Hrev_master The magnitude of perinatal mortality rate and associated risk factors among deliveries at Dilla University Referral Hospital, Southern Ethiopia: A case-control study Kefale Lelamo Legu, Alemu Tamiso Debiso, Kaleb Mayisso Rodamo Hawassa University College of Medicine and Health Sciences, Hawassa Abstract The perinatal mortality rate is the sum of stillbirths and early neonatal deaths divided by the number of pregnancies of seven or more months’ duration. In Ethiopia, the death rate was 33 deaths/1000 total births in 2016. We aimed to identify the perinatal mortality rate and associated risk factors among deliveries in Dilla University Referral Hospital; January, 2016 - December, 2018. A hospital based retro- spective case-control study was conducted using subgroup binary logistic regression analysis including 138 cases and 296 con- trol group. The proportion of hospital peri- natal deaths was 30% with 90% of the deaths were occurred as a result of still- births and antepartum hemorrhage. Adjusted odds ratios revealed that history of still birth, very low birth weight, short inter- val and nonuse of partograph found to be independent predictors of both stillbirths and early neonatal deaths besides to preg- nancy induced hypertension and antepartum hemorrhage. The risk of perinatal mortality may be increased by not treating chronic ill- nesses, obstetrics complications and risk factors causing low birth weight as well as short birth intervals and not using parto- graph during labour. Introduction The Perinatal Mortality Rate is defined as the sum of the number of perinatal deaths (stillbirths and early neonatal deaths/ENND) divided by the number of pregnancies of seven or more months’ dura- tion (all live births plus stillbirths). PMR is determined by the distinct features of antepartum, intra partum, and neonatal peri- ods (a death in the first seven days of a child born alive). Stillbirth is intrauterine death occurs either before onset of labour (antepartum death) or during labour (intra partum death) and classified as a fetal deaths ≥28 weeks gestation, weight of ≥1000 grams or a body length of ≥35 cm.1-3 Annually 2.5 million neonatal deaths and 2.6 million stillbirths occur globally4,5 of which 1.3 million are intra partum still- births; and 41% of new-born deaths. 70% of stillbirths could be averted with an integrat- ed cost effective antenatal and essential obstetric care.1,3,5 Perinatal death assumed to be an impor- tant public health problem and key indicator of poor health status in low income settings where the death is maintained in very high rates. Birth asphyxia, birth injury, preterm low birth weight, birth interval of <2 years, young maternal age at birth, low level edu- cation, poor maternal nutrition, absence of Antenatal Care (ANC) and complications during labour were primary causes initiat- ing the cascade of perinatal death.6,8 In Ethiopia, the PMR was 33 deaths/1000 pregnancies of seven or more months’ dura- tion (30 still births and 29 neonatal deaths), despite neonatal deaths shown a reduction of 17% over the past 5 years with signifi- cant disparities per 1,000 live births of 43 PMR in rural and 41 PMR in urban settings.9,10 In the southern region, threefold of the national figure was reported, with Pregnancy Induced Hypertension (PIH) being a single most important risk factor and obstructed labor accounted for 26 % of hospital deaths with 50% of still births and 2-5 folds of ENND.10,11 To the authors’ knowledge, data on the rate of perinatal deaths were very old and were obtained from urban residents rather than from remote population where the rate and the risk factors of perinatal deaths were poorly documented.12 Hence, the study aimed to assess the rate and risk factors associated with perinatal mortality so as to fill the existing gap of data and knowledge using local epidemiological study findings of these kinds and to establish a foundation of knowledge and understanding. Materials and Methods Study site A hospital based case-control study was conducted from April 1st-30th, 2019 at Dilla University Referral Hospital (DURH); a referral hospital for Gedeo zone where the total population is estimated to be 1 million, as extrapolated from the 2007 national cen- sus. The hospital was selected based on availability of both delivery and neonatal intensive care units. Study population All deliveries conducted from January Healthcare in Low-resource Settings 2021; volume 9:9960 Correspondence: Kaleb Mayisso Rodamo, Hawassa University College of Medicine and Health Sciences, P.O Box 1560, Hawassa. Tel.: +251.919532392 Fax: +251.0462208755 