Hype or Hope? Selection and Performance of Accountable Care Organizations

Abstract

The performance of pay-for-performance (P4P) models in healthcare has been mixed. While prior studies have explored various population-based payment models, there is a lack of empirical evidence related to selection of high risk, high-reward payment models in the context of value-based healthcare. Further, the performance implications of selection into new types of payment models is not well understood. We study the rollout of Accountable Care Organizations (ACO) under the Medicare Shared Savings Program (MSSP) and identify factors that explain their selection into two-sided risk models, which offer greater rewards as well as penalties. Specifically, we study whether such ACO selection decisions are associated with performance improvements based on their shared cost savings as well as quality of health outcomes. Our longitudinal analysis is based on publicly available Medicare ACO data for the six-year period between 2013 and 2018, and explores the antecedents and consequences of ACO selection into two-sided risk models. We find that ACOs with greater organizational scope, based on their scale, service variety and patient segments, are also more likely to switch to a two-sided risk model. Further, we observe that ACOs that switched into two-sided models exhibit greater savings and marginally higher quality, compared to ACOs that remained in a one-sided risk model. However, our analysis indicates that the initial gains after switching are not sustained over time, as these ACOs exhibit significant reduction in the rate of improvement of shared savings and quality, in the three-year period after switching. Our results indicate that ACOs with superior prior capabilities reap the advantages associated with MSSP incentives, and imply that incentive programs that promote short-term goals help participants who enjoyed greater a priori advantages in terms of their extant resources and capabilities. However, long-term sustainable performance improvements remain an elusive goal, and our research suggests that the incentives in the current ACO program need to be modified to reward improvements in their operational capabilities.