Spatial clustering of zero dose children aged 12 to 59 months across 33 countries in sub-Saharan Africa: A multiscale geographically weighted regression analysis

by Chamberline E. Ozigbu, Zhenlong Li, Bankole Olatosi, James W. Hardin, Nicole L. Hair

While prior studies have identified sociodemographic correlates of zero-dose status within populations in sub-Saharan Africa (SSA), few have applied spatial regression techniques to explore geographic variability in these relationships. We aimed to address this gap using data from Demographic and Health Surveys conducted in SSA between 2010 and 2020. Our sample comprised children aged 12–59 months in 33 countries and 329 survey regions. Data were aggregated to the first-level administrative unit prior to analysis. First, using ordinary least squares regression, we documented global relationships between theoretically important sociodemographic characteristics and zero-dose prevalence. Next, we identified patterns, i.e., geographic clustering, of zero-dose prevalence. Finally, using multiscale geographically weighted regression, we described spatial variability in relationships between sociodemographic characteristics and zero-dose prevalence. We detected 27 regions with higher than expected concentrations of zero-dose children. All but one of these hot spots were observed in 7 Western and Central African countries; only 1 was located in an Eastern African country. Regions with higher proportions of mothers with no antenatal care visits were consistently found to have higher rates of zero-dose children. In contrast, relationships between zero-dose prevalence and indicators of religious affiliation, delivery site, maternal age, maternal education, and maternal employment were found to vary locally in terms of their strength and/or direction. Study findings underscore spatial disparities in zero-dose prevalence within SSA and, further, highlight the importance of geographically informed strategies to effectively address immunization gaps. Implementing targeted interventions based on regional sociodemographic dynamics is crucial for achieving comprehensive vaccination coverage in SSA.