Background: Early neonatal death is a serious concern, both in the developing and developed worlds. Early neonatal death remains a health problem and is the biggest component of the neonatal mortality rate. Therefore, substantial reduction in early neonatal mortality is a crucial pre-requisite for achieving further gains in child survival in the country. However, the distribution of deaths in the community within the first week is poorly understood. Therefore, this study was conducted to assess time to death and its predictors among early neonates in the neonatal intensive care unit of Dessie Referral Hospital.

Method: A retrospective cohort study was conducted in the early neonatal patient in Dessie referral hospital on the total sample size of 416 selected patients by using a simple random sampling technique. Data were collected by document review. Life table used to estimate the probability of survival, log-rank test to compare survival in two or more groups and the Cox proportional hazard model was used to determine factors associated with time to death of early neonate. The hazard ratio with 95% confidence level was used to declare statistical significant association.

Result: A total of 416 early neonates were included in the study. There were 224(53.8%) male neonates. The overall early neonatal mortalities were 47(11.3%), of which 23.41% of them died on the first day of age. early neonatal mortality was associated with neonatal sepsis (AHR=3.349,95% CI,1.842-6.089), low birth weight of neonates (AHR=3.563%,95% CI,1.925-6.595), unable to breastfeeding during delivery (AHR=4.094,95% CI,1.603-10.452), neonatal respiratory distress (AHR=4.018,95% CI,1.733-9.371), perinatal asphyxia (AHR =2.540,95% CI,1.186-5.847).

Conclusion: The causes of early neonatal death described in this study were preventable. Managing low birth weight, initiating exclusive breastfeeding, refining quality of service, and confirming the continuity of care are recommended to increase the survival of neonates.

Keywords: time to death, early neonatal, predictors, nicu

NICU, neonatal intensive care unit; ENM, early neonatal mortality; HEP, health care provider

Early neonatal death accounts for 75% of all neonatal deaths worldwide.⁵ Globally, 2.4 million children died in the first month of life in 2019, with approximately 6,700 neonatal deaths every day, most of which close to three quarters dying in the first week, with one-third dying on the first day.⁶ Over the past decades, mortality during the early neonatal period declined more slowly than late neonatal and post neonatal infant mortality.⁷ All inclusive, stamped incongruities in neonatal mortality exist over districts and nations. Across the globe, neonatal mortality rates range from 5 per 1000 in developed countries to 50 per 1000 in the least developed countries.⁸ Sub-Saharan Africa is by far the risk region to be born. In sub-Saharan Africa, babies are more than 7 times as likely to die on the day they are born as babies in industrialized countries.⁹

In 2015, the Joined together Countries summit on maintainable improvement formally received the 2030 plan for maintainable advancement. In the Sustainable Development Goals, the goal is to end preventable neonatal and under-five child deaths by 2030, with all nations pointing to decrease neonatal mortality to at slightest as moo as 12 deaths per 1000 live births and under-five mortality to at slightest as moo as 25 deaths per 1000 live births.¹⁰ However, Sub-Saharan Africa is the only SDG region to see no decline in the number of neonatal deaths from 1990 to 2019.⁶

Early neonatal survival is the most important indicator of improving health care during childbirth and the life quality of a given population.¹¹ There are various factors that may be associated with early neonatal mortality and they have been the reasons for the wide variation in mortality rates among the health facilities reporting. One of the three major causes of ENM is intrapartum-related neonatal death, especially birth asphyxia^12,13. Another significant risk factor is preterm birth, which leads to small size at birth, or a timely delivery but that the child is small-for-gestational-age^12,14 often caused by malnutrition of the mother.¹⁵ Small size at birth is correlated to neonate hypothermia and infections, which both increase the risk of the neonate dying during the ENM period.¹⁶ The third major cause of ENM is severe infections, where sepsis and pneumonia are the two major killers, but tetanus and diarrhea are also important causes.^12,13

Ethiopia is one of the sub-Saharan countries that have a high neonatal mortality rate, and even one of the top ten countries in Africa.¹⁷ To reduce neonatal deaths, the Ethiopian government has implemented many health interventions such as training midwives, enhancing the referral system, integrating health services, implementing packages of the Health Extension Program (HEP), and routine immunization.¹⁸ In spite of many efforts by the government and other stakeholders, a non-significant and very sluggish decline has been achieved. Distinguishing the determinants and causes of early neonatal mortality at the nearby setting may be an exceptionally vital and convenient issue. Past literature on the subject matter has continued to give more emphasis to either under-five mortality^19–21 or neonatal mortality^22–24 ignoring the significant proportion of early neonatal deaths as a critical component of both neonatal and under-five mortalities.