E-mail: kalebmayisso@gmail.com Key words: Perinatal death; case-control study; Ethiopia. Acknowledgments: We are very grateful to the college of medicine and health sciences, Hawassa University, for funding the research. Contributions: KLL contributed to conduct interviews, to the conception or design of the work; or the acquisition, analysis, or interpre- tation of data for the work, KMR contributed to drafting the work or revising it critically for important intellectual content; writing the manuscript, final approval of the version to be published and ATD contributed as content expert and to manu-script editing, to ensure the accuracy or integrity of any part of the work are appropriately investi-gated and resolved. All authors read and approved the final manuscript. Conflict of interests: The authors have no con- flicts of interest to declare. Further information: This work was supported by a budget for a student research of college of medicine and health sciences, Hawassa University. Availability of data: All data generated or ana- lyzed during this study are included in this the arti-cle. Ethics approval and consent to participate: Ethical approval and clearance were obtained from Institutional Review Board of Hawassa University, College of Medicine and Health Sciences. Since this analysis used entirely reg- istered data, there was no need of obtaining informed written consent. Therefore, verbal informed consent was obtained from the mothers using telephone call to prove that dis- charged neonates were alive up to 7 completed days. The consent was incorporated with con- sideration of the right of volunteer participa- tion and consent based subject’s best interest to participate in the research, both initially and during the course of the research based on Helsinki’s declaration. Moreover, the consent was approved by the Institutional Review Board of Hawassa University, College of Medicine and Health Sciences. Received for publication: 7 August 2021. Revision received: 23 November 2021. Accepted for publication: 29 November 2021. This work is licensed under a Creative Commons Attribution 4.0 License (by-nc 4.0). ©Copyright: the Author(s), 2021 Licensee PAGEPress, Italy Healthcare in Low-resource Settings 2021; 9:9960 doi:10.4081/hls.2021.9960 [Healthcare in Low-resource Settings 2021; 9:9960] [page 39] 1st 2016 – December 30th 2018 were used as a source population. Fetal deaths at 28 com- plete weeks of gestation and live newborns died in the first week of life were used as cases, while control group were newborn delivered alive who did not die before the age of first seven days of life. All medical charts with incomplete records and referred from others institution because of associat- ed complications were excluded. The minimum required sample size was calculated using single proportion formula in a population as described in open EPI- info 2002, version 3, open source calculator.13 We assumed 54.3% frequency of the coexisting newborns low birth weight, one of the factors strongly associat- ed with perinatal mortality from the study conducted in Addis Ababa public hospitals, with Adjusted Odds Ratio (AOR) of 16.45; 95% CI (9.57–28.26).12 The value of 95% confidence interval (CI), power of 80% (1- B) and 5% margin of error; and a non- response rate of 15% and a control to case ratio of 2:1, an estimated sample size of 438 (138 cases and 300 control group) sufficient to determine associated factors. From 2016-2018 delivery reports, cases were traced systematically using delivery, operation and neonatology log books of the hospital. For each case, 2 other babies delivered in the same day as the cases, one before and one after each case, were select- ed and prepared based on inclusion criteria. For subgroup analysis, the matching of cases in the control group was changed from 1:2 in the total cases to control to 1:2.4 in stillbirth to control and to 1:20 in ENND to control group. Variables Outcome variables: this study selected the perinatal death as an outcome variable which is categorized as “Yes” for who expe- rienced still birth or ENND and “No” for did not experienced still birth or ENND. The independent variables selected for the purpose of analyses were: Socio-demo- graphic variables: Age of the mother: the mother’s age at child birth and categorized as: i) less than 18 years, 18–34 years, and 35–49 years; ii) family size: based on the number of family members and dichotomized as small, medium and large family: iii) maternal education: categorized as: no education, elementary education, and higher education; iv) marital status of the mother, dichotomized as