Spatial variation, pooled prevalence, and factors associated with perinatal mortality in Sub-Saharan Africa, evidence from demographic and health surveys 2015–2023: a geospatial regression approach
Belayneh Jejaw Abate, Alemakef Wagnew Melesse, Helen Brhan, Muluken Chanie Agimas
Abstract
Background Sub-Saharan Africa (SSA) bears the greatest burden of perinatal mortality in the world, and the magnitude of the problem varied based on geographical location. A detailed understanding of spatial variation is important to improve the targeting of interventions, to identify the most affected community, and for designing evidence-based health policies. Hence, this study aimed to assess pooled prevalence, spatial variation, and factors contributing to perinatal mortality in SSA. Methods A cross-sectional study using Demographic and Health Survey datasets (2015–2023) of 25 SSA countries with a total of 201,566 weighted samples was used for this study. The global spatial autocorrelation was explored using global Moran's-I, and the spatial variation of perinatal mortality was examined using hot spot analysis (Local Getis-Ord Gi∗ statistic). Spatial regression analyses (ordinary least squares, spatial error model, spatial lag model, geographically weighted regression, and multiscale geographically weighted regression) were conducted. Models were assessed using corrected Akaike information criteria and adjusted R 2 . A p-value threshold of 0.05 was set to identify statistically significant spatial predictors, and the corresponding local coefficients were illustrated on a map. Findings The pooled prevalence of perinatal mortality in SSA was 46.63 per 1000 total births (95% CI: 42.48, 51.17), and its spatial distribution was found to be clustered (Global Moran's I = 0.18, p < 0.01). Significant hotspot areas were located in Nigeria, Madagascar, Rwanda, Malawi, Burundi, Gambia, Uganda, Côte d'Ivoire, Angola, Ethiopia, Burkina Faso, and Senegal, while significant cold spots were located in Kenya, Gabon, South Africa, Ghana, Mali, and Mauritania. The multi-scale geographic weighted regression model explained 85% of the spatial variation of perinatal mortality in SSA. No antenatal care visit, birth interval less than 15 months, women undergoing cesarean section delivery, unemployed women, and households without children were significant spatial predictors of perinatal mortality in SSA. Interpretation Perinatal mortality in SSA was high and varied across regions. We identified five predictors for perinatal mortality that might be a priority for policymakers. Enhancing antenatal care and family planning services and empowering women through employment opportunities is crucial to decreasing perinatal mortality in the region. Funding None.