Did Payment Reform Lead to Patient Selection in Hip and Knee Arthroplasties? An Observational Study Using New York State Data

Background: While the comprehensive care for joint replacement (CJR) bundled payment program for total joint replacement (TJR) emphasizes value, concerns persist regarding unintended consequences, primarily hospital selection of healthier, younger patients. Purpose: We sought to assess changes in patient characteristics and outcomes after CJR implementation in New York State. Methods: This retrospective cohort study included primary total hip and total knee arthroplasties from the New York Statewide Planning and Research Cooperative System (SPARCS) database. Procedures performed before (July 2014 to March 2016; n = 58,610) and after (April 2016 to December 2017; n = 78,728) CJR implementation were compared. Primary outcomes were patient characteristics: Deyo-Comorbidity Index and age. Secondary outcomes were increased hospitalization cost, discharge to institutional post-acute care, and prolonged length of stay. A difference-in-differences analysis estimated changes after CJR implementation, comparing CJR to non-CJR hospitals. Results: We found that CJR implementation (in 49 of 144 New York State hospitals) coincided with slightly older and more comorbid TJR recipients. The CJR program coincided with significantly reduced hospitalization cost and discharge to institutional post-acute care but not length of stay. Some CJR effects appear to have affected non-Medicare patients, as well. Conclusion: This retrospective analysis suggests that in New York State, the CJR bundled payment program did not result in hospitals selecting younger and healthier TJR recipients and coincided with decreased costs and fewer discharges to institutional postacute care.