Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 1 OJPHI Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 Carla Bezold1, Erin Sizemore2, Heather Halter1, Diana Bartlett2, Kelly Hay1, Hammad Ali2 1. MITRE Corporation, United States 2. Centers for Disease Control and Prevention, United States Abstract Objectives: Health department personnel conduct daily active symptom monitoring for persons potentially exposed to SARS-CoV-2. This can be resource-intensive. Automation and digital tools can improve efficiency. We describe use of a digital tool, Sara Alert, for automated daily symptom monitoring across multiple public health jurisdictions. Methods: Eleven of the 20 U.S. public health jurisdictions using Sara Alert provided average daily activity data during June 29 to August 30, 2021. Data elements included demographics, communication preferences, timeliness of symptom monitoring initiation, responsiveness to daily messages, and reports of symptoms. Results: Participating jurisdictions served a U.S. population of over 22 million persons. Health department personnel used this digital tool to monitor more than 12,000 persons per day on average for COVID-19 symptoms. On average, monitoring began 3.9 days following last exposure and was conducted for an average of 5.7 days. Monitored persons were frequently < 18 years old (45%, 5,474/12,450) and preferred communication via text message (47%). Seventy-four percent of monitored persons responded to at least one daily automated symptom message. Conclusions: In our geographically diverse sample, we found that use of an automated digital tool might improve public health capacity for daily symptom monitoring, allowing staff to focus their time on interventions for persons most at risk or in need of support. Future work should include identifying jurisdictional successes and challenges implementing digital tools; the effectiveness of digital tools in identifying symptomatic individuals, ensuring appropriate isolation, and testing to disrupt transmission; and impact on public health staff efficiency and program costs. Keywords: COVID-19, digital health, contact tracing, public health practice, informatics, symptom assessment Abbreviations: Code of Federal Regulations (CFR), coronavirus disease 2019 (COVID-19), standard deviations (SD) DOI: 10.5210/ojphi.v14i1.12449 Copyright ©2022 the author(s) Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 2 OJPHI This is an Open Access article. Authors own copyright of their articles appearing in the Online Journal of Public Health Informatics. Readers may copy articles without permission of the copyright owner(s), as long as the author and OJPHI are acknowledged in the copy and the copy is used for educational, not-for-profit purposes. Introduction Contact tracing, including active monitoring of identified contacts, is a key public health strategy to contain SARS-CoV-2, the virus that causes coronavirus disease 2019 (COVID-19) [1]. Close contacts of persons with COVID-19 are recommended to quarantine and monitor symptoms for 14 days following exposure; under certain conditions, options exist to end quarantine after 7–10 days [2]. Daily active monitoring of persons in quarantine can help improve compliance with quarantine and promptly identify symptomatic persons who may need clinical evaluation [3]. Effective monitoring must be initiated in a timely manner [4]. Daily monitoring is resource- intensive for state, tribal, local, and territorial public health staff [1, 4, 5]. Because of resource constraints, some health departments use technology, including digital tools, to automate aspects of daily monitoring. Digital tools can expedite initiation of symptom monitoring, streamline communication with monitored persons, and integrate data about their status into other public health information systems [6]. Sara Alert (MITRE Corporation, McLean, VA) is an open-source [7], web-based, automated symptom monitoring tool launched in April 2020 [8]. This digital tool can be used for daily symptom monitoring of persons with COVID-19, close contacts recently exposed to persons with COVID-19 [9], and other groups with potential exposure to SARS-CoV-2, including travelers [10], students in K-12 schools [11], and critical infrastructure employees [12]. Each public health jurisdiction determines how to incorporate Sara Alert into