 97%. in the virological studies reporting age breakdown, the median rate of nasal shedding in 0-2 years old camels was 34% (718/2612; range 0-100%) of cases, compared with 2% (91/1142; range 0-100%) in camels older than 2 years. some individual studies show a significantly higher seroprevalence in female camels compared with males [27, 39] , while others show the opposite [38] or do not find any significant difference [17, 35] . similar disagreeing results are published for the presence of mers-cov rna in male vs. female camels [17, 27, 38, 39] . in the studies in this review where sex of camels was recorded, a total of 4810 serum samples from female camels and 3458 samples from male camels were collected and analysed for mers-cov antibodies, compared with 2007 vs. 2505 nasal swabs for viral rna testing. approximately three times more female camels were sampled at farms, while male camels were in the majority in studies that looked at mers-cov prevalence of camels at slaughterhouses, live animal markets and quarantine areas. the overall median seroprevalence of male and female camels in our review is 50% and 67%, respectively (range 0-100%; excluding results from israel and kazakhstan). the median percentage of presence of viral rna is 18% in nasal swabs of male camels (range 0-21%) compared with 9% in female camels (range 0-100%), in our review. in several studies, camels that were sampled at animal markets or quarantine facilities were seropositive more often than camels at farms [17, 22, 27, 34] . combining serological laboratory results of camels in our review with sufficient background information with regard to the sampling location does not result in the same pattern, with a median seroprevalence of 84% (5632/8115; range 0-100%; excluding australia and spain) in camels from farms and 80% (943/1005; range 28-98%) in the camel population sampled at markets and quarantine facilities. studies in egypt found a significantly higher pcr positivity rate in camels sampled in abattoirs or quarantine facilities, but these results could not be confirmed by other papers in this review [17, 27] . when comparing differences in seroprevalence or virus rnapositive rate in nomadic vs. sedentary camel herds, some authors did not find a statistical difference between the two herd management types [39, 40] , while others found some evidence of higher seroprevalences in nomadic herds [27, 36] . one study in kenya looked at the differences between herds with different levels of isolation, and did not find significant differences in mers-cov antibody levels [40] . most studies that compared local camels with imported camels suggested that imported camels are seropositive for mers-cov more often [9, 17, 27, 34, 41] , although not all differences were significant. two studies in egypt found a significantly higher rna positivity rate in imported camels from east africa compared with domestically bred camels [17, 27] , while another study executed in the ksa found a significantly higher number of mers-cov rna-positive results amongst local camels vs. camels from sudan and somalia [22] . although mers-cov was detected almost year-round in camels, some studies show a relatively higher seroprevalence and viral detection during the cooler winter months [17, 20, 27, 38] . mers-cov antibodies have been detected in llamas and alpacas in israel and in alpacas in qatar [42, 43] . to date, no mers-cov antibodies or viral rna have been detected in bactrian camels [4, 37, [44] [45] [46] [47] (table 1 and table 3 ). swine, goats and horses that were included in the field surveys in our review all tested negative for mers-cov rna and antibodies [4, 17, 31, [48] [49] [50] [51] [52] . mers-cov antibodies were detected in two studies in sheep in egypt and qatar, although in very low numbers [17, 51] . however, most surveys that investigated sheep did not find evidence of mers-cov infection or exposure [4, 23, 29, 31, 34, [48] [49] [50] [51] 53] . the publications in this review show that the mers-cov mainly circulates in dromedary camel populations in the middle east and part of africa, and has been infecting dromedary camels in africa for more than three decades. antibodies have also been found in arabic camel sera from the early 90s [31, 32] . however, mers-cov was discovered until 2012, after the first human cases appeared [1] , which is probably due to the minor clinical symptoms of mers-cov infections in camels [18] . most camel surveys were conducted in the middle east and some northern and eastern african countries, but significant data gaps currently still exist in the north and west of africa, in countries that have camel populations of 100 000 to more than a million animals, such as algeria, libya, mauritania and niger. even less is known about the central asian region. some evidence of mers-cov circulation in camels of pakistan and bangladesh was recently published, but data is lacking from afghanistan and india. knowledge on the presence of mers-cov in the animal reservoir is a crucial first step to assess whether mers-cov could be a relevant public health threat in these regions. mers-cov infections are mainly detected in calves and young camels [30, 31] . the research included in this review shows that the igg positivity rate increases gradually in dromedary camels of increasing age while the mers-cov rna detection rate decreases. maternal igg antibodies in camels are acquired through the intake of colostrum during the first 24 h post-parturition. after 24 h, antibody levels in the dam's milk decrease rapidly [54] . one study showed that maternal antibodies in calves peak at 7 days postparturition and decline in the following 6 months. after 5-6 months, over half of the calves did not have maternal neutralizing antibodies in their serum any longer [30] . however, in other field studies, the titre of mers-cov-specific antibodies is still low at 1 month of age and increases with age in dromedary calves [27, 55] . a lower or undetectable antibody levels in young camels is likely to explain the higher mers-cov rna detection rate. in adult camels, a much higher mers-cov seroprevalence can be found, which is probably due to a long-lasting immune response against a mers-cov infection or multiple re-infections with mers-cov. immunity is not sterilizing, as mers-cov infection and shedding have also been shown in adult camels that have mers-cov antibodies [19, 21, 23, 24, 30, 31] . several articles have analysed seroprevalence and virus shedding data in relation to factors, other than age, that