currently married and not married; v) house hold wealth: was dichotomized as low, medium and high wealth scores; vi) place of residence: was dichotomized as rural or urban residences; vii) maternal occupation: was categorized as: house wife, employed, daily laborer and farmer; and viii) the household’s religion was dichotomized as: protestant, orthodox or Muslim. Maternal obstetric conditions: Obstetrics complications associated with pregnancy, labor and child birth were: i) APH, multiple pregnancy, premature rup- ture of membrane, prolonged and obstruct- ed labor, cord accidents and PIH; ii) parity of the mother: categorized as primiparous, multipara and grand multipara; iii) pregnan- cy intention: based on women’s self-reports of their desire to become pregnant right before the conception occurred was catego- rized as: intended, mistimed or unwanted. Previous history of perinatal death: was categorized as: i) still birth, early neonatal death and no history of perinatal death; ii) history of abortion, both spontaneous and medically induced termination of pregnan- cy before the 28th week of gestation and dichotomized as yes (ever experienced)/no (never experienced); iii) fetal presentations: was trichotomized as vertex, breech and transverse; iv) mode of deliver: was tri- chotomized as spontaneous vaginal deliv- ery, caesarean section and instrumental delivery; and v) antenatal follow up and use of partograph were dichotomized as yes/no. Coexisting medical conditions of the mothers: i) history of chronic illnesses: dia- betes mellitus, renal and cardiac diseases; ii) tuberculosis, HIV/AIDS and malaria were considered as common infectious dis- eases; and iii) maternal anemia: we tri- chotomized based on hemoglobin level of first, second and third trimester of pregnan- cy as severe anemia, moderate anemia, mild anemia and no anemia. Fetal and newborn conditions: i) gesta- tional age of the new born: was calculated from the last menstrual period and catego- rized as: preterm, term and post term neonates; ii) duration of the labor: was cat- egorized based on the length for the stage of labor as: 12 hour-19 hour-normal first stage, >20 hour -prolonged first stage- 20 min-2 hours-second stage labor, and third stage- 5 hour-30 hour; iii) weight at birth: was cate- gorized as smaller than average, average and larger than average; iv) newborn Apgar score: was generally done at 1st and 5th min- utes after birth and categorized based on the scores as: - normal, fairly low and critically low; v) birth interval: was dichotomized as preceding interval < 2 years and preceding interval 2 or more years; vi) birth order, was trichotomized as: first/second-born, third- born, and fourth/higher order born; vii) neonatal sepsis: dichotomized based on the time of presentation after birth as early- onset sepsis and late-onset sepsis; viii) birth asphyxia: dichotomized into two grades of severity, pale asphyxia and acute asphyxia; and ix) neonatal anemia: trichotomized based on hemoglobin level as severe ane- mia, moderate anemia, and mild anemia. Data collection Two midwives who were trained for 5 days collected the data so that mothers would be comfortable to discuss reproduc- tive health matters that they may not be comfortable to discuss with men. A phone call was made to prove that the discharged neonates were alive up to 7 completed days after verbal informed consent was obtained. For a neonate discharged alive and died before seven completed days, the next alive neonate was taken as a control group. The mother was taken as a non-respondent if she didn’t respond to a phone call or if she was not available. Social, demographic and eco- nomic data of the households were taken from admission register and medical log books. All data of fetal and neonatal condi- tions such as birth weight, gestational age of the neonate etc. and coexisting medical and obstetrics conditions were obtained from antenatal records, maternal admission regis- ter and log books, delivery and labor sum- mary of the mother, ultrasonography records and medical records of neonates. Data analysis The data were coded, checked and Article Table 1. Distribution of perinatal deaths (cases) and control group by year of study, Dilla Univer-sity Referral Hospital, Ethiopia, 2016-2018. Year 2016 2017 2018 Total Total babies born 398 538 688 1624 Cases 36 43 59 138 Control group 74 98 128 300 PMR 90 86 86 85 Stillbirths 35 38 50 123 ENND 4 5 6 15 SB to ENND ratio 9:1 8:1 8:1 8:1 PMR= perinatal mortality rate, ENND= early neonatal death, SB to ENND ratio= still birth to early neonatal death ratio. [page 40] [Healthcare in Low-resource