Spatial econometric modeling of malaria distribution in Burkina Faso

Abstract

This study focuses on the spatial distribution of malaria rate in Burkina Faso. For this purpose, we used 2020 geographical and confirmed malaria cases data from the National Institute of Statistics and Demography of Burkina Faso. We also accessed climate data on the French weather history website. We deployed relevant spatial statistical tools to address the notions of neighborhood matrix, spatial autocorrelation, and spatial heterogeneity between geographical observations. Ordinary Least Squares (OLS) regression model, Spatial Error Model (SEM), Spatial Autoregressive model (SAR), and the spatial Durbin Model (SDM) were computed using cross-validation to ensure the reliability of our findings. The Akaike information criterion (AIC) was used to select the most appropriate model for our study. The specification tests conclude that there is a spatial dependence between the observations. The SDM was chosen as the best-fitting data for modeling geolocated malaria rates. This outcome reveals that environmental indicators, population literacy rate, and rural population size by region significantly affect the country’s geolocated malaria rate. Policymakers can use these findings to make informed decisions related to malaria spread.