their contact tracing workflow [13]. The objective of this report is to describe the use of Sara Alert across multiple jurisdictions for monitoring for COVID-19 symptoms following exposure in terms of the persons monitored, timeliness of initiation in daily monitoring, responsiveness to daily messages, and reports of symptoms. Methods Automated Monitoring Sara Alert can be used to monitor symptoms following confirmed or suspected exposure to SARS- CoV-2 (“exposure monitoring”) or following identification of a confirmed or probable case of COVID-19 (“isolation monitoring”). Each jurisdiction determines the population to be monitored via Sara Alert, and implements procedures to initiate monitoring, including securing permissions as appropriate. Monitored persons must have access to a telephone, mobile phone, or computer to interact with Sara Alert. Monitored persons are sent daily reminders to report symptoms during their monitoring period, via their preferred method (telephone call, text message, texted weblink, or emailed weblink). Monitored persons can respond directly, through a parent or other proxy, or a public health staff member can enter responses collected through follow-up contact. Public health staff are alerted to persons reporting potential COVID-19 symptoms or not responding to that day’s Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 3 OJPHI reminder [13]. Once monitoring is completed and a record is closed, potentially identifiable data are purged to maintain privacy. Data Source Jurisdictions actively using Sara Alert during summer 2021 were invited to participate in this study (N=20). Only those providing consent to share data are included. Data on persons refusing automated monitoring (opt-out) are not consistently available through Sara Alert. Notable data elements and collection methods for this report are: (1) monitored persons or a designated proxy (e.g., parent or guardian) self-report date of birth (required), and sex, race, and ethnicity (all optional); (2) preferred mode of communication (required); (3) preferred contact time and primary language (optional); and (4) public health staff enter last date of exposure during enrollment when known, otherwise, date of last exposure defaults to date of enrollment into Sara Alert. Additional data elements details are available in the data dictionary [14]. Data Analysis We extracted aggregate data on currently monitored persons daily from June 29 to August 30, 2021. To focus on monitoring of symptoms during a quarantine period following a confirmed or possible exposure (as opposed to continuous monitoring of persons with ongoing exposure), we limited our analysis to jurisdictions where automated monitoring was conducted for a mean of ≤ 14 days over the duration of our data collection period, aligning with recommended duration of quarantine for close contacts following SARS-CoV-2 exposure at the time of analysis. We determined the number of monitored persons, summarized their demographic characteristics, calculated symptom monitoring timeliness initiation, examined responsiveness to daily messages, and summarized use of the digital tool to report potential COVID-19 symptoms. We totaled categorical data (sex, race, ethnicity, age group, and contact preferences) across jurisdictions and calculated percentages among all persons monitored. We weighted continuous data by the proportion of persons contributed by each jurisdiction and calculated weighted means and standard deviations (SD). This non-research activity was reviewed by CDC and conducted consistent with applicable federal law (45 CFR part 46 did not apply) and CDC policy. Results Of the 20 jurisdictions invited to participate: five were excluded because automated monitoring was conducted for a mean >14 days; three did not respond by consent deadline; and one declined to participate. Eleven jurisdictions (5 state, 1 territorial, 5 local) contributed to this analysis: Arkansas; Commonwealth of the Northern Mariana Islands; Weld County, Colorado; Berrien County, Michigan; Clay County, Missouri; Jackson County, Missouri; Montana; Vermont; Virginia; Washington; and Teton County, Wyoming. The population within the 11 participating jurisdictions is 22,369,182 [15]. In these jurisdictions, each day on average, public health staff Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 4 