Settings 2021; 9:9960] entered using Epi-Info version 7 and exported into SPSS version 22.0 computer software programs for analysis. Still-birth, perinatal and ENND death rates were calcu- lated using descriptive statistics and cross tabulation. Bivariate logistic regression analysis was conducted to measure the association between the dependent and independent variables and those variables associated with significance level of p value <0.05 were transferred into multivariate logistic regression model to identify the important determinants by controlling pos- sible confounders, and the strength of asso- ciation was measured using OR with 95% CI of a p value <0.01. A total of 10 dichoto- mous household asset variables were involved to generate wealth index using principal component analysis. According to the index, households were divided in to quintiles ranging from the poorest 50% to the richest 2%. Results In a period of three years, a total of 1624 deliveries of all types were registered with a perinatal deaths of 138 (123 still- births and 15 ENND); and overall PMR of 85 per 1,000 total pregnancies of 7 or more months’ duration (Table 1). Still-births accounted for more than two-third (67%) of the mothers admitted to the hospital due to fetal deaths. Among the hospital admissions due to fetal deaths, thirty mothers were admitted with positive fetal heartbeats and reported as hospital stillbirths later on. The proportion of hospital deaths, the death of fetus recorded as “alive” on admission and later on reported as stillbirths plus ENND new borne babies died in the hospital before seven days of life were a quarter (33%) of a total perinatal deaths. Antepartum Hemorrhage (APH) contributed for about 44% of overall perinatal deaths, 40% of the hospital perinatal deaths and 53% of ENND (Table 2). The majority (91%) of the moth- ers in the cases and 270 (90%) in the control groups were married and (96%) of the mothers in the cases and 57% in the control groups were rural dwellers with the literacy rate of 50% in both the groups. Because of only 10 and 11% of the mothers in cases and in control group respectively were employed, 62% of them in the cases and 56% in the control group lied with in low wealth quintiles (Table 3). The study demonstrated 22% of the mothers in the control group and 20% in the cases were primiparous and least (9%) of the mothers in cases had planned for the current pregnancy. More than three-fourth of the mothers in the control group had ANC follow up and two-third used parto- graph during the labor. Nearest to three- fourth of the mothers in the cases had not ANC follow up and more than three-fourth were not used partograph during the labor. The majority (96%) of the mothers in the control group and in the cases (82%) expe- rienced vertex presentation and more than Article Table 2. The proportion of perinatal deaths before and after arriving at the hospital by type of ob-stetric complications, Dilla University Referral Hospital, Ethiopia, 2016-2018. Obstetric complication Total cases Stillbirths before arrival Hospital deaths (N) Stillbirths ENND Number % Number % Number % Number % PIH 61 44 41 44 12 40 8 53 APH 61 44 39 42 16 53 6 40 Obstructed labor 6 4 4 4 2 7 0 0 PROM 5 4 5 5 0 0 0 0 Mal-presentation 2 1 1 1 0 0 1 7 Cord accident 3 2 3 3 0 0 0 0 Total 138 93 30 15 ENND=early neonatal death, APH=antepartum hemorrhage, PIH= pregnancy induced hyperten-sion, PROM=premature rapture of membrane. Table 3. Socio economic and demographic variables of the mothers, Dilla University Referral Hospital, Ethiopia, 2016- 2018 (n=438). Cases (n=138) Control group (n=300) Number Percent Number Percent Marital status Currently married 125 91 270 90 Currently single 13 9 30 10 Occupation House wife 96 69 240 80 Employed 12 9 34 11 Daily labor 18 13 10 3 Farmer 12 9 12 4 House wife 96 69 240 80 Residences Urban residents 15 11 128 43 Rural residents 133 96 170 57 Maternal education No education 72 52 157 52 Primary education (1-6) 53 38 106 35 Secondary education (7+) 13 10 37 12 Wealth quintiles Low wealth score 85 62 168 56 Medium wealth score 35 25 90 30 High wealth score 18 13 42 14 Family size Small family (2-4) 8 6 15 5 Medium family (5-10) 50 36 185 62 Large family (>10 m) 80 58 100 33 [Healthcare in Low-resource Settings 2021; 9:9960] [page 41] [page 42] [Healthcare in Low-resource Settings 2021; 9:9960] three fourth in the cases delivered by Spontaneous Vaginal Deliver (SVD) and the remaining delivered with caesarean sec- tion and instrumental deliveries. 