may explain differences in seroprevalence and mers-cov rna-positive rate in camels, such as sex, sampling location, herd characteristics and animal origin. our review shows that there is considerable heterogeneity in results. in addition, comparison between studies is difficult given the lack of standardisation of study designs. a key factor to consider when comparing studies is the difference in distribution of male and female camels amongst different disciplines of camel husbandry. females are mainly used for milking and reproduction. as a result, they often stay at farms. male camels, especially of young age (<1 year old), are the predominant sex in slaughterhouses and amongst camels used for transport [39, 56] . this also influences the risk profile of acquiring a mers-cov infection. female camels are in closer contact with calves, who are more susceptible to infection and shed virus in higher quantities compared with older camels [30] . on the other hand, meat and transport camels (predominantly male) travel more, leading to increased contact with other camels and camel herds, and therefore a higher chance of exposure to mers-cov. some papers in this review suggest that there is a generally lower infection rate of domestically bred camels and camels on farms compared with imported camels and camels on animal markets or in quarantine facilities. this may be explained by the same increased contact rate and mixing of camel herds, leading to an increased chance of mers-cov exposure and spread. the increase in mers-cov circulation in winter and spring can have multiple explanations. firstly, the winter is the calving season [10] , which leads to a larger proportion of young animals that usually have a higher number of mers-cov infections and virus excretion. moreover, in winter season, there is a major increase of camel and human movements due to camel racing competitions, camel breeding, trading and movements to grazing grounds, which increases the chance of virus spread. additionally, cooler temperatures may facilitate coronavirus survival in the environment [57] . in experimental studies, llama's and alpaca's are shown to be susceptible to infection with mers-cov [58, 59] , which was confirmed by two papers in our review, describing serologically positive llamas and alpacas in israel and alpacas with mers-cov neutralizing antibodies in qatar [42, 43] . in experimental settings, animal-to-animal transmission has been shown for alpacas, making them a possible risk population for human infections [58] . two studies in our review also found anti-mers-cov antibodies in sheep [17, 51] but experimental inoculation of sheep did not result in mers-cov replication or antibody development [59, 60] . however, the dpp4 receptor, the entry receptor for mers-cov, is present in sheep tissues, making it possible for the virus to bind to the sheep respiratory tract which may explain the finding of mers-cov antibodies [61] . pigs also express the dpp4 receptor in their respiratory tract, and viral replication in experimental settings has been shown for pigs, but no antibodies or mers-cov rna have been found in pigs during field surveys [48, 59] . this may be explained by the limited viral shedding in pigs and the absence of animal-to-animal transmission [62, 63] . we show that dromedary camels are present in large parts of the african and asian continent, and that mers infections in dromedary camels are widespread. however, human infections due to spill-over from the dromedary camel reservoir have not been reported in africa [10] . several explanations for the difference in human cases between the arabian peninsula and africa have been suggested, such as differences in cultural habits, camel husbandry, prevalence of comorbidities, under detection or genetic factors in the local population [64] . moreover, west african viruses were found to be phylogenetically and phenotypically distinct from the mers-cov viruses that caused human disease in the middle east [65] . increased knowledge on the animal reservoir of mers-cov needs to be combined with research on mers prevalence and risk factors in humans to assess the true public health risk. moreover, the absence of human disease, combined with the mild symptoms in camels, caused by mers, will likely have a negative effect on the willingness to implement interventions and the cost-effectiveness of possible interventions in some areas. since the discovery of mers-cov in 2012, the dromedary camel has been identified as the animal reservoir of human infections with the mers-cov. however, the exact route of human primary infections is still unknown. moreover, the scale of the spread and prevalence of mers-cov in the camel reservoir is not fully known yet since there is still a lack of mers-cov prevalence data in some countries that harbour a very significant proportion of the world camel population. however, knowledge of the animal reservoir of mers-cov is essential to develop intervention and control measures to prevent human infections. prospective studies that include representative sampling of camels of different age groups and sex, within the different husbandry practices, are needed to fully understand the patterns of mers-cov circulation. such studies are important as they may give more information on critical control points for interventions to reduce the circulation of mers-cov and/or exposure of humans. author orcids. r. s. sikkema, 0000-0001-7331-6274 financial support. this study was financially supported by the european commission's h2020 programme under contract number 643476 (http:// www.compare-europe.eu/). none. isolation of a novel coronavirus from a man with pneumonia in saudi arabia mers situation update march middle east respiratory syndrome coronavirus in dromedary camels: an outbreak investigation middle east respiratory syndrome coronavirus neutralising serum antibodies in dromedary camels: a comparative serological study middle east respiratory syndrome coronavirus: risk factors and determinants of primary, household, and nosocomial transmission global summary and assessment of risk mers-cov spillover at the camel-human interface further evidence for bats as the evolutionary source of middle east respiratory syndrome coronavirus mers coronaviruses in dromedary camels cov situation updates available at la production de viande de chameau: ã©tat des connaissances, situation actuelle et perspectives distribution and abundance of the feral camel (camelus dromedarius) in australia mers and the dromedary camel trade