OJPHI initiated automated monitoring of 1,318 persons, monitored 12,450 persons automatically, and sent 70,218 automated messages. Among 11,165 persons with sex reported, the majority were female (5,823, 52%). Nearly half (5,474, 45%) of persons monitored were aged ≤ 18 years (Table 1). Race and ethnicity were not available for 64% and 30% of monitored persons, respectively. Monitoring preferences were to be contacted via text message (47%), in the morning from 8:00-12:00 AM (51%). Among 10,616 persons with a recorded self-reported primary language, their preference was predominately English (10,350, 97%%). Table 1. Demographic characteristics and communication preferences among persons monitored for COVID-19 symptoms using an automated symptom monitoring tool (Sara Alert), 11 U.S. public health jurisdictions, June 29-August 30, 2021 (N=12,450) Characteristic Monitored persons, No. (Percent)a Sex Male 5,342 (43%) Female 5,823 (47%) Unknown 33 (<1%) Not reported 1,252 (10%) Race White 3,316 (27%) Black or African American 917 (7%) American Indian or Alaska Native 15 (<1%) Asian 146 (1%) Native Hawaiian or Pacific Islander 10 (<1%) Multiple races 111 (1%) Not reported 7,935 (64%) Ethnicity Hispanic or Latino 698 (6%) Non-Hispanic or [not] Latino 7,879 (63%) Not reported 3,735 (30%) Ageb group, years Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 5 OJPHI ≤ 18 5,474 (45%) 19-29 1,540 (13%) 30-39 1,604 (13%) 40-49 1,426 (12%) 50-59 1,044 (9%) 60-69 652 (5%) 70-79 299 (2%) ≥ 80 100 (1%) Probable miscalculationb 312 (3%) Preferred mode of communication from public health jurisdiction Telephone call 833 (7%) Text message 5,900 (47%) Texted weblink 4,275 (34%) Emailed weblink 1,442 (12%) Preferred contact time Morning (8:00 to 12:00) 6,317 (51%) Afternoon (12:00 to 16:00) 3,226 (26%) Evening (16:00 to 20:00) 1,111 (9%) Not reported 1,796 (14%) Primary language English 10,350 (83%) Spanish 247 (2%) Otherc 19 (<1%) Not reported 1,832 (15%) aNumbers and percentages may not sum to 100 due to rounding, since values are daily averages across the study period. bSara Alert calculates age as the difference between date of birth and the current date. Outlier records with calculated age > 110 years are flagged as probable miscalculations. cAs of August 2021, Sara Alert also supports French, Somali, and Puerto Rican Spanish messaging from the health department. Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 6 OJPHI Public health staff-initiated symptom monitoring in Sara Alert a weighted mean of 3.9 days (SD = 2.8; range 0-7.8) after last exposure, and persons were monitored for a weighted mean of 5.7 days (SD = 5.8; range 1.7-8.3). Within an individual’s monitoring period, 49% (6,159/12,450) responded to every automated message, and 74% (9,242/12,450) responded to at least one automated message. Of those responding, 86% of responses were submitted directly via the digital tool (5,281/6,159 for every message; 7,944/9,242 for at least one message). Of the 51% (6,291/12,450) who did not respond to every automated message sent during their monitoring period, a weighted mean of 3.6 (SD = 0.7; range 1.4-6.7) consecutive days passed without response. Eleven percent (1,313/12,450) of monitored persons reported potential COVID-19 symptoms [16] during their monitoring period. Discussion Public health staff in 11 jurisdictions used Sara Alert to automate symptom monitoring for an average of more than 12,000 persons per day from June 29 to August 30, 2021. The high number of automated messages sent, and large number of individuals engaging directly with the tool suggests the potential benefit of digital tools to increase public health staff capacity and efficiency for symptom monitoring beyond what is achievable through traditional, non-automated methods. Our findings build upon previously published experience with Sara Alert from Maine [9]. Our study incorporates a broader geographic area and monitoring experiences 18 months after the World Health Organization’s declaration of the COVID-19 pandemic. The jurisdictions included in this analysis have a combined population of more than 22 million persons and include areas of higher overall COVID-19 incidence than Maine. Both the report from Maine and our study explore use of the tool during a time when COVID-19 case counts were increasing towards a single day peak, namely