90% of the mothers in the control group and 84% in the cases had no previous history of abortion and 86% of them in cases and 98% in con- trol group had no previous history of peri- natal deaths (Table 4). The mean age in the case of perinatal deaths was 26.5±6.3 and 27±5.8 years in the control group. Among 132 (96%) of the cases of perinatal deaths who experienced single gestation, 76 (55%) had given birth to their first babies. Close to 88% of the cases of perinatal deaths had smaller than average weight at birth and only two new borne were weighing 2500-3999 grams. More than two-third of the new borne in the control group had average birth weight and 10% weighted more than 4000 grams. More than three-fourth of the cases in perinatal deaths had delivered at an interval of less than two years of preceding birth while the majority (93%) in the control group had delivered at an interval of more than two years of preceding birth. Gestational age at delivery in two-third of the cases was term and 7% were post-term; and three-fourth of the cases of the perinatal death had pro- longed labor. Gestational age at deliver in the majority (92%) of the control group were at term and 3% were post term; and only 3% of the control group had 25-48 hour’s median duration of labor. The major- ity (92%) of the cases in neonatal death had critically low Apgar score during 1st minutes and 3% had seven and above Apgar score during 1st.minutes. Similarly, 94% of the cases in neonatal deaths had critically low Apgar scores during 5th. minutes. Three- fourth of the cases in neonatal deaths had a hemoglobin level of <70 g/L whereas simi- lar proportion of the control had a hemoglo- bin level of 110-90 g/L. Seventy four (54%) of the cases in neonatal deaths had devel- oped early onset neonatal sepsis and 8% had developed neither types of neonatal sepsis. Three-fourth of the cases in neonatal deaths had developed severe asphyxia and the least (8%) had neither types of neonatal asphyxia (Table 5). The majority 88% of the mothers of the cases of perinatal deaths had previous histo- ry of chronic medical illnesses and 12% had no previous history of the illnesses. Majority of the mothers’ were negative of HIV/AIDS infection in the cases and in the control group whereas neither of the moth- ers in both the cases and in the control group was positive of Hepatitis B virus infection. The least (16%) of the mothers in the cases were reactive for Venereal Disease Research Laboratory (VDRL). Nearest to two-third of the mothers in the cases had not ever infected either with malaria or tubercu- losis (Table 6). The association of obstetric conditions and perinatal deaths as cases was tested in binary and multiple logistic regression analysis. Those obstetrics conditions shown significant association in crude analysis with a p-value of <0.05 had been transferred to multivariate logistic regression model for adjusted analysis to rule out possible con- founding factors. Adjusted analysis showed that ENND was highest among mothers who had short birth interval (<2 years between births) (AOR = 0.37, 95% CI: 0.17–0.82), a p value <0.01, than mothers who had birth interval of >2 years. Very low birth weight newborns had 1.9 times increased risk of ENND (AOR=0.19, 95% CI: 0.06-0.52), a p value <0.01, than normal weight newborns. Babies born to women with history of chronic illness showed 0.2 times higher odds of still births (AOR=0.13, 95% CI: 0.05-0.33), a p value <0.01 than babies born to women with no history of chronic illness. Labor not followed up with pantographs had 0.1 times higher odds of still births (AOR=0.10, 95% CI: 0.03– 0.26), a p value <0.00, as compared to labor followed up with pantographs. In subgroup analysis, APH (AOR=0.03, 95% CI: 0.01- 0.02), a p value <0.01 and PIH (AOR=0.02, 95% CI: 0.01-0.03), a p value<0.00 had higher risk of still births more than other obstetrics complications including obstruct- ed labour, cord accident and premature rap- ture of membrane (Table 7). Article Table 4. Obstetrics characterstics of mothers, Dilla University Referral Hospital, Ethiopia, 2016- 2018. Obstetric variables Cases (138) Control group (300) No. % No. % Age <18 years 16 12 23 8 18-34 years 98 71 232 77 35-49 years 24 17 45 15 Pregnancy Intended 12 9 260 87 Mistimed 89 64 30 10 Unwanted 37 27 10 3 ANC follow up Yes 37 27 231 77 No 101 73 69 23 Partograph used Yes 27 20 194 65 No 111 80 106 35 Parity I 100 72 213 71 II-IV 27 20 67 22 V+ 11 8 20 7 Fetal presentation Vertex 113 82 287 96 Transvers 13 9 5 1 Breech 12 9 8 3 Mode of delivery SVD 107 78 258 86 C/S 21 15 36 12 Instrumental 10 7 6 2 History of abortion Yes 22 16 31 10 No 116 84 269 90 History of perinatal deaths Still births 4 3 6 2 