Therefore, understanding early neonatal mortality in connection to these factors is pivotal. This is because there are highly feasible and cost-effective interventions that could avert this mortality, and this can only be achieved if countries adopt locally relevant and focused interventions that are guided by evidence. So this study was interested in dealing with an estimate of the survival rate in addition to identifying primary etiological factors responsible for early neonatal mortality and the relative contribution of these factors to early neonatal mortality. So the result of this study provides a crucial data of survival and predictors of survival of neonate to the health care provider, at a different level, so that it will allow for closer follow up of high-risk patients and more targeted interventions will be made that reduces mortality and improve quality of life of patients.

Study design, setting and period

A retrospective follow up study was conducted in the Dessie referral hospital of Dessie town, Amhara Regional State, Ethiopia from January 4, 2017, to January 6, 2020.

Population

The source population was all early neonates who were admitted to the Dessie referral hospital intensive care unit, and the study population was all early neonates who were admitted to Dessie Referral Hospital from January 4, 2017-January 6, 2020.

Inclusion and exclusion criteria

All early neonates who were admitted to the neonatal intensive care unit in Dessie referral hospital were included, and early neonates with incomplete data were excluded.

Sample size determination and sampling procedures

The total sample size was determined by Epi info version 7.2, which included 416 early neonates. Based on the assumptions from the previous study of 2.03 hazard-ratios, 10% outcome in an unexposed group with a gestational age of neonate.²⁵ There was an accepted error of 5%, power of 80%, and 95% confidence level.

All the study participants who were found in neonatal registration logbooks of the past three years (January 4, 2017 – January 6, 2020) were checked for chart retrieval. After that, a chart that had accessed was taken for sampling by using computer-generated random numbers was used to select the predetermined required sample size. Charts with incomplete data were excluded, and the next random number was taken.

Study variable

Time to death of early neonates was the dependent variable and the independent variable included neonatal characteristics and maternal characteristics.

Data collection tools and Quality control

Data were collected by using a checklist developed based on the maternal card, neonatal card, and neonatal patient registration logbook which is adopted by the federal minister of health. Its quality was controlled by designing proper data collection tools, pre-testing, and continuous supervision and before actual data collection, training was provided to BSC nurse data collectors for two days on the data collection techniques to familiarize data collectors with the tool. Data were collected by reviewing maternal cards, neonatal cards, and a neonatal registration log book.

Data processing and analysis

Data were entered in Epi data 3.1 and it was checked, cleaned, and edited before analysis, and it was exported to SPSS version 23 for analysis. Describe the patient cohort characteristics in terms of the mean value for continuous data and percentage for categorical data. The survival time was calculated in a day using the time interval between the date of birth and the end of the follow-up period. A life table was applied to estimate the probability of the survival of the patient and a log-rank test was used to compare the KM curve for two or more categories of patients. The multivariable Cox proportional hazard model was used to determine the relationship between the independent variable and the outcome variable. The Cox proportional hazard assumption was checked graphically. The first bivariable analysis was made for each independent variable to the outcome variable, and those variables resulting p-value less than 0.3 was entered to the multivariable Cox regression model. In the final model, those variables with a p-value less than 0.05 were considered as statistically significant, and it was presented by hazard ratio (HR), with a 95% confidence level (CI) to show the strength and direction of the association.

Ethical consideration

Ethical clearance was obtained from the Institutional Review Board (IRB) of Wollo University, College of medicine, and health science. Following the approval by the IRB, the official letter of co-operation was written to the concerned bodies by Wollo University. As the study was conducted through a review of medical records, the individual patients were not subjected to any harm as far as confidentiality is kept. To preserve confidentiality, no personal identifier was used on the data collection form, and the recorded data was not accessed by a third person except the principal investigator.

