Spatiotemporal variation and decomposition of early neonatal mortality in Ethiopia using demographic health survey data.

Early neonatal mortality is a serious public health issue in Ethiopia. Therefore, this study aimed to map the regional disparities and identify factors contributing to early neonatal mortality in Ethiopia over time. Ethiopian demographic and health survey (2000-2019) dataset were utilized. A total of 80,286 early neonates was included in this study. Logit based decomposition analysis was employed to understand the contributing factors for the change in early neonatal mortality over time. Getis Ord GI* statistic was performed to identify the cold and hot spots of the early neonatal mortality in Ethiopia. In addition, kriging interpolation was used to predict the burden of early neonatal mortality in the unsampled areas of the country based on the observed data. Early neonatal mortality trends in Ethiopia has been decline from 43 in 2000 to 33 in 2019 per 1000 live births. It was spatially clustered, with significant hotspots in the Benishangul Gumuz and some areas of Oromia, Tigray, Amhara and Somali regions. In the logit multivariable decomposition analysis babies born in rural (B - 0.0002, 95% CI - 0.002 to - 0.001), Women had partner (B - 0.0005, 95% CI - 0.0007 to - 0.0003), preceding birth interval ≥ 2 years (B - 0.008, 95% CI - 0.005 to - 0.001), Health facility delivery (B - 0.001, 95% CI - 0.003 to - 0.001), had ANC visits (B - 0.02, 95% CI - 0.03 to - 0.01), early initiations of breastfeeding (B - 0.002, 95% CI - 0.003 to - 0.001), multiple pregnancies (B - 0.002, 95% CI - 0.003 to - 0.001), and mothers education higher (B - 0.004, 95% CI - 0.006 to - 0.002) were a substantial factors that contribute to the change in the decline in early neonatal mortality in Ethiopia over time. Strengthen maternal and newborn care, expand skilled birth attendance, enhance antenatal and postnatal services, and utilize spatial evidence to inform targeted policy and resource allocation.