between africa and the middle east somalia registers record exports of 5 million livestock in 2014 cross-sectional surveillance of middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels and other mammals in egypt replication and shedding of mers-cov in upper respiratory tract of inoculated dromedary camels mers coronavirus in dromedary camel herd, saudi arabia mers-cov in upper respiratory tract and lungs of dromedary camels, saudi arabia an orthopoxvirus-based vaccine reduces virus excretion after mers-cov infection in dromedary camels co-circulation of three camel coronavirus species and recombination of mers-covs in saudi arabia high prevalence of middle east respiratory coronavirus in young dromedary camels in jordan. vector-borne and zoonotic diseases longitudinal study of middle east respiratory syndrome coronavirus infection in dromedary camel herds in saudi arabia novel betacoronavirus in dromedaries of the middle east high proportion of mers-cov shedding dromedaries at slaughterhouse with a potential epidemiological link to human cases systematic, active surveillance for middle east respiratory syndrome coronavirus in camels in egypt middle east respiratory syndrome coronavirus (mers-cov) rna and neutralising antibodies in milk collected according to local customs from dromedary camels epidemiological investigation of middle east respiratory syndrome coronavirus in dromedary camel farms linked with human infection in abu dhabi emirate time course of mers-cov infection and immunity in dromedary camels middle east respiratory syndrome coronavirus infection in dromedary camels in saudi arabia mers coronavirus neutralizing antibodies in camels presence of antibodies but no evidence for circulation of mers-cov in dromedaries on the canary islands middle east respiratory syndrome coronavirus antibodies in dromedary camels serologic evidence for mers-cov infection in dromedary camels antibodies against mers coronavirus in dromedary camels antibodies against mers coronavirus in dromedary camels cross-sectional study of mers-cov-specific rna and antibodies in animals that have had contact with mers patients in saudi arabia risk factors for mers coronavirus infection in dromedary camels in burkina faso, ethiopia, and morocco serological evidence of mers-cov antibodies in dromedary camels (camelus dromedaries) in laikipia county diversity of middle east respiratory syndrome coronaviruses in 109 dromedary camels based on full-genome sequencing mers-cov infection of alpaca in a region where mers-cov is endemic middle east respiratory syndrome coronavirus specific antibodies in naturally exposed israeli llamas, alpacas and camels absence of mers-coronavirus in bactrian camels absence of middle east respiratory syndrome coronavirus in bactrian camels in the west inner mongolia autonomous region of china: surveillance study results from absence of middle east respiratory syndrome coronavirus in camelids middle east respiratory syndrome coronavirus infection not found in camels in japan seroepidemiology for mers coronavirus using microneutralisation and pseudoparticle virus neutralisation assays reveal a high prevalence of antibody in dromedary camels in egypt middle east respiratory syndrome coronavirus antibody reactors among camels in dubai middle east respiratory syndrome (mers) coronavirus seroprevalence in domestic livestock in saudi arabia middle east respiratory syndrome coronavirus (mers-cov) serology in major livestock species in an affected region in jordan serologic assessment of possibility for mers-cov infection in equids dromedary camels in northern mali have high seropositivity to mers-cov. one health studies on the supply of immunoglobulin g to newborn camel calves (camelus dromedarius) acute middle east respiratory syndrome coronavirus infection in livestock dromedaries the camel today: assets and potentials stability of middle east respiratory syndrome coronavirus (mers-cov) under different environmental conditions experimental infection and response to rechallenge of alpacas with middle east respiratory syndrome coronavirus livestock susceptibility to infection with middle east respiratory syndrome coronavirus inoculation of goats, sheep, and horses with mers-cov does not result in productive viral shedding host species restriction of middle east respiratory syndrome coronavirus through its receptor, dipeptidyl peptidase 4 middle east respiratory syndrome coronavirus experimental transmission using a pig model domestic pig unlikely reservoir for mers-cov mers-cov antibodies in humans mers coronaviruses from camels in africa exhibit region-dependent genetic diversity genetic diversity and phylogeographic structure of bactrian camels shown by mitochondrial sequence variations evidence for camel-to-human transmission of mers coronavirus human infection with mers coronavirus after exposure to infected camels, saudi arabia seroepidemiology of middle east respiratory syndrome (mers) coronavirus in saudi arabia (1993) and australia (2014) and characterisation of assay specificity middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels isolation of mers coronavirus from a dromedary camel asymptomatic mers-cov infection in humans possibly linked to infected dromedaries imported from oman to united arab emirates middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels in nigeria prevalence of middle east respiratory syndrome coronavirus (mers-cov) in dromedary camels in abu dhabi emirate phylogenetic analysis of merscov in human and camels in iraq no serologic evidence of middle east respiratory syndrome coronavirus infection among camel farmers exposed to highly seropositive camel herds: a household linked study identification of diverse viruses in upper respiratory samples in dromedary camels from united arab emirates sero-prevalence of middle east respiratory syndrome coronavirus (mers-cov) specific antibodies in dromedary camels in tabuk, saudi arabia the prevalence of middle east respiratory syndrome coronavirus (mers-cov) infection in livestock and temporal relation to locations and seasons key: cord-293525-c7nwygl1 authors: saldanha, i. f.; lawson, b.; goharriz, h.; rodriguez-ramos fernandez, j.; john, s. k.; fooks, a. r.; cunningham, a. a.; johnson, n.; horton, d. l. title: extension of the known distribution of a novel clade c betacoronavirus in a wildlife host date: 2019-04-03 journal: epidemiol infect doi: 10.1017/s0950268819000207 sha: doc_id: 293525 cord_uid: c7nwygl1 disease surveillance in wildlife populations presents a logistical