Data were collected from all 416 study participants born between January 4, 2017 and January 6, 2020.

Neonatal characteristics

The mean age of the neonates was found to be 3.62 days (SD=±4.4). Among the cohort, 224 (53.8%) were male. Of these (416) neonates, 20.9% of them had neonatal sepsis, 17.1% had respiratory distress syndrome and 15.1% had perinatal asphyxia. Regarding birth weight, 65.1% of them had a normal birth weight (2.5-4 kg). With respect to their anthropometrics, more than three-fourth of (91.5%) neonates had the appropriate size for gestational age (Table 1).

Maternal characteristics

The mean age of mothers during delivery was 26.95 years (SD=±4.4). Among the total number of mothers recruited for the study, 98.3% had received two doses of the TT vaccine, 98.6% had breastfed during delivery, 1.7% had a history of stillbirth, and 1% had a history of abortion (Table 2).

Follow up outcome

All the study subjects were followed for seven consecutive days, but for different reasons participants were retained in the cohort for different lengths of time, they stay for a minimum of one, a maximum of seven days, and the median time was unreached (Figure 1). From a total of 416 cases, 22(5.3%) of them were referred to higher health facilities, and the remaining 11(2.6%) of them were against the treatment.

Figure 1 The median survival time for early neonate in Dessie Referral hospital NICU in 2020.

Survival

As it’s mentioned before, the study subjects were stayed for different lengths of time in the cohort making the general population at risk of 1508 person day follow up. The total of death in the study period was 47 this makes the instantaneous hazard rate 47/1508=9.3 per day per 1000 Neonates. In this study, 23.4% of the death occurred on the first day, and 76.6% of death occurred between 2-7 days. As per the analysis of the data from all study participants (416) the cumulative probability at 1 and 7 days is 97% and 77% respectively (Table 3).

Predictors of Early neonatal mortality

The relation between the main variable and the risk of death was analyzed using the bivariable Cox proportional hazard model. Only a total of 383 neonates who had different lengths of follow-up were included in bivariable and multivariable Cox regression analysis by excluding those transferred and against patients.

In the final multivariable Cox regression analysis, Neonatal sepsis (P<0.001), neonatal birth weight (p<0.001), neonatal respiratory distress syndrome (p=0.001), Perinatal asphyxia (p=0.021), and mother’s breastfeeding during delivery (p=0.003) remain in the final model as statistical significant predictor with the timing of early neonatal mortality (Table 4).

Hazard Function of the Predictor Variables

Neonatal sepsis had a significant association with the timing of early neonatal mortality, with statistical significant difference among the categories. (diseased vs non diseased) (Log-rank test, p value <0.001) (Figure 2).

Figure 2 Hazard plot of neonatal sepsis among different categories of early neonate, 2020.

Perinatal asphyxia was the other determinant factors which had a significant association with the timing of early neonatal mortality with statistical significant difference among categories (diseased vs non diseased)(log-rank test, p value <0.001) (Figure 3).

Figure 3 Hazard plot of perinatal asphyxia among different categories of early neonate, 2020.

The birth weight of the neonate was the other determinate factor which had a significant association with the timing of early neonatal mortality with statistical significant difference among different categories. (>=2.5kg vs<2.5kg) (log-rank test, p value <0.001) (Figure 4).

Figure 4 Hazard plot of birth weight among different categories of early neonate, 2020.

The other categories of predictor were which had an association with the timing of early neonatal mortality was breastfeeding of a mother during delivery, there was a statistical significance difference among categories (breastfeed vs non breastfeed) (log-rank test value <0.001) (Figure 5).

Figure 5 Hazard plot of breast feeding during delivery among categories of early neonatal, 2020.

The other categories of predictor were neonatal respiratory distress syndrome which has an association with the timing of early neonatal mortality with the statistical difference among categories (diseased vs non diseased) (log- rank test value <0.001) (Figure 6).

Figure 6 Hazard plot of neonatal respiratory syndrome among categories of early neonatal, 2020.

