Exploring spatiotemporal distribution of under-five mortality in Ethiopia: further analysis of Ethiopian Demographic and Health Surveys 2000, 2005, 2011 and 2016

Objective The aim of this study was to explore spatiotemporal distribution of under-five mortality in Ethiopia using data from four (2000, 2005, 2011 and 2016) consecutive demographic and health surveys. Methods A total of 41 498 children were included from four consecutive Ethiopian Demographic and Health Surveys. The geospatial analysis was conducted by using ArcGIS V.10.7 and saTScan V.9.6. Thus, spatial autocorrelation, hotspot analysis, spatial interpolation and spatial scan statistics were carried out for each survey separately to show the temporal pattern of geographically risk areas of under-five mortality in Ethiopia. Finally, the most under-five mortality rate (U5MR) risk areas in each survey period were mapped. Results Under-five mortality was spatially clustered in Ethiopia (Moran's Index: 0.046-0.096, p<0.01). The Benishangul-Gumuz region was consistently at a higher risk in the last two decades. Additional hotspot areas were detected at Afar and Amhara (in 2000, 2005, 2016), at Gambala (in 2011) and the South Nation Nationality and People's (SNNP) Region (in 2016). Moreover, 160 primary clusters were identified. Of these, 85 clusters (log-likelihood ratio (LLR)=13.10, p<0.01) were from Benishangul-Gumuz and Amhara regions (in 2000); 67 clusters (LLR=12.93, p<0.01) were from Afar and Amhara regions (in 2005); 4 clusters (LLR=10.54, p<0.01) were from Benishangul-Gumuz region (in 2011); and another 4 clusters (LLR=11.85, p<0.01) were from Afar region (in 2016). Conclusion High-risk areas were detected mainly in the Benishangul-Gumuz and Afar regions. As a result, designing under-five population targeted intervention programmes in those high-risk geographical regions was vital to reduce under-five mortality in Ethiopia.