challenge, yet is critical in gaining a deeper understanding of the presence and impact of wildlife pathogens. erinaceus coronavirus (ericov), a clade c betacoronavirus, was first described in western european hedgehogs (erinaceus europaeus) in germany. here, our objective was to determine whether ericov is present, and if it is associated with disease, in great britain (gb). an ericov-specific bryt-green(®) real-time reverse transcription pcr assay was used to test 351 samples of faeces or distal large intestinal tract contents collected from casualty or dead hedgehogs from a wide area across gb. viral rna was detected in 10.8% (38) samples; however, the virus was not detected in any of the 61 samples tested from scotland. the full genome sequence of the british ericov strain was determined using next generation sequencing; it shared 94% identity with a german ericov sequence. multivariate statistical models using hedgehog case history data, faecal specimen descriptions and post-mortem examination findings found no significant associations indicative of disease associated with ericov in hedgehogs. these findings indicate that the western european hedgehog is a reservoir host of ericov in the absence of apparent disease. the betacoronavirus genus includes multiple bat-associated viruses and several human pathogens, including severe acute respiratory syndrome coronavirus (sars-cov) and middle eastern respiratory syndrome coronavirus (mers-cov) [1, 2] . in 2013, a novel clade c betacoronavirus was discovered in the western european hedgehog (erinaceus europaeus) in germany [3] . the virus was subsequently detected in the same species in france [4] . characterisation of these erinaceus coronavirus (ericov) nucleotide sequences revealed high nucleotide identity to mers-cov [3] , the cause of an acute respiratory syndrome in humans with high case fatality rates [5, 6] . the structural spike protein is key to coronavirus virulence and host specificity [7] and the gene that encodes it plays a pivotal role in coronavirus diversity [8] . analysis of the receptorbinding region in the s protein of ericov revealed that it shared only 36.7% amino acid identity with that of mers-cov [3] . however, there have been no further published ericov whole genome sequencing attempts, and many questions remain over the origins and evolution of ericov [2] . whilst there is no known threat to public health from ericov, further investigation of the virus is warranted to gain a fuller understanding of the impact of coronavirus infection on wildlife populations. bats are widely recognised and accepted as the natural reservoir hosts of a variety of coronavirus species, including sars-cov [9, 10] . both bats and camels have been cited as possible reservoirs of mers-cov [6, 11] . it is considered that bats are the likely natural reservoir of mers-cov [2, 5] , and that dromedaries (camelus dromedarius), and potentially other species, act as an intermediate host for human transmission [6, 12, 13] . mers-cov-like virus sequences have only been detected in one other species, e. europaeus [3] . hedgehogs are members of the order eulipotyphla, which is closely related to chiroptera, the order of bats [14] . this close association is what initially led to hedgehogs being investigated as a potential coronavirus reservoir [3] . many animal species seem to have the capacity for coronavirus infection in the absence of apparent disease, including bats [15] , aquatic birds [16] and rabbits when inoculated with mers-cov [17] . however, dromedaries infected with mers-cov develop pyrexia with mild upper respiratory signs, consisting of serous to purulent nasal discharge [18] . lesions in infected dromedaries are comparable to those seen in the common cold in humans, including inflammation and loss of pseudostratified epithelial cells in the upper respiratory tract. coronaviruses that affect the gastrointestinal system, such as porcine transmissible gastroenteritis virus, typically cause similarly mild lesions, including destruction of enterocytes and villus atrophy [19] . although still widespread throughout much of western and northern europe, the western european hedgehog population in great britain (gb) has seen a rapid decline in recent decades, with an estimated population decrease of 95% between the 1950s and 1990s [20, 21] . the reasons for this decline are not fully understood, and it is likely to be multi-factorial (e.g. habitat loss and fragmentation, road traffic death and predation) [21, 22] . the possible role of infectious disease also needs to be investigated, including the consequences of infection with newlydiscovered viral agents. here, we identify the presence of ericov in british hedgehogs, sequence the full genome, explore the spatial distribution of infection and investigate if there is any association of ericov with disease. a total of 351 samples from individual hedgehogs were tested in this study. of these, 225 were voided faecal samples collected from live hedgehogs that had been found injured, sick or abandoned juveniles and were consequently admitted to wildlife centres for treatment and rehabilitation in 2014 and 2015. a total of nine wildlife centres across gb collected and supplied samples, using the methods described by sangster et al. [23] . a fresh faecal sample was collected non-invasively shortly after admission and associated metadata, including clinical signs, signalment, location and history, were collected where available. faecal samples were stored at 4 or −20°c for a maximum of 1 week at rehabilitation centres prior to submission. a subsample of each sample collected by sangster et al. [23] was used in the current study. an additional four voided faecal samples were collected from the wild as part of a greater london-based hedgehog population survey conducted in 2014 to give a total of 229 faecal samples from live animals. the remaining 122 samples were of distal large intestinal tract contents collected from dead hedgehogs during post-mortem examination (pme) from 2011 to 2015 and comprised animals with infectious and non-infectious (e.g. trauma) causes of death. these hedgehogs had either died in wildlife centres within 48 h of admission or had been found dead by members of the public. where possible, pmes were performed on fresh carcases, but frozen carcases that had been archived at −20°c were also examined. systematic examination of external and internal body systems was performed using a standardised protocol, followed by microbiological, parasitological and