Causal Inference with Artificial Intelligence
image credit to this websiteAbstract
Cardiovascular infarctions, such as myocardial infarction (MI), and cerebrovascular infarctions, such as ischemic stroke, are leading causes of death and disability in Canada. Both conditions result from tissue damage caused by a blocked blood supply to the heart or brain. Self-rated health (SRH), an individual’s perception of their own health status, is predictive of quality of life, disease recurrence, and mortality. Stroke and MI can impact SRH through various pathways, while psychosocial factors may buffer the challenges posed by such serious health events, leading to varying levels of SRH. In this project, we investigate the causal impact of stroke and MI on SRH, aiming to inform interventions that help maintain good subjective health despite serious medical events, particularly in the context of an aging population. Using propensity weighting and matching with parametric and machine learning methods, we demonstrate that stroke and MI negatively affect individuals’ subjective ratings of having very good or excellent health.
