Abstract Body (Do not enter title and authors here): Objective: U.S. healthcare spending is rising, driven in large part by spending on cardiovascular disease. In response, the Centers for Medicare & Medicaid Services (CMS) is implementing payment models to incentivize more efficient care. The Transforming Episode Accountability Model (TEAM) is a mandatory bundled payment model launching in 2026. Hospitals will be accountable for spending on episodes of care, spanning admission to 30 days after discharge, for 5 surgeries, including coronary artery bypass grafts (CABGs). If they meet spending targets, they will receive a financial bonus from CMS; if not, they must pay a penalty. Little is known about the hospitals mandated to participate in TEAM. Understanding how participants differ from nonparticipants is important as CMS scales insights from TEAM to the broader cardiovascular payment landscape.

Design: We identified TEAM participation from CMS and hospital characteristics from the American Hospital Association survey, Area Health Resources File, and Medicare enrollment and claims data. t-tests compared characteristics.

Findings: We identified 716 hospitals in TEAM. Compared to nonparticipants, participants were larger (263 vs 214 beds), more often urban (100.0% vs 90.6%), and part of the safety net (14.4% vs 8.3%). Participants served a higher share of patients dually enrolled in Medicare and Medicaid. 30-day episode spending ranged from $29,778 for lower extremity joint replacements to $50,092 for CABGs, but for all bundles, spending was higher among participants than nonparticipants. The contribution of post-acute care (PAC) spending to episode spending varied from $5,705 (11.4% of 30-day spending) for CABG to $21,556 (50.1% of 30-day spending) for hip/femur fracture bundles.

Conclusions: Hospitals mandated to participate in TEAM were larger, more urban, cared for a higher share of marginalized patients, and were more often part of the safety net. This raises concerns that such hospitals, which have fared poorly in prior payment models, could bear disproportionate penalties under TEAM. In addition, our results suggest that spending reductions might be heterogeneous across bundles. Bundled payment participants generally reduce overall episode spending by reducing PAC spending. Greater spending reductions might thus be expected in procedures for which we found PAC to represent a greater portion of episode spending, such as the hip/femur fracture bundle, as opposed to the CABG bundle.