In this study, the early neonatal mortality proportion was 11.3% with 95%CI (8.4-14.7). This proportion was lower when to compare with the study done in Black Lion specialized hospital.²⁶ This difference might be due to the complicated cases that were referred to Black Lion Hospital. But Dessie referral hospital NICU mortality proportion was comparable with the study done in Gondar and Bahir Dar.^27,28 However, it was higher than a study conducted in southwest Ethiopia.²⁹ These variations might be due to methodological differences among studies and dissimilarities in socio-cultural, health service utilization, and economical variations among study participants.

The median survival time of the study participants was found to be unreached. Which was similar to the study finding in Tigray Ethiopia.³⁰ But the median survival time was reached in a different place like North West Ethiopia.³¹ This difference might be due to the short follow- up period and a higher number of right-censored cases.

In accordance with this study finding, the neonates who develop neonatal sepsis had a commutative hazard was found to be 3.34 times at high risk of death when compared to those who do not develop neonatal sepsis. The relationship between neonatal sepsis and mortality among neonates was also shown in similar studies done in Gondar and Eritrea.^27,32 Being infected with neonatal sepsis indicates that the neonates cannot suck, and the neonate had vomiting, diarrhea, poor perfusion, and hypothermia, which are associated with increased risk of mortality.

Neonates with low birth weight had a commutative hazard was found to be 3.56 times high risk of death when compared to those who had normal birth weight. The association observed between low birth weight and mortality in this study was similar to previous studies done in Addis Ababa, South West Ethiopia, Eretria, and Egypt.^26,29,32,33 Low birth weight indicators of poor intrauterine growth and low birth weight of neonates were at high risk of a motor, cognitive, behavioral, and emotional problem, which leads to mortality.

Those neonates who develop respiratory distress syndrome had a commutative hazard was found to be 4 times the high risk of death when compared to those who do not develop respiratory distress syndrome. This finding was supported by a study done in South West Ethiopia and Egypt^29,33. Various studies showed this respiratory distress syndrome related to preterm, poor prenatal care, and low maternal body mass index which increased the risk of mortality of neonates.^34,35

Neonates who developed neonatal asphyxia had a commutative hazard was found to be 2.54 times the high risk for early neonatal mortality than those who did not develop neonatal asphyxia. This finding is also supported by studies done in Addis Ababa, Gondar and South West Ethiopia^26,27,29. In different studies showed this perinatal asphyxia is related to antepartum, intrapartum, and postnatal related factors which increase the risk of neonatal mortality.^36,37

A mother who did not breastfeed during the first hour of delivery increased neonatal risk of death by the commutative hazard of 4.1 times when compared with those who got breastfeeding during delivery. The association observed between breastfeeding and neonatal mortality in this study is supported by studies done in Gondar, Bahir Dar, and North Shoa.^27,28,38 Breastfeeding was important to boost the immunity of the neonates; therefore, no breastfeeding indicates that the neonates were exposed to infection which increases neonatal morbidity and mortality.

First and foremost we would like to express my deepest gratitude to Wollo University for creating this opportunity. our special thanks also go to Dessie referral hospital human resource management office and neonatal intensive care unit staff members for providing us important information for writing this thesis.

All authors contributed to data analysis, drafting or revising the item, have agreed at the journal to which the object could be submitted, gave final approval for the model to be published, and agreed to be responsible for all components of the work.

No financial support was provided.

The datasets employed in the current study can be available from the corresponding author upon a reasonable request.

The authors declared no potential competing interest with respect to the publication of this paper.

