Developing Causal Machine Learning Models in Health Informatics to Assess Social Determinants Driving Regional Health Inequities and Intervention Outcomes

Persistent regional health inequities stem from complex, interrelated social determinants such as income, education, housing, access to care, and environmental exposure. Traditional statistical approaches often capture correlations without disentangling the underlying causal mechanisms that drive these disparities. Causal machine learning (CML) provides a transformative pathway for advancing health informatics by combining the predictive power of artificial intelligence with the interpretive rigor of causal inference. Through methods such as structural causal models (SCMs), directed acyclic graphs (DAGs), and counterfactual reasoning, CML enables researchers to identify not only which factors correlate with poor health outcomes, but which interventions could most effectively mitigate inequities. In the context of health informatics, CML frameworks integrate diverse datasets electronic health records, census data, geospatial indicators, and social vulnerability indices to infer how structural and behavioral factors jointly influence population health. By linking social determinants of health (SDOH) with outcomes such as chronic disease prevalence, hospitalization rates, and treatment adherence, these models can simulate hypothetical interventions, assess regional policy effects, and quantify potential causal impacts. Moreover, advances in explainable AI enhance transparency, allowing stakeholders to trace decision pathways and ensure equity-centered policy formulation. This paper conceptualizes an integrated causal health informatics framework for evaluating SDOH-driven disparities and intervention outcomes at regional and community levels. It outlines model design principles, data fusion methodologies, and ethical considerations essential for fair and reproducible analysis. Ultimately, CML enables a shift from descriptive analytics to actionable insight, supporting evidence-based strategies that directly target the root causes of health inequity and optimize intervention effectiveness.