Personalized Healthcare Outcome Analysis of Cardiovascular Surgical Procedures.

This article focuses on evaluating hospital performance heterogeneity across patient groups to improve personalized healthcare outcomes and pay-for-performance programs. Using patient-level data from 35 New York State hospitals for six cardiovascular surgeries, the study applies an extended multiple-treatment instrumental variable tree (MT-IVT) approach to identify patient subgroups with significantly different hospital outcomes, accounting for endogeneity in hospital choice. Results show that outcome differences between hospitals vary not only by procedure type but also by patient age and comorbidities, with about 80% of patients having different best-quality hospitals when assessed by patient-centric rather than population-average information. The study further demonstrates that patient-centric information can reduce complication rates, guide patients to more appropriate hospitals, and enable payers like CMS to better align financial incentives with hospital performance, thereby encouraging targeted quality improvements and strategic focus among hospitals.