Infant mortality and its determinants in Uganda 2016: Using a geographically weighted regression approach

Abstract

Infant mortality (IM) represents the overall health of a country or region as it relates to access to medicine, health care, and clean water in a population. IM remains understudied in many areas of Uganda, as many studies are from urban the capital (Kampala). The long-term goal of this research is the mitigation of IM and poor pregnancy outcomes in Uganda. Insights gained from geographic distribution of IM will allow adaptation of diagnosis, treatments, and interventions within the studied areas. Through using OLS and geographically weighted regression, this study has explored the significant factors and their heterogeneous and scaling effects in 2016 across Uganda. The empirical findings from this study include a significant association between IM and both being unmarried and preferring to speak Luganda when interviewed. Those unmarried may lack a social network to assist with income, childcare, and household chores representing decreased resources. Additionally, being interviewed in Luganda was associated over a large geographic area, which may represent not being comfortable in English, which is the language of education, commerce, and presumably health care, thus suggesting a disconnect with health care settings. These data suggest that strides can be made in Uganda by providing targeted resources to areas with high rates of unmarried mothers and those areas with high rates of Luganda as their language of choice.