Resource Allocation Models for Primary Healthcare in the Context of Global Health Inequity

Abstract

This investigation creates a model for resource allocation powered by AI to help resolve global disparities in the management of chronic diseases in resource-poor areas. This study uses multi-country healthcare data from Asia, Africa, and Latin America. It applies reinforcement learning with equity constraints to optimise resource allocation for diabetes, hypertension, chronic obstructive pulmonary disease (COPD), and cardiovascular diseases (CVD). The model combines machine learning-based optimisation with principles of distributive justice, incorporating equity-weighted socioeconomic vulnerability indices and spatiotemporally-augmented real-time disease burdens. Experimental results show a reduction of 42.3% of chronic disease related Disability Adjusted Life Years (DALYs) with more than half of the improvement arising from the poorest population quintiles when compared to traditionally allocated methods. The model achieved 38.7% higher resource allocation efficiency while observing patterns of equitable allocation. Validation across countries confirmed the adjustability of the model to different health care systems using dynamic hyper-parameter tuning proving the robustness of the approach. The results empirically validate the potential in AI-enabled solutions as radical, transformative drivers towards the attainment of universal health coverage and the persistent issues of under-provisioning in lower-resourced contexts.