CAUSAL MACHINE LEARNING ALGORITHMS FOR ROBUST EPIDEMIOLOGICAL POLICY IMPACT ESTIMATION

We examine how epidemiological policy implementation influences disease control outcomes using a causal machine learning framework applied to national public health data in Ghana. The analysis relies on the Global Epidemiological Policy Intervention Dataset covering epidemiological policy actions and health outcomes from 2020 to 2025 and estimates the Policy Impact Causal Learning Model to evaluate how vaccination strategy, public health communication, and disease surveillance infrastructure shape epidemiological control outcomes under different governance conditions. Evidence indicates that expansion of vaccination coverage, stronger communication outreach, and improved surveillance capacity correspond with sustained declines in infection rates and mortality while improving outbreak containment and health system resilience. Governance capacity strengthens these effects by improving policy coordination and implementation efficiency across health institutions. The results introduce a structured causal learning framework that explains how coordinated policy systems rather than isolated interventions drive epidemiological control performance. The model offers a scalable analytical tool for evaluating public health policies across national health systems and provides practical guidance for governments seeking to strengthen epidemic preparedness and institutional response capacity.