1. Ganatra HA, Stoll BJ, Zaidi AKM. International perspective on early-onset neonatal sepsis. Clin Perinatol [Internet]. 2010;37(2):501–23. Available from: http://dx.doi.org/10.1016/j.clp.2010.02.004
2. Roseveare MP. Perinatal and neonatal mortality. Br Med J. 1980;281(6238):515.
3. Beck D, Ganges F, Goldman S, Long P. Care of the newborn - Saving Newborn Lives. Reference Manual. 2004;
4. Rajaratnam JK, Marcus JR, Flaxman AD, Wang H, Levin-Rector A, Dwyer L, et al. Neonatal, postneonatal, childhood, and under-5 mortality for 187 countries, 1970-2010: a systematic analysis of progress towards Millennium Development Goal 4. Lancet [Internet]. 2010;375(9730):1988–2008. Available from: http://dx.doi.org/10.1016/S0140-6736(10)60703-9
5. Lloyd LG, de Witt TW. Neonatal mortality in South Africa: How are we doing and can we do better? South African Med J. 2013;103(8):518–9.
6. UN Inter-agency Group for Child Mortality Estimation, UNICEF, WHO, WBO. Levels & Trends in Child Mortality Estimation Child Mortality [Internet]. Un Igme. 2020. 48 p. Available from: https://www.unicef.org/media/79371/file/UN-IGME-child-mortality-report-2020.pdf.pdf
7. Lohela TJ, Nesbitt RC, Pekkanen J, Gabrysch S. Comparing socioeconomic inequalities between early neonatal mortality and facility delivery: Cross-sectional data from 72 low- and middle-income countries. Sci Rep [Internet]. 2019;9(1):1–11. Available from: http://dx.doi.org/10.1038/s41598-019-45148-5
8. Rajab AM, .Ghareba AM. Neonatal Mortality Rate in the Special Care Baby Unit. J Med Sci Clin Res [Internet]. 2013;1(5):195–208. Available from: http://jmscr.igmpublication.org/v1-i5/1 jmscr.pdf
9. Children S the. Surviving the First Day [Internet]. Save the Children, State of the World’S Mothers 2013. 2013. 1–88 p. Available from: www.savethechildren.org
10. United Nations. The sustainable development goals report 2019. United Nations Publ issued by Dep Econ Soc Aff [Internet]. 2019;64. Available from: https://undocs.org/E/2019/68
11. P. Barker,N. van den Broek,S. Currie,C. Hanson,K. Hill KH. Standards for improving quality of maternal and newborn care in health facilities. World Health Organization. 2016.
12. E Lawn , S. Cousens JZ. MDGs and newborn babies. Lancet [Internet]. 2005;365:891–900. Available from: http://www.measuredhs.com
13. Sankar MJ, Natarajan CK, Das RR, Agarwal R, Chandrasekaran A, Paul VK. When do newborns die? A systematic review of timing of overall and cause-specific neonatal deaths in developing countries. J Perinatol [Internet]. 2016;36(S1):S1–11. Available from: http://dx.doi.org/10.1038/jp.2016.27
14. Lawn JE, Blencowe H, Oza S, You D, Lee ACC, Waiswa P, et al. Every newborn: Progress, priorities, and potential beyond survival. Vol. 384, The Lancet. 2014. p. 189–205.
15. Black RE, Allen LH, Bhutta ZA, Caulfield LE, de Onis M, Ezzati M, et al. Maternal and child undernutrition: global and regional exposures and health consequences. Lancet. 2008;371(9608):243–60.
16. Bhutta ZA, Das JK, Bahl R, Lawn JE, Salam RA, Paul VK, et al. Can available interventions end preventable deaths in mothers, newborn babies, and stillbirths, and at what cost? Lancet. 2014;384(9940):347–70.
17. Berhe M, Medhaniye AA, Kahsay G, Birhane E, Abay M. Essential neonatal care utilization and associated factors among mothers in public health facilities of Aksum Town, North Ethiopia, 2016. PLoS One. 2017;12(4):1–11.
18. Draft F. Federal Democratic Republic of Ethiopia Ministry of Health Health Sector Development Program IV October 2010 Contents. 2014;(October 2010).
19. Bereka SG, Habtewold FG. Under-Five Mortality of Children and its Determinants in Ethiopian Somali Regional State, Eastern Ethiopia. Heal Sci J. 2017;11(3):1–8.