the July 2020 peak for Maine and September 2021 peak for our study [17]. Collectively, our findings suggest that use of a digital tool for symptom monitoring has utility, both for public health staff and monitored persons, throughout the duration of the COVID-19 pandemic and across a broad geographic area. Uniquely, our study contributes data on engagement by monitored persons with a digital tool for symptom reporting. Only 49% of monitored persons responded to every automated message. We think this finding reflects the difficulty in getting complete monitoring data, despite using a digital tool. However, 74% of monitored persons responded to at least one automated message. This number is higher than what has been reported in other studies examining the use of automated monitoring tools for other infectious diseases [18, 19]. The majority persons monitored that responded to automated messages did so directly via Sara Alert; suggesting that persons monitored accepted use of a digital tool to report symptoms. Interestingly, we found that 45% of monitored persons were 18 years of age or younger. Considering trends in case counts during July and August, 2021, where the age groups with the highest incidence of cases were among children or persons of reproductive age [17], combined with lower vaccination coverage in persons <18 [17, 20], it is plausible that a large portion of close contacts requiring symptom monitoring by public health departments would be 18 years of age or Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 7 OJPHI younger. Additionally, K-12 schools resumed in-person classes in some jurisdictions during our study period, potentially reflecting increased exposures for children attending school [21]. Limitations A key limitation of this report is the fact that we are unable to document the reason each monitored person was enrolled in Sara Alert, and this report likely includes persons not identified as close contacts following a SARS-CoV-2 exposure. The populations eligible for automated symptom monitoring via Sara Alert shifted over time due to jurisdictional changes in policy, community transmission levels, and resource availability [9]. Based on their evolving COVID-19 response needs, some jurisdictions primarily used Sara Alert for symptom monitoring of close contacts, while others used it to monitor travelers entering their jurisdiction or persons with ongoing exposures such as critical infrastructure employees. Information to differentiate between these uses is not captured consistently across jurisdictions. However, persons were enrolled in Sara Alert based on health department policies and priorities, and as such our findings can provide relevant insights regarding the acceptability of automated symptom monitoring for a range of purposes. This report is subject to at least three additional limitations. First, as our analysis period represents a snapshot of the COVID-19 pandemic after initiation of vaccination of persons 12 years and older and during a time of increasing but non-peak case counts, findings may not be generalizable beyond similar epidemiologic settings. Second, our findings are neither representative of all jurisdictions using Sara Alert nor of the United States. Generalizability to the U.S. population is also limited by available data on race and ethnicity of monitored persons and overrepresentation of persons aged 18 years and younger compared with the U.S. population (45% v. 22% in 2019 U.S. Census). Third, because deidentified aggregate data extracted from Sara Alert were used in this analysis, we were not able to link to other data sources (e.g., laboratory results, case surveillance) or verify values (e.g., last date of exposure, which might be biased toward zero because of system defaults). Conclusion Numerous digital tools emerged to support contact tracing during the COVID-19 response [6], but limited data describing their use are available. This report shows that incorporating this digital tool into public health workflows might improve public health capacity for daily symptom monitoring, allowing staff to focus their time on interventions for persons most at risk or in need of support. Evaluation priorities for future work include implementation lessons learned, challenges and successes; the effectiveness of digital tools compared to traditional non-automated