Percutaneous Coronary Intervention Mortality, Cost, Complications, and Disparities after Radiation Therapy: Artificial Intelligence-Augmented, Cost Effectiveness, and Computational Ethical Analysis.

Abstract

The optimal cardio-oncology management of radiation therapy and its complications are unknown despite the high patient and societal costs. This study is the first known nationally representative, multi-year, artificial intelligence and propensity score-augmented causal clinical inference and computational ethical and policy analysis of percutaneous coronary intervention (PCI) mortality, cost, and disparities including by primary malignancy following radiation therapy. Bayesian Machine learning-augmented Propensity Score translational (BAM-PS) statistics were conducted in the 2016-2020 National Inpatient Sample. Of the 148,755,036 adult hospitalizations, 2,229,285 (1.50%) had a history of radiation therapy, of whom, 67,450 (3.00%) had an inpatient AMI, and of whom, 18,400 (28.69%) underwent PCI. Post-AMI mortality, costs, and complications were comparable with and without radiation across cancers in general and across the 30 primary malignancies tested, except for breast cancer, in which PCI significantly increased mortality (OR 3.70, 95%CI 1.10-12.43, p = 0.035). In addition to significant sex, race, and insurance disparities, significant regional disparities were associated with nearly 50 extra inpatient deaths and over USD 500 million lost. This large clinical, cost, and pluralistic ethical analysis suggests PCI when clinically indicated should be provided to patients regardless of sex, race, insurance, or region to generate significant improvements in population health, cost savings, and social equity.