Personalized Health Care Outcome Analysis of Cardiovascular Surgical Procedures

Abstract

This study addresses the challenges of generating patient-centric outcome information. Using patient-level data from thirty-five hospitals for six cardiovascular surgeries in New York State, we identify patient groups that exhibit significant differences in outcomes with a recently developed instrumental variable tree approach. We find that outcome differences between hospitals are heterogenous not only across procedure types, but also along other dimensions such as patient age and comorbidities. For around 80% of patients, the best hospitals indicated by patient-centric information are different than those indicated as best by population-average information. We compare potential outcomes when patients are treated at the best hospitals based on the two types of information, and estimate that complications could be reduced by 66.7% by using patient-centric information instead of population-average information. We also use our model to illustrate how patient-centric outcome information can enhance pay-for-performance programs offered by payers and guide providers in targeting quality improvement efforts.

Keywords: Healthcare outcome analysis, patient-centric healthcare, provider quality, machine learning