Causal decomposition of summary measures of population health

Abstract

Summary measures of population health facilitate the identification of public health priorities by quantifying the relative impact of diseases, injuries, and risk factors. Causal attribution within these measures primarily utilizes two traditions: categorical attribution and counterfactual analysis. Categorical attribution assigns health events to a single underlying cause based on mutually exclusive rules, such as those defined in the International Classification of Diseases. This method enables additive decomposition, which is essential for certain policy applications, yet it often fails to account for multicausality. In contrast, counterfactual analysis evaluates health outcomes by comparing current population health to hypothetical scenarios, such as a theoretical minimum risk distribution. This framework is vital for assessing individual and environmental risk factors and accommodates the interacting components of the causal web. It further distinguishes between attributable burden from past exposures and avoidable burden that may be prevented by current interventions. While counterfactual methods lack the inherent additivity of categorical approaches, they offer a more theoretically robust estimation of the impact of various health determinants. Effective population health assessment requires integrating both methods to capture the structural determinants of health loss while maintaining the practical utility of additive decomposition in health gap measures. – AI-generated abstract.