Health system planners are increasingly interested in using population-level data to inform system planning. Precision public health offers to improve the health systemâs ability to predict and prevent public health risks by developing targeted public health interventions aimed at specific, high-risk sub-populations. In Canada, premature mortality, deaths before the age of 75, has remained stagnant. But gaps in premature mortality have widened across sex, socioeconomic status, and geography, indicating that health systems are not functioning equitably. We propose developing a population-based risk prediction tool to predict the five-year incidence of all-cause premature mortality in Canadian cities using machine learning methods. By incorporating data on social and environmental determinants of health at the individual and area-levels, we aim to capture unfairness from the various structural and social factors that shape individual health outcomes. Health policymakers will use this prediction tool to create population risk segments, identify high-risk sub-populations, and inform policy and social and environmental interventions that can minimize health care expenditures and ensure healthy living conditions for all.
Read full abstract