Leveraging Artificial Intelligence to Strengthen MEL Systems in Immunization Programs: Insights from Cameroon’s Fragile Contexts

Abstract

This study introduces an Artificial Intelligence (AI)-enhanced framework for Monitoring, Evaluation, and Learning (MEL) in fragile immunization contexts, using Cameroon as a case study. Drawing on routine service delivery and community-level data, we trained a Random Forest model to predict zero-dose hotspots and assess the drivers of immunization gaps. The predictive features included geographic accessibility, security risks, community engagement, and health system capacity. The results highlight that the distance to vaccination posts, community leader involvement, and availability of cold-chain infrastructure are key determinants of coverage. The model demonstrated a strong classification performance, offering actionable insights for targeted interventions. While this approach reduces reliance on manual triangulation and enhances real-time decision-making, it requires careful handling of data quality and contextual constraints. This research provides a practical framework for applying AI to improve equity, efficiency, and planning in fragile immunization systems.