Quantifying Disease Burden in Low-Resource Settings: Statistical Insights and Approaches with Model-Based Geostatistical Modeling

Accurately quantifying disease burden is essential for guiding public health interventions and resource allocation, particularly in low- and middle-income countries (LMICs). However, LMICs often lack comprehensive and reliable disease registries, leading to biased estimates, as populations with better healthcare access tend to be overrepresented. This paper explores the application of Generalized Linear Geostatistical Modelling (GLGM), an extension of Generalised Linear Models (GLM), to address these data limitations and improve the estimation of disease burden. GLGM integrates statistical modelling with geospatial analysis to incorporate alternative data sources, such as hospital records, population surveys, and demographic data, allowing for more precise disease mapping even without complete registries. Using spatial correlations and adjusting for covariates such as demographic and environmental factors, GLGM enables the estimation of the disease burden at high resolution (up to 1 km2). The use of GLGM in LMICs offers multiple benefits, including identifying disease hotspots, evaluating intervention strategies, and prioritizing resources for the most affected populations. In addition, it facilitates the exploration of complex relationships between disease outcomes and risk factors, providing deeper insights into disease patterns. This paper underscores the value of statistical approaches, particularly GLGM, in improving the estimation of the disease burden in LMICs with incomplete registers. We demonstrate its application by mapping anaemia prevalence in children under five years of age in Malawi, using data from the 2015/16 Demographic and Health Survey. By leveraging available data and spatial techniques, GLGM provides policymakers and public health practitioners with a valuable tool for informed decision-making and equitable resource allocation.