histological investigations where indicated by macroscopic findings and when permitted by the state of carcase preservation [24] . the contents of the distal large intestinal tract were sampled to most closely reflect freshly voided faeces, as the latter had contained the highest ericov viral titre in a previous study [3] . faecal sample submission forms provided information on: admittance date, sample collection date, the location where the hedgehog was found, sex, live body weight, faecal characteristics (including colour and consistency) and reason for admittance. pme records were available for 121 of 122 hedgehogs from which samples of distal large intestinal contents were tested. these comprised data on the location where the dead hedgehog was found, the date the carcase was found, sex, age class, carcase weight, subjective assessment of body condition (based on muscle mass and body fat deposits), gross pme findings and the results of ancillary diagnostic tests. rna stabilising agent rnalater (ambion, life technologies europe, nl) was added to, and mixed with, a pea-sized amount of each sample (faeces or distal large intestinal tract contents) in a 1.5 ml microcentrifuge tube. each sample was then stored at either −20 or −80°c. faeces were homogenised and pooled (five samples per pool). each 140 µl of total pooled sample was added to 560 µl buffer avl-carrier rna solution and extracted immediately following the spin protocol alongside negative extraction controls. a guanidine and column-based method (qiaamp viral mini kit qiagen, hilden, de) was used for rna extraction, according to the manufacturer's protocol. rna was then eluted in 60 µl of buffer ave and stored at −80°c. a specific oligonucleotide primer set (ericovr and ericovf, see supplementary table s1 ) was designed based on the ericov rna-dependent rna-polymerase (rdrp) gene sequence of the german virus sequence (kc545383) with oligoarchitect™ (sigma-aldrich®, st louis, mo, usa) and used with the bryt-green based gotaq® 1-step real-time reverse transcription pcr (rt-pcr) system kit (promega corporation, madison, wi, usa) to detect ericov rna in an applied biosystems 7500 fast real-time system. a 10 −2 dilution of extracted rna from a positive ericov sample (r618/14, previously confirmed through sequencing) was used as the positive control. nuclease-free water (promega corporation) was used as the negative control. following the amplification stages, melt curve analysis was performed in order to verify the specificity of the amplicons [25] . in addition, a 110 bp ericov ultramer oligonucleotide was used to quantify the limits of detection and assess the sensitivity of the designed assay. this was a sequence within the ericov primer rdrp target region, but with the removal of 10 nucleotides at the centre to differentiate it from wild-type ericov (see supplementary table s1 ). a 14-fold dilution series was performed and a standard curve was generated using graphpad prism 6 (graphpad software, ca, usa). to assess assay specificity, rna of the gammacoronavirus, infectious bronchitis virus (ibv), was also included as a control. to assess the primer dynamic range and potential rna degradation, a dilution series of rna extracted from faeces was tested. a 'spiked' dilution series was set containing 25% (5 µl in 20 µl) faecal rna extraction from an ericov-negative pool and a subset of negative samples (n = 11) was tested using primers for the 18s ribosomal rna subunit. all samples extracted from faeces produced positive results confirming the presence of amplifiable rna (data not shown). real-time rt-pcr assays using the ericov primers were carried out for all pooled samples. single samples from positive pools were then extracted and tested individually. a subset of amplicons from individual samples (n = 10) was also assessed by gel electrophoresis and compared to the positive control. sanger sequencing was conducted on three amplicons with ericovr and ericovf primers using standard protocols. amplicon sequences were 2 i. f. saldanha et al. manually checked then aligned against the german ericov rdrp gene fragment sequence retrieved from genbank (accession number kc545385) to confirm their identity (mega 6.0). whole genome sequencing was performed on rna extracted from one faecal sample collected in 2014 (r618/14) which was identified as ericov-positive by real-time rt-pcr. sequencing was performed using synthesis technology (illumina, san diego, ca, usa). rna was extracted as before and dna was depleted from this sample using the on-column dnase treatment (rneasy® plus mini kit) following the manufacturer's instructions (qiagen); the resultant rna was then eluted in 30 µl molecular grade water [26] . cdna was synthesised from 50 ng rna using a random cdna synthesis system (cdna synthesis kit, merck, de), according to the manufacturers' instructions. the cdna was purified using ampure xp magnetic beads (beckman coulter, high wycombe, uk), 1 ng of which was processed for sequencing using the nextera xt dna sample preparation kit (illumina). a sequencing library was prepared according to the manufacturers' instructions and sequenced using an illumina miseq with 2 × 150 bp paired end reads following standard illumina protocols. for confirmation of the insertions and deletions of the resulting sequence, seven extra sets of primers were designed (available on request). the sequences of these short amplicons were compared with the sequence derived by illumina sequencing. the total reads (12 908 682) were mapped to a reference ericov sequence from germany (genbank accession number kc545383). a modified samtools/vcfutils script was used to generate an intermediate consensus sequence in which any indels relative to the original reference sequence were appropriately called. the intermediate consensus was used as the reference for subsequent iteration of mapping and consensus calling. the total number of assembled viral reads was 1 217 783 (9.43% of the total reads). despite the low proportion of viral sequence detected within the total data set, coverage of the entire genome was obtained (average coverage depth of 4546 reads). gross pme findings were categorised into whether respiratory abnormalities (including, but not limited to, pneumonia, consolidation, haemorrhage, congestion, granuloma and high crenosoma sp. burden) and/or gastrointestinal abnormalities (including, but not limited to, haemorrhage, inflammation, intussusception, rupture, abscess and faeces of abnormal colour or consistency) were detected. gross