ENND 15 11 0 0 No history 119 86 294 98 [Healthcare in Low-resource Settings 2021; 9:9960] [page 43] Article Table 5. Fetal and Newborn characteristics, Dilla University Referral Hospital, Ethiopia, 2016-2018. Variables Cases (138) Control group (300) No. % No. % Number of gestation Single 132 96 293 98 Multiple 6 4 7 2 Birth weight in grams 1000-1499 128 92.7 14 4 1500-2499 8 5.7 56 19 2500-3999 2 1.4 200 67 ≥4000 0 0 30 10 Birth interval <2 years 128 92.7 21 7 >/=2 years 10 7.2 279 93 Birth order First 76 55 126 42 Second & third 32 23 137 46 Fourth/higher 30 22 37 12 Gestational age Preterm 91 66 10 3 Term 38 28 275 92 Post term 9 6 15 5 Duration of labor 12 -19 hours 8 6 250 83 >20 hours 23 17 42 14 25-48 hours 104 75 8 3 5-30 hours 3 2 0 0 Apgar score at 1st minute 3 & below 127 92 0 15 4 to 6 7 5 44 65 7 & above 4 3 256 85 Apgar score at 5th minute 3 & below 130 94 2 1 4 to 6 7 5 26 8 7 & above 1 1 272 91 Neonatal anemia Sever 103 75 10 3 Moderate 22 16 68 23 Mild 13 9 222 74 New born sepsis Early onset 74 54 4 1.3 Late onset 53 38 2 0.6 No sepsis 11 8 294 98 Birth asphyxia Sever asphyxia 104 75 0 0 Acute asphyxia 22 16 5 2 No asphyxia 12 8 295 98 Table 6. Coexisting medical characteristics of the mothers, Dilla University Referral Hospital, 2016-2018. Variables Cases (138) Control group (300) No. % No. % *History of chronic illness Yes 121 88 28 9 No 17 12 272 91 History of hepatitis B-virus Positive 1 1 0 0 Negative 137 99 300 100 VDRL status Reactive 22 16 26 9 Non-reactive 116 84 274 91 HIV/AIDS Positive 11 8 13 4 Negative 127 92 287 96 Malaria infection Yes 56 41 120 40 No 90 65 180 60 Tuberculosis infection Yes 45 23 14 5 No 93 67 286 95 Maternal anemia Severe 63 45 12 4 Moderate 33 24 23 8 Mild 30 22 102 34 No anemia 12 9 163 54 *Chronic illness (diabetes mellitus, renal disease, cardiac disease). [page 44] [Healthcare in Low-resource Settings 2021; 9:9960] Discussion In this analysis of the hospital dataset, we found that perinatal deaths were adversely associated with a number of obstetric outcomes during perinatal period. We observed that the PMR was nearly three-fold of the national PMR estimated for 2016, Ethiopia Demographic and Health Survey (EDHS).10 The possible explana- tions of the differences in the PMR were the variation in the scope of the studies and the level that the results were inferred for. However, the finding was almost compara- ble with the hospital-based studies conduct- ed in different parts of Ethiopia. The main reason for increased rate of perinatal mor- tality in both studies is probably because most of the mothers of the cases came very late and with serious obstetric complica- tions and mechanical causes, mainly obstructed labor with or without uterine rupture. Therefore, perinatal mortality is highly prone for overestimation in hospital based studies.11-14 But, the PMR in this study was found to be high as compared to population-based studies conducted in north and northwest Ethiopia.15,16 We observed that birth interval of <2 years, low birth weight, nonuse of parto- graph and history of chronic illness found to had significant association with both still births and ENND. The findings were com- parable with the findings from EDHS, 2016 where the causes of stillbirths and early neonatal deaths were closely linked.9 However, predictive models analysis of population based prospective cohort study in low and middle income countries dis- agreed with the current findings: gestational age at enrollment, maternal age, birth order, parity hypertension, and severe pre-eclamp- sia, or eclampsia found to be important pre- dictors of intra partum stillbirth in the pre- natal and pre-delivery.17 Our result suggested that birth interval <2 years had 0.037 times greater risk of perinatal deaths in comparison to birth interval of more than two years. This is in line with the finding of the longitudinal study conducted in northwest Ethiopia.16 Too many closely spaced pregnancies is a phenomena related to sibling competitions recognized as the maternal depletion syn- drome,18 and associated with premature rup- ture of membranes and puerperal endometritis which can cause perinatal deaths.19 Low birth weight was found to be 0.19 times at higher risk of neonatal death [AOR 0.19; 95% CI (0.02-0.052)] in com- parison to normal birth weight. This is con- sistent with retrospective study conducted in India,20 and predictive models analysis of population based prospective cohort study in low and middle income countries where birth weight was the most important vari- able for predicting the risk of neonatal mor- tality that provided the strongest evidence