monitoring methods in identifying symptomatic individuals, ensuring appropriate isolation, and testing to disrupt transmission chains; and how these digital tools helped health departments use human resources more efficiently, resulting in public health cost savings. Correspondence: Erin Sizemore, esizemore@cdc.gov Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 8 OJPHI Acknowledgements We would like to thank all monitored persons who enrolled in Sara Alert; the public health staff at each jurisdictions contributing data to this report; Margaret (Peggy) Honein, Melanie Taylor, Nicholas DeLuca, Jody McLean, Lauren Finklea and Calvin Hightower for their support and contributions to Sara Alert through CDC’s COVID-19 emergency response; Dawn Heisey-Grove for critical review of this manuscript; Adam Holmes, Maeve Kokolus, and Kathy Lewis for assistance acquiring data; Paul Jarris and Suzette Stoutenburg for contributions to Sara Alert as MITRE project leadership; and Association of Public Health Laboratories, Association of State and Territorial Health Officials, CDC Foundation, Council of State and Territorial Epidemiologists, and National Association of County and City Health Officials for their partnership in Sara Alert. Financial Disclosure The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work, including no-cost, jurisdictional implementation of Sara AlertTM and technical assistance, was supported by the Centers for Disease Control and Prevention [Contract 75FCMC18D0047]. The Sara Alert trademark is currently registered with the United States Patent and Trade Office. This technical product was produced for the U.S. Government under Contract Number 75FCMC18D0047, and is subject to Federal Acquisition Regulation Clause 52.227-14, Rights in Data-General. No other use other than that granted to the U.S. Government, or to those acting on behalf of the U.S. Government under that Clause is authorized without the express written permission of The MITRE Corporation. For further information, please contact The MITRE Corporation, Contracts Management Office, 7515 Colshire Drive, McLean, VA 22102-7539, (703) 983-6000. (c) 2021 The MITRE Corporation All Rights Reserved. Approved for Public Release 21-3316. Competing Interests No Competing Interests. References 1. Spencer KD, Chung CL, Stargel A, Shultz A, Thorpe PG, et al. 2021. COVID-19 Case Investigation and Contact Tracing Efforts from Health Departments - United States, June 25- July 24, 2020. MMWR Morb Mortal Wkly Rep. 70(3), 83-87. PubMed https://doi.org/10.15585/mmwr.mm7003a3 https://pubmed.ncbi.nlm.nih.gov/33476317 https://doi.org/10.15585/mmwr.mm7003a3 Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 9 OJPHI 2. Centers for Disease Control and Prevention. When to Quarantine. Available at: https://www.cdc.gov/coronavirus/2019-ncov/if-you-are-sick/quarantine.html. 3. Centers for Disease Control and Prevention. Science Brief: Options to Reduce Quarantine for Contacts of Persons with SARS-CoV-2 Infection Using Symptom Monitoring and Diagnostic Testing. Available from: https://www.cdc.gov/coronavirus/2019-ncov/science/science- briefs/scientific-brief-options-to-reduce-quarantine.html. 4. Kretzschmar ME, Rozhnova G, Bootsma MCJ, van Boven M, van de Wijgert J, et al. 2020. Impact of delays on effectiveness of contact tracing strategies for COVID-19: a modelling study. Lancet Public Health. 5(8), e452-59. Epub Jul 2020. doi:https://doi.org/10.1016/S2468-2667(20)30157-2. PubMed 5. Ruebush E, Fraser M, Poulin A, Allen M, Lane JT, Blumenstock JS. COVID-19 Case Investigation and Contact Tracing: Early Lessons Learned and Future Opportunities. J Public Health Manag Pract. 2021 Jan/Feb;27 Suppl 1, COVID-19 and Public Health: Looking Back, Moving Forward:S87-S97. 6. Centers for Disease Control and Prevention. Digital Contact Tracing Tools. Available at: https://www.cdc.gov/coronavirus/2019-ncov/php/contact-tracing/contact-tracing-plan/digital- contact-tracing-tools.html. 7. MITRE Corporation. Sara Alert Open Source Code. Available at: https://github.com/SaraAlert/SaraAlert. 8. MITRE Corporation. Sara Alert. Available at: https://saraalert.org/. 9. Krueger A, Gunn JKL, Watson J, Smith AE, Lincoln R, et al. 2020. Characteristics and Outcomes of Contacts of COVID-19 Patients Monitored Using an Automated Symptom Monitoring Tool — Maine, May–June 2020. MMWR Morb Mortal Wkly Rep. 69(31), 1026- 30. PubMed. https://doi.org/10.15585/mmwr.mm6931e2 10. Pacific Island Times News. CNMI announces new arrival procedure for travelers. 