Supply of interventional cardiologists and the provision of lower-value Percutaneous Coronary Interventions (PCI)

Background: Percutaneous coronary interventions (PCI) can be lifesaving for patients with acute coronary syndromes but of lower value for patients with stable coronary artery disease (CAD). Previous studies suggest that larger provider supply drives higher healthcare utilization. Whether this is true for lower-value PCI is unknown. Objective: To examine the association between the regional supply of PCI-performing physicians and PCI use and lower-value use. Design: Cross-sectional. Setting: 100% Medicare fee-for-service (FFS) claims and Medicare Advantage encounter data, linked to publicly available sources of population and hospital data. Participants: All Medicare FFS and MA beneficiaries undergoing PCI at US hospitals, 2019–2021. Measurements: Our exposure was the supply of PCI-performing physicians per million population in the hospital-referral-region (HRR). Our primary outcome was lower-value PCI provision defined as the share of all PCIs performed for stable CAD at the hospital-level, unadjusted and with multilevel linear regression adjustment for hospital and regional-population factors. We also examined both PCI use and lower-value use at the regional (HRR) level. Results: Our final dataset included 1,580 hospitals across 306 HRRs with 5,505 PCI-performing physicians. In our primary, hospital-level analysis, the lower-value PCI provision rate averaged 17.3% among hospitals located in HRRs in the lowest quintile of PCI provider density and 22.6% among hospitals in HRRs in the highest quintile. In our fully adjusted model, hospitals in the highest provider density quintile had a 5.77 percentage point higher lower-value provision rate than those in the lowest quintile (95% CI 3.70, 7.84; p