pme findings were available for both organ systems from 94 of the 121 hedgehog examined post mortem and were included in these analyses. a subset of hedgehogs (n = 9) examined postmortem that were ericov-positive was selected for histological examination. selection criteria focused on carcases likely to provide the optimal state of tissue preservation from those available, prioritising tissues from carcases examined fresh or when frozen, from carcases with the least evidence of autolysis. a range of available tissue samples from each selected carcase was fixed in neutral-buffered 10% formalin and was prepared for histopathological examination and stained with haematoxylin and eosin using standard techniques. particular attention was paid to examination of the respiratory and alimentary tracts, when available, as these are the organ systems frequently targeted by mammalian covs. associations between ericov-positive status, case history signalment and pathological findings were explored. variables were first re-categorised and coded using microsoft excel before being exported to ibm® spss® version 22 for windows (ibm®, chicago, il, usa) for statistical analyses. cross-tabulation and pearson's χ 2 tests were undertaken for the binary dependent variable (ericov presence or absence) against all other categorical independent variables. in addition, where sample numbers were sufficient, binary logistic regression (blr) was performed to include consideration of continuous independent variables, such as age, and to quantify the strength of association for multiple variables simultaneously (year, region). a p-value of <0.05 was considered statistically significant. arcgis pro version 11 software (esri, redlands, ca, usa) was used to map the distribution of hedgehogs for which location data were available. of 351 samples examined, 38 (24 faecal samples and 14 distal large intestinal tract samples) were positive for ericov, giving a positive percentage of 10.8%. the viral copies detected in positive faecal samples ranged from 1.15 × 10 8 to 2.7x10 4 per 4 µl extracted rna. the limit of detection for the bryt® green i based ericov primer rt qpcr assay was calculated as 2.68 × 10 3 copies per 1 µl of extracted rna sample. ibv rna at 10 −2 dilution was not detected by the ericov-specific assay. the complete genome of the hedgehog ericov from sample (r618/14, genbank accession number: mk679660) is 30 173 nucleotides (nt) in length, which is comparable to those sequenced from germany: kc545383 and kc545386 (30 148 and 30 175 nt, respectively). complete genome comparison with open reading frames (orfs) from the german ericov (kc545383) gave identities ranging from 79% (orf3b) to 98% (orf6, encoding the envelope protein). the genes encoding the spike protein of the british and german ericov viruses shared 93% identity (table 1) . the locations for all hedgehogs with available data (n = 344) are displayed in figure 1 , along with their ericov status. hedgehogs sampled for the study covered a wide area across gb, from devon in south west england to northern scotland. hedgehogs positive for ericov were found in southern england, east of england, north east england and wales. we found no evidence of ericov in hedgehogs from scotland, even though 17.4% (n = 61) of sampled hedgehogs were submitted from this country. there was a bias of submissions from the east of england, accounting for 44.4% (n = 123) of the total, with norfolk alone contributing 23.1% (n = 64) of the hedgehog samples examined, which is explained by a single, large participating wildlife centre in this region. although all regions of gb were represented, some appeared to be under-represented; for example, only 1.1% (n = 3) of submissions came from north east england. the majority of hedgehog samples were submitted in 2014 and 2015. the proportion positive in 2015 was higher (28/144, 19%) than in 2014 (10/182, 5%) but there were too few submissions from other years to determine if there was an association with year of sampling. there was a higher number of juvenile hedgehogs included in the study in 2015 (n = 68) than 2014 (n = 24), but the proportion of juveniles positive in 2015 (19.1%) was comparable to the proportion of adults positive (20.3%). as coronaviruses are associated with diarrhoea in other species, we tested for the possibility of an association between ericov status and abnormal faeces. the proportion of samples from hedgehogs with reportedly normal faecal colour or consistency, which tested positive for ericov (14/102), was not significantly different from the proportion of all faecal samples with abnormal consistency and/or abnormal colour (8/115) (p = 0.099). however, a significant association was observed between ericov-positive hedgehogs and those with green faeces (p = 0.004). a significant association was also observed between ericov-positive hedgehogs and those with yellow faeces (p = 0.034). colour variation within a single sample of hedgehog faeces was not uncommon: 42.6% (n = 20) of hedgehogs with green coloured faeces were also reported to have faeces of brown, yellow or black colouration. no other statistically significant association was shown across the other independent variables, including body condition ( table 2 ). a blr model testing all but the non-mutually exclusive variables also showed no statistical significance. the proportion of hedgehogs with respiratory abnormalities reported at pme that tested positive for eri-cov (5/55) was not significantly different from the proportion of those without respiratory abnormalities that tested positive (7/39) (p = 0.205). the proportion of hedgehogs with gastrointestinal tract (git) abnormalities that tested positive (5/55) was not significantly different from those that had no git abnormalities (6/39) (p = 0.35). a statistically significant association was shown between ericov status and geographical region of origin (p = 0.008) ( table 2 ). the highest proportion of ericov-positive hedgehog samples were submitted from the south of england (34/217, 16%); however, blr showed no significant association (p = 0.678) between ericov infection status and wider region when other factors including age and year were included. histopathological examination was conducted on available tissues from nine of the ericov-positive hedgehogs in this study that were examined post-mortem: results are presented with the signalment, macroscopic findings and other ancillary diagnostic tests for interpretation in supplementary table s2 . interpretation was complicated