2020. Available at: https://www.pacificislandtimes.com/post/2020/07/07/cnmi-announces-new- arrival-procedure-for-travelers. 11. News Center Maine. Contact tracers at the Maine Department of Education getting the word out about “Sara Alert.” 2020. Available at: https://www.youtube.com/watch?v=xTZx4A_o- PA. 12. Newsroom WLAD. Danbury City employees using Sara Alert app to monitor COVID-19 symptoms. 2020. Available at: https://wlad.com/local-headlines/551471. 13. MITRE Corporation. About Sara Alert. Available at: https://saraalert.org/. https://www.cdc.gov/coronavirus/2019-ncov/if-you-are-sick/quarantine.html https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/scientific-brief-options-to-reduce-quarantine.html https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/scientific-brief-options-to-reduce-quarantine.html https://doi.org/10.1016/S2468-2667(20)30157-2 https://pubmed.ncbi.nlm.nih.gov/32682487 https://www.cdc.gov/coronavirus/2019-ncov/php/contact-tracing/contact-tracing-plan/digital-contact-tracing-tools.html https://www.cdc.gov/coronavirus/2019-ncov/php/contact-tracing/contact-tracing-plan/digital-contact-tracing-tools.html https://github.com/SaraAlert/SaraAlert https://saraalert.org/ https://pubmed.ncbi.nlm.nih.gov/32759918 https://doi.org/10.15585/mmwr.mm6931e2 https://www.pacificislandtimes.com/post/2020/07/07/cnmi-announces-new-arrival-procedure-for-travelers https://www.pacificislandtimes.com/post/2020/07/07/cnmi-announces-new-arrival-procedure-for-travelers https://www.youtube.com/watch?v=xTZx4A_o-PA https://www.youtube.com/watch?v=xTZx4A_o-PA https://wlad.com/local-headlines/551471 https://saraalert.org/ Sara Alert: An automated symptom monitoring tool for COVID-19 in 11 jurisdictions in the United States, June – August, 2021 10 OJPHI 14. MITRE Corporation. Sara Alert Data Dictionary. Available at: https://saraalert.org/blog/user_guide/data-dictionary/. 15. US Census Bureau. American Community Survey, 2019 American Community Survey 5- Year Estimates. Available at: https://data.census.gov/cedsci/. 16. Council of State and Territorial Epidemiologists. Update to the standardized surveillance case definition and national notification for 2019 novel coronavirus disease (COVID-19) Available at: https://cdn.ymaws.com/www.cste.org/resource/resmgr/ps/positionstatement2020/Interim-20- ID-02_COVID-19.pdf. 17. Centers for Disease Control and Prevention. COVID Data Tracker [Internet]. 2021. Available at: https://covid.cdc.gov/covid-data-tracker/#datatracker-home. 18. Stewart RJ, Rossow JA, Eckel SA, Bidol S, Ballew GA, et al. 2019. Text-Based Illness Monitoring for Detection of Novel Influenza A Virus Infections During an Influenza A (H3N2)v Virus Outbreak in Michigan, 2016: Surveillance and Survey. JMIR Public Health Surveill. 5(2), e10842. PubMed https://doi.org/10.2196/10842 19. Stephenson LM, Biggs JS, Sheppeard V, Oakman TL. 2016. An evaluation of the use of short message service during an avian influenza outbreak on a poultry farm in Young. Commun Dis Intell Q Rep. 40(2), E195-201. PubMed 20. Murthy BP, Zell E, Saelee R, Murthy N, Meng L, et al. 2021. COVID-19 Vaccination Coverage Among Adolescents Aged 12-17 Years — United States, December 14, 2020-July 31,2021. MMWR Morb Mortal Wkly Rep. 70, 1206-13. PubMed https://doi.org/10.15585/mmwr.mm7035e1 21. Siegel DA, Reses HE, Cool AJ, Shapiro CN, Hsu J, et al. 2021. Trends in COVID-19 Cases, Emergency Department Visits, and Hospital Admissions Among Children and Adolescents Aged 0-17 Years — United States, August 2020-August 2021. MMWR Morb Mortal Wkly Rep. 70(36), 1249-54.PubMed. https://doi.org/10.15585/mmwr.mm7036e1 https://saraalert.org/blog/user_guide/data-dictionary/ https://data.census.gov/cedsci/ https://cdn.ymaws.com/www.cste.org/resource/resmgr/ps/positionstatement2020/Interim-20-ID-02_COVID-19.pdf https://cdn.ymaws.com/www.cste.org/resource/resmgr/ps/positionstatement2020/Interim-20-ID-02_COVID-19.pdf https://covid.cdc.gov/covid-data-tracker/#datatracker-home %20PubMed https://doi.org/10.2196/10842 https://pubmed.ncbi.nlm.nih.gov/27522128 https://pubmed.ncbi.nlm.nih.gov/34473680 https://doi.org/10.15585/mmwr.mm7035e1 PubMed https://doi.org/10.15585/mmwr.mm7036e1