that the risk of mortality increased with decreasing in birth weight in both the deliv- ery/day 1 and post-delivery/day 2 scenarios. APH in the predictive models analysis of population based prospective cohort study in low and middle income countries was the most important predictor of neona- tal deaths. The finding is consistent with the current study in which APH had 0.03 times high risk of perinatal deaths [AOR 0.03; 95% CI (0.01-0.023)].17 However, the find- ing of the current study was inconsistent with the findings of the study conducted in Addis Ababa public hospitals.10,11 The study demonstrated follow up of labor with pan- tographs had protective effect of 10% [AOR 0.10; 95% CI (0.01-0.02)] against perinatal death than not followed up with pan- tographs. The finding is consistent with perinatal death audit carried out in semi- urban hospital in Kampala, Uganda where updates on use of pantographs reduced a deaths rate by 5 per 1,000 total births after introduction of the audits compared to the death rate before the audit.21 The cases of perinatal deaths who were suffered from PIH and having a history of chronic illness had 0.02/ 0.13 times higher risk of perinatal deaths [AOR 0.02; 95% CI (0.01-0.033)] and [AOR 0.13; 95% CI (0.05-0.33) than their counter parts respec- tively. The findings were consistent with the study conducted in Odisha, India, which demonstrated ranges of fetal pregnancy out- comes: fetal deaths that were occurred as a result of fetal complications including a low birth weight, preterm birth and intra uterine growth retardation resulted in acidosis and less oxygen supply to fetus. In addition, some pre-eclamptic cases lead to severe preeclampsia associated with various maternal complications including the risk of stroke; kidney and liver dysfunction.22 Conclusions Low birth weight, birth interval <2 years, previous history of chronic illness, APH and PIH were identified as increasing the risk of perinatal deaths. However, fol- lowing labor with partograph was identified as decreasing the risk of perinatal deaths. Birth interval of <2 years resulted from nonuse of modern contraceptive and poor quality of intra partum care reflected by nonuse of partograph for labor follow up Article Table 7. Factors associated with perinatal mortality (still births and early neonatal deaths), Dilla University Referral Hospital, Ethiopia, 2016-2018. Variables Perinatal Deaths (138) 95% CI for 95% CI for P-value Still Births (123, %) ENND (15, %) (COR) (AOR) Birth interval in years >2 5 (4.1) 5 (33.3) 1 1 <2 118 (95.9) 10 (66.6) 0.23(0.13-0.21) 0.037 (0.17-0.82)** 0.01 Chronic illness Yes 121 (98.4) 9 (60) 4.89(0.19-0.29) 0.13 (0.05-0.33)** 0.00 No 2 (1.6) 6(40) 1 1 Birth weight in gram 1000-1499 121 (98.4) 12(80) 0.13(0.07-0.27) 0.19 (0.06-0.52)** 0.00 >2500 -3999 2 (1.6) 3(20) 1 1 Use of parto-graph Yes 2 (1.6) 1(6.7) 1 1 No 121(98.4) 14 (93.3) 0.08 (.04-.17) 0.10 (0.03-0.26)** 0.00 Obstetric com-plications APH 55(45) 6(40) 0.07(0.02-1.23)* 0.03(0.01-0.023)** 0.02 Mal-presentation 2(2) 0(53) 1 1 PIH 53(43) 8(53) 0.09(0.01-1.33)* 0.02(0.01-0.033)** 0.00 Mal-presentation 2(2) 0(20) 1 1 1: Indicates the reference categories*: Indicates significant association (P-value < 0.25. **: Indicate highly significant association (P-value <0.05). [Healthcare in Low-resource Settings 2021; 9:9960] [page 45] are the important determinant factors for perinatal loss. The study suggests that the risk of perinatal deaths could be improved by early investigation of pregnant mothers during ANC follow up. Appropriate moni- toring of labor using partograph, immediate newborn care and interventions to prolong birth interval could be resulted in signifi- cant reductions of perinatal mortality. Lots of ENND caused as a result of low birth weight and could be averted by optimizing immediate newborn care and neonatal resuscitation. Limitations of the study The hospital record based data was uti- lized in addition to the participants driven information to overcome the possibilities of recall biases. Intra partum and antepartum stillbirth differentiation can be associated with identification errors, but training of the data collectors, several quality checks and subgroup analyses were made to minimize this error. There is a possibility of potential confounders like individual health status and a wide range of sociocultural norms that were not captured. Nonetheless, as the study results are based on a sufficient num- ber of participants over 3 years, the poten- tial impact of confounders on the study result should be low. Although generaliz- ability of the results could be questioned due to the limited scopes of the data, the results are likely to be pertinent to many of the hospitals similar to those of the study settings. References 1. Allanson E, Tunçalp Ö, Gardosi J, et al. Classify the causes of perinatal death. World Health Organization 2016;94:79. available at https://www.researchgate.net/publica- tion/292676625. 