by the state of tissue preservation in the majority of cases, limiting the ability to appraise evidence of either a host inflammatory response or the presence of mild abnormalities (e.g. loss of pseudostratified epithelial cells in the respiratory tract or intestinal villus atrophy, as has been reported with some coronavirus infections in other species [27] ). examination of the respiratory and gastro-intestinal tract identified frequent metazoan parasite infections of variable severity, sometimes in combination with bronchopneumonia. a range of other significant findings was detected, supported by ancillary testing, including trauma and yersiniosis. we detected ericov, a novel clade c betacoronavirus, in wild western european hedgehogs in gb. this virus had previously only been reported from germany and france; in both instances from this same species. our results, therefore, indicate that ericov is widespread in this european wild mammal. full genome sequence of ericov from gb showed 94% identity with a ericov previously detected in germany, and 78% identity with the virus responsible for mers [3] . there was no evidence to indicate that ericov was causing clinical disease in hedgehogs. the lack of association with detection of ericov and abnormal faecal consistency, poor body condition (i.e. thin or emaciated), respiratory and digestive tract abnormalities suggests ericov is unlikely to be a primary pathogen in this species. this concurs with studies reporting natural cov infections in wild animal species, including bats [15] and aquatic birds [16] , which do not manifest as clinical disease. this is also a common feature of a reservoir or maintenance host [28, 29] . coronavirus infections of man and other animals are known to predispose to secondary infections, for example, a study recently isolated feline coronavirus and several other enteropathogens from cats with diarrhoea [30] , thus making any association difficult to detect should it occur with ericov infection in the hedgehog. in the current study, histopathological examination was complicated by concurrent parasitic and bacterial infections in several cases and limited by freeze-thaw artefact and the state of tissue preservation, therefore microscopic lesions caused by viral infection could have been missed or obscured. experimental infection studies and the development of in situ hybridisation to localise ericov in tissues would be worthwhile in the future to further elucidate the clinical significance of ericov infection. despite the lack of association between ericov and faecal abnormalities overall, there was an apparent association between ericov status and green or yellow faeces. green faeces is not an unusual finding in hedgehogs, and can often be associated with crenosoma sp., isospora or salmonella sp. infection [31] . many of the hedgehogs reported to have green faeces also had faeces of another colour, suggesting that faecal colour change is not unusual. the colour change or the reporting thereof may not be consistent and therefore this result should be interpreted with caution. however, the possibility of a causative relationship between ericov infection and abnormal faeces colour cannot be excluded. significantly stronger amplification of cov rna has previously been described in juvenile and lactating female myotis sp. bats [32] , and younger age has been shown to be associated with increased coronavirus shedding across various bat species [33, 34] . although no evidence for such an association for ericov in the hedgehog was found in the current study, juveniles accounted for only 27.6% of the hedgehogs sampled. examination of a larger sample size of juvenile animals would be worthwhile to further explore the possibility of age-structured prevalence. hedgehogs positive for ericov were detected over a wide region of england and wales, but not from scotland even though 17 .4% of all samples tested were submitted from this country (fig. 1) . inferring population prevalence from these data is inappropriate due to the sampling strategy, relying on convenience sampling and the catchment area of rehabilitation centres. there is apparent sampling bias since 47% of all hedgehog samples were from the east of england. demonstrating regional differences in population prevalence and absence of infection would require a much larger, random sample of the population, and information regarding the population structure of hedgehogs in gb. annual variation in the proportion of ericov-positive samples was found. the proportion of hedgehogs sampled in 2015 that tested positive for ericov was higher than that in 2014; a difference that was consistent irrespective of sample type or hedgehog age class (table 2 ). this temporal variability in detection rate is consistent with previous multi-year studies in bats where significant annual variation in coronavirus detection rate was reported [34] . seasonal variation in contact rates, environmental factors and fluctuating levels of cov immunity in the population have all been postulated as causes for such fluctuations [34] . proving viral persistence and establishing critical community size is challenging for wildlife populations [29] . whilst there is now a wealth of studies investigating bat covs, research is lacking for other wildlife covs. there is evidence that the gregarious roosting habits of some species of bat facilitate cov transmission [35] , but the transmission dynamics of ericov in the more-solitary hedgehog [20] are likely to be very different. establishing whether infection is endemic in british hedgehogs will require longitudinal sampling, augmented by age-specific serology [36] . overall, the percentage of ericov-positive hedgehogs detected, at 10.8%, was lower than that reported previously, where 58.9% (148/248) and 50% (37/74) of specimens tested positive for ericov in germany and france, respectively [3, 4] . this difference could be due to assay sensitivity or sample degradation. although the limit of detection of the previously used assays was not given, average copy numbers in the faecal samples were well over the sensitivity threshold of the assay used here. all three studies included hedgehogs in wildlife centres. in the german study, no details were given for the timing of sampling. in the current and french studies, however, faecal samples were taken soon after admission (typically within 48 h) in order to reduce the possibility of nosocomial transmission. dog kennels and shelters are recognised as important environments for maintaining