RESUMO
Neonatal mortality is the death of a live-born baby within the first 28 days of birth. For the selected households, neonatal mortality was collected from children aged 0-28 days and women aged 15-49. The neonatal period is a significant 4-week period in human life because it carries a greater mortality risk. To identify the determinant factors of neonatal mortality in Ethiopia based on EDHS 2016 data with the application of count regression models. In this study, all neonates in Ethiopia were born within the 5 years preceding EDHS 2016 of the source population in the selected EAs from September to December 2015. Count regression models were used to analyze the data. A total of 10,641 live-born neonates within the previous 5 years of EDHS 2016 had neonatal mortality of women aged 15-49, which was considered in the study to be 7193. The data were found to have excess zeros (96.6%), and the variance (0.052) was higher than its mean (0.04). The count regression model (ZINB) was best fitted to the data with maximum likelihood parameter estimation methods. The average neonatal mortality difference in multiple births was increased by IRR = 8.53 times compared with a single birth. The average number of neonatal deaths experienced during breastfeeding was lower (IRR = 0.38) than that experienced by mothers who did not experience breastfeeding their child. The average neonatal mortality difference in rural residences was increased by IRR = 3.99 times compared to urban mothers' residences. In this study, the prevalence of Neonatal mortality in Ethiopia was higher. For selected ZINB count regression models of explanatory variables, such as multiple birth types, having rural residence factors of neonatal mortality increased the risk of death. However, having early breastfeeding, a female household head, and antenatal visits (1-4) and (5-10) during pregnancy decrease the risk of neonatal death.