2. Bradley SEK, Winfrey W, Croft TN. Contraceptive Use and Perinatal Mortality in the DHS: An Assessment of the Quality and Consistency of Calendars and Histories. DHS Program, 2015. Available from: www.dhspro- gram.com/publications/publication- MR17-Methodological-Reports.cfm. 3. WHO, USAID, Panamerican Health Organization, et al. Latin American Center for Perinatology Women and Reproductive Health. Plan of action to accelerate the reduction of maternal mortality and severe maternal morbidi- ty: monitoring and evaluation strategy. Montevideo: CLAP/WR; 2012. (CLAP/WR. Scientific Publication; 1593-02). http://perinatal.bvsalud.org/. 4. UNICEF. Levels and trends in child mortality. UN Inter-agency Group for Child Mortality Estimation. Published 2018. Accessed September 1, 2020. Available from: h t t p s : / / d a t a . u n i c e f . o r g / w p - conten t /up loads /2018/10/Chi ld- Mortality-Report-2018.pdf. 5. Liu L, Oza S, Hogan D, et al. Global, regional and national causes of child mortality in 2000-13, with projections to inform post-2015 priorities: an updat- ed systematic analysis. Lancet 2015;385:430-40. 6. Richardus JH, Graafmans WC, Verloove SP, et al. The perinatal mortal- ity rate as an indicator of quality of care in international comparison. Medical Care 1998;36:54-66. 7. World Health Organization. The Global Strategy for Women’s, Children’s and Adolescents’ Health (2016–2030): Sustanable Development Goal Report, Every Woman Every Child 2015. https://www.everywomaneverychild.or g/global-strategy. 8. Little GA, Keenan WJ, Niermeyer S, et al. Neonatal nursing and helping babies breathe: an effective intervention to decrease global neonatal mortality. Newborn and Infant Nursing Reviews 2011;11:82-8. 9. Central Statistical Agency (CSA). Ethiopia mini demographic and health survey 2014. Addis Ababa: Central Statistical Agency [Ethiopia]. Available f r o m : http://www.healthdata.org/ethiopia 10. Central Statistical Agency (CSA) Ethiopia and ICF. The 2016 Ethiopia demographic and health survey key findings. CSA and ICF, Addis Ababa, Ethiopia, and Rockville, Maryland, USA, 2017. Available from: https://dhsprogram.com/pubs/pdf/SR24 1/SR241.pdf. 11. Bayou G, Berhan Y. Perinatal mortality and associated factors: A case control study. Ethiop J Health Sci 2012;22:153- 62. 12. Getiye Y, Fantahun M. Factors associat- ed with perinatal mortality among pub- lic health deliveries in Addis Ababa, Ethiopia, an unmatched case control study. BMC Pregnancy Childbirth 2017;17:245. 13. Sullivan KM, Soe MM. Sample size for cross-sectional & cohort studies & clin- ical trials. OpenEpi, 2007. Available f r o m : http://www.openepi.com/PDFDocs/SS CohortDoc.pdf. 14. Asefa D, Muleta G, Araya F, et al. Pattern of perinatal mortality among deliveries at Jimma University Teaching Hospital, South-West Ethiopia. J Woman Health Issues 2016:5:1000252. 15. Yirgu R, Molla M, Sibley L, Gebremariam A. Perinatal Mortality Magnitude, Determinants and Causes in West Gojam: Population-Based Nested Case-Control Study. PLoS ONE 2016;11:e0159390. 16. Andargie G, Berhane Y, Worku A, et al. Predictors of perinatal mortality in rural population of Northwest Ethiopia: a prospective longitudinal study. BMC Public Health 2013;13:168. 17. Shukla VV, Eggleston B, Ambalavanan N, et al. Predictive modeling for perina- tal mortality in resource limited set- tings: Original investigation. JAMA Netw Open 2020;3:e2026750. 18. David McKinney, Melissa House , Aimin Chen,etal. The influence of interpregnancy interval on infant mor- tality.Am J Obstet Gynecol 2017;216:316.e1-316.e9. 19. Simetka O, Michalec I, Crkvenjas ZN, et al. Toxic epidermal necrolysis com- plicating antibi-otic treatment of puer- peral endometritis: A case report. Ginekol Pol 2015;86:315-7. 20. Shilpi Srivastava, Sanjaya Sharma, Sushila Kharkwal, etal. A study of caus- es of perinatal mortality in tertiary cen- ter in Bundelkhand region. Int J Reprod Contracept Obstet Gynecol 2015;4:43- 46. 21. Nakibuuka VK, Okong P, Waiswa P, et al. Perinatal death audits in a peri-urban hospital in Kampala, Uganda. Afr Health Sci 2012;12:435-42. 22. Jena M, Mishra S, Jena S, et al. Pregnancy induced hypertension & pre eclampsia: Pathophysiology & recent management trends: A review. J Pharmaceut Res Allied Sci 2016;5:326- 34. Article