enteric canine covs, with increased length of stay associated with increased probability of cov infection [37] . however, nosocomial transmission seems to be an unlikely cause of the differences in sample percentage of positive samples in this study. another potential factor is that the relatively high percentage of positive samples in the french and german studies could be simply due to a much higher prevalence of ericov in the mainland europe hedgehog population: the recruited wildlife centres (from which the hedgehogs were sampled) were all located in north and north western france, with one located close to the border with germany [4] . more studies are required to determine this. hedgehogs in northern europe hibernate for 2-5 months of the year [20] . ericov was detected in both 2014 and 2015, suggesting that the virus overwinters in the hedgehog population, but the method of viral persistence is not known. in the current study, no association was detected with the month of sampling but to explore this further, longitudinal sampling of individual hedgehogs could be conducted, including from animals before and after hibernation, to determine if there is persistence of infection in individuals. deriving full genome sequence is an important step in characterising this virus to assess relationships with other covs, including pathogens of domestic animals and man. overall identity to the previous coronavirus detected in hedgehogs from mainland europe was high (94%), but lower than the identity between two german ericov isolates (97%) [3] . previous studies have demonstrated that ericov shares a common ancestor with mers-cov, and bat coronaviruses hku5 and hku4, raising the possibility of a lineage that has been positively selected for and which has evolved with the potential for cross-species transmission. host tropism in the coronaviruses is largely determined by the spike protein and its ability to bind to host cell receptors. both the german and the gb ericovs have <40% amino acid similarity with mers-cov at the crucial receptor-binding domain on the spike protein, suggesting that binding to the human receptor (dipeptidyl peptidase 4) is unlikely. virus culture was not attempted in this study, but was previously unsuccessful and this might be due to the cell tropism of the virus [3] . the detection of ericov in british hedgehogs is a notable finding, and the first report outside mainland europe. wildlife disease research in gb has greatly increased in recent years, particularly for popular wildlife species such as the hedgehog, where citizen science offers a cost-effective approach to achieving national disease surveillance [38] . it is important that this continues if we are to gain a deeper understanding of endemic and emerging pathogens of wildlife and their impact on wildlife, livestock and human populations. a p-value of <0.05 was considered statistically significant. discovery of seven novel mammalian and avian coronaviruses in the genus deltacoronavirus supports bat coronaviruses as the gene source of alphacoronavirus and betacoronavirus and avian coronaviruses as the gene source of gammacoronavirus and deltacoronavirus ecology, evolution and classification of bat coronaviruses in the aftermath of sars characterization of a novel betacoronavirus related to middle east respiratory syndrome coronavirus in european hedgehogs identification of alpha and beta coronavirus in wildlife species in france: bats, rodents, rabbits, and hedgehogs middle east respiratory syndrome coronavirus in bats, saudi arabia middle east respiratory syndrome coronavirus (mers-cov) origin and animal reservoir the coronavirus spike protein is a class i virus fusion protein : structural and functional characterization of the fusion core complex recombination, reservoirs, and the modular spike: mechanisms of coronavirus cross-species transmission bats are natural reservoirs of sars-like coronaviruses bat origin of human coronaviruses the emergence of the middle east respiratory syndrome coronavirus middle east respiratory syndrome coronavirus in dromedary camels: an outbreak investigation dromedary camels and the transmission of middle east respiratory syndrome coronavirus (mers-cov) a close relationship of chiroptera with eulipotyphla (core insectivora) suggested by four mitochondrial genes ecoepidemiology and complete genome comparison of different strains of severe acute respiratory syndrome-related rhinolophus bat coronavirus in china reveal bats as a reservoir for acute, self-limiting infection that allows recombination events avian coronavirus in wild aquatic birds asymptomatic middle east respiratory syndrome coronavirus infection in rabbits replication and shedding of mers-cov in upper respiratory tract of inoculated dromedary camels molecular characterization and pathogenesis of transmissible gastroenteritis coronavirus (tgev) and porcine respiratory coronavirus (prcv) field isolates co-circulating in a swine herd the state of britain's hedgehogs 2011. people's trust for endangered species national predictors of hedgehog erinaceus europaeus distribution and decline in britain detection and molecular characterisation of cryptosporidium parvum in british european hedgehogs (erinaceus europaeus) streptococcus pyogenes infection in a freeliving european hedgehog (erinaceus europaeus) product differentiation by analysis of dna melting curves during the polymerase chain reaction next generation sequencing of viral rna genomes pathogenesis of coronavirus-induced infections. review of pathological and immunological aspects identifying reservoirs of infection: a conceptual and practical challenge assembling evidence for identifying reservoirs of infection enteropathogen co-infection in uk cats with diarrhoea the importance of faecal indices in assessing gastrointestinal parasite infestation and bacterial infection in the hedgehog (erinaceus europaeus) amplification of emerging viruses in a bat colony influence of age and body condition on astrovirus infection of bats in singapore: an evolutionary and epidemiological analysis alphacoronaviruses in new world bats: prevalence, persistence, phylogeny, and potential for interaction with humans diversity of coronavirus in bats from eastern thailand emerging viruses deciphering serology to understand the ecology of infectious diseases in wildlife cross sectional and longitudinal surveys of canine enteric coronavirus infection in kennelled dogs: a molecular marker for biosecurity. infection citizen science and wildlife disease surveillance