The spillover effects of Medicare’s comprehensive care for joint replacement (CJR) model in California

TL;DR

This study examines how a Medicare joint replacement payment model in California affected older patients beyond those directly covered by Medicare.

Contribution

The study reveals spillover effects of the CJR model on non-Medicare patients in California.

Findings

Patients in treated hospitals had shorter hospital stays and higher home discharge rates.

Spillover effects were observed for non-Medicare patients in treated hospitals.

Changes in outcomes were significant for Medicare and non-Medicare patients alike.

Abstract

The Comprehensive Care for Joint Replacement (CJR) model, a bundled Medicare payment for lower extremity joint replacement (LEJR), was initially randomized across the United States, providing a unique opportunity to study the broad impact of this alternative payment model. This study aimed to determine the spillover effects of the CJR model on older patients in California covered outside of the traditional Medicare program. The study analyzed hospitalizations for hip and knee joint replacement in California between January 2014 and December 2017 from the California Patient Discharge Dataset. The study used event study and difference-in-differences models to estimate changes in discharge-related outcomes in hospitals in treated and control areas before versus after CJR implementation (April 2016). Main outcomes were hospital length of stay and home discharge rates. All LEJR patients admitted to the treated or control hospitals were included in the study regardless of their primary payers. Of 312,914 analyzed LEJR hospitalizations (mean [SD] age, 68.3 [11.3] years; 189,575 [60.6%] women; 15,374 [4.9%] black), 113,590 (36.3%) were covered by traditional Medicare (TM), 83,277 (26.6%) were covered by Medicare Advantage (MA), and 116,047 (37.1%) were without Medicare coverage. After program implementation, TM and non-Medicare patients in treated hospitals experienced reductions in length of stay (-4.0% & -1.0%, p < 0.05) and TM, MA and non-Medicare patients in treated hospitals experienced increases in home discharge rates (3.4%, 4.7% & 2.3%, p < 0.001) relative to patients in untreated hospitals. CJR affected health care for non-targeted populations. Evaluating the program based on traditional Medicare beneficiaries alone does not capture the entire effect of the program on older adults.

Figures