20. Ayele DG, Zewotir TT. Comparison of under-five mortality for 2000, 2005 and 2011 surveys in Ethiopia. BMC Public Health [Internet]. 2016;16(1):1–10. Available from: http://dx.doi.org/10.1186/s12889-016-3601-0
21. Ababa A. College of natural and computational science department of statistics determinants and trends of under-five child mortality in ethiopia by Mulugeta Tadesse. 2018;
22. Hibstu DT, Ayele TA, Mengesha ZB. Determinants of Neonatal Mortality in Ethiopia: A Case Control Study, 2013. OALib. 2014;01(06):1–9.
23. Yirgu R, Molla M, Sibley L. Determinants of neonatal mortality in rural Northern Ethiopia: A population based nested case control study. PLoS One. 2017;12(4):1–10.
24. Kolobo H, Chaka T, Kassa R. Determinants of neonatal mortality among newborns admitted to neonatal intensive care unit Adama, Ethiopia: A case–control study. J Clin Neonatol. 2019;8(4):232.
25. de Castro ECM, Leite ÁJM, de Almeida MFB, Guinsburg R. Perinatal factors associated with early neonatal deaths in very low birth weight preterm infants in Northeast Brazil. BMC Pediatr. 2014;14(1).
26. Worku B, Kassie A, Mekasha A, Tilahun B, Worku A. Predictors of early neonatal mortality at a neonatal intensive care unit of a specialized referral teaching hospital in Ethiopia. Ethiop J Heal Dev. 2012;26(3):200–7.
27. Demisse AG, Alemu F, Gizaw MA, Tigabu Z. Patterns of admission and factors associated with neonatal mortality among neonates admitted to the neonatal intensive care unit of University of Gondar Hospital, Northwest Ethiopia. Pediatr Heal Med Ther. 2017;Volume 8:57–64.
28. Tewabe T, Mehariw Y, Negatie E, Yibeltal B. Neonatal mortality in the case of Felege Hiwot referral hospital, Bahir Dar, Amhara Regional State, North West Ethiopia 2016: A one year retrospective chart review. Vol. 44, Italian Journal of Pediatrics. Italian Journal of Pediatrics; 2018. p. 1–5.
29. Debelew GT, Afework MF, Yalew AW. Determinants and causes of neonatal mortality in jimma Zone, Southwest Ethiopia: A multilevel analysis of prospective follow up study. PLoS One. 2014;9(9).
30. Mengesha HG, Wuneh AD, Lerebo WT, Tekle TH. Survival of neonates and predictors of their mortality in Tigray region, Northern Ethiopia: Prospective cohort study. BMC Pregnancy Childbirth [Internet]. 2016;16(1):1–13. Available from: http://dx.doi.org/10.1186/s12884-016-0994-9
31. Ayalew S YS. Survival Analysis of Premature Infants Admitted to Neonatal Int ensive Care Unit (NICU) in Northwest Ethiopia using Semi-Parametric Fr ailty Model. J Biom Biostat. 2015;06(01):1–12.
32. Shah S, Zemichael O, Meng HD. Factors associated with mortality and length of stay in hospitalised neonates in Eritrea, Africa: A cross-sectional study. BMJ Open. 2012;2(5):1–9.
33. Mohamed, Eman M and Soliman, Asmaa MA and El-Asheer OM. Predictors of mortality among neonates admitted to neonatal intensive care unit in pediatric Assiut University Hospital, Egypt. Am J Sci. 2011;7(6):606--11.
34. Kommawar A, Borkar R, Vagha J, Lakhkar B, Meshram R, Taksandae A. Study of respiratory distress in newborn. Int J Contemp Pediatr. 2017;4(2):490.
35. Kumar A, Bhat BV. Epidemiology of respiratory distress of newborns. Indian J Pediatr. 1996;63(1):93–8.
36. Lawn JE, Lee ACC, Kinney M, Sibley L, Carlo WA, Paul VK, et al. Two million intrapartum-related stillbirths and neonatal deaths: Where, why, and what can be done? In: International Journal of Gynecology and Obstetrics. 2009. p. 5–19.
37. WHO. Guidelines on basic newborn resucitation. WHO Libr Cat Data. 2012;1–61.
38. Kolola T, Ekubay M, Tesfa E, Morka W. Determinants of neonatal mortality in North Shoa Zone, Amhara regional state, Ethiopia. PLoS One. 2016;11(10):1–11.