Hospital participation in Medicare ACOs: no change in admission practices and spending

Abstract

Objectives: Hospital participation in accountable care organizations (ACOs)-Medicare's signature alternative payment model-continues to grow despite mixed evidence on spending and quality. This study examines whether hospital ACO participation is associated with changes in emergency department (ED) admission practices, hospital length of stay (LOS), and spending for unplanned admissions.

Study design: A difference-in-differences analysis of Medicare fee-for-service ED visits and hospitalizations (2008-2019).

Methods: Medicare claims were linked to ACO tracking data from Torch Insight to identify hospitals that joined an ACO between 2012 and 2017 (6 cohorts, followed for a maximum of 5 years), the start date of their initial contract, and ACO characteristics. Key outcomes included ED admission and observation stay rates, hospital LOS for emergent admissions, and total costs for an index ED event.

Results: Among the 995 hospitals (27.6% of the short-term hospitals in our study) that joined a Medicare ACO during the study period, program participation up to 5 years was not associated with changes in the rate of hospitalization from the ED, hospital LOS, or total costs of the index event. Findings remained consistent across ACO program, contract risk levels, year of program entry, and overall ACO performance (eg, whether the ACO generated shared savings).

Conclusions: Hospitals did not significantly alter care delivery for unplanned hospitalizations after joining an ACO. These findings suggest that hospital-led ACOs may have limited impact on reducing costs for emergent admissions, raising concerns about their ability to drive meaningful care transformation.