How Hospitals Respond to Incentives in Bundled Payment Models for Joint Surgery

The Issue

Episode-based payments, also known as bundled payments, are intended to incentivize hospitals and other providers to deliver more efficient care. However, some question whether hospitals in these models can take advantage of incentive structures in a way that undercuts the aim to deliver lower-cost care. One example is a hospital selecting low-risk patients for inpatient surgery when lower-cost outpatient surgery is just as appropriate.

Medicare’s Comprehensive Care for Joint Replacement (CJR) model offers bundled payments for hip and knee joint replacements. A Commonwealth Fund–supported study in JAMA Internal Medicine looked at six years of Medicare claims data for hospitals in CJR. Researchers estimated changes in episode savings as CJR transitioned from mandatory to voluntary in parts of the country and analyzed how responses by hospitals might explain changes in savings.

What the Study Found

• After changes to the CJR program — including making participation voluntary for some hospitals and introducing outpatient total knee replacements nationally that were not eligible for CJR bundles — participation was no longer associated with statistically significant savings by the fourth year. From the second to the fourth year of the program, savings in CJR versus control hospitals dropped from a significant savings of $976 to much smaller, nonsignificant savings of $331 per episode.
• In areas where CJR participation was mandatory, hospitals tended to treat healthier patients and were more likely to perform knee replacements in the inpatient setting (a procedure covered by CJR bundles) when outpatient surgery (a procedure not covered by CJR bundles) would be a less costly alternative. These hospital responses to program changes explained most of the drop in savings from the program’s second to fourth years for mandatory participants.
• In areas where CJR participation was voluntary, a greater percentage of hospitals with the highest-cost patients dropped out of the program, compared with the lowest-cost hospitals (93.3% in the highest quartile of patient cost vs. 51.8% in the lowest quartile). These costlier hospitals, which typically cared for the most complex patients, also tended to have more savings per episode in the early years of CJR. This study was among the first to suggest that dropout from the program was greater among hospitals that generated the most savings per episode and also the highest-cost hospitals, based on regional benchmarks.
• Using regional benchmarks, rather than benchmarks based on a hospital’s own historical spending, creates trade-offs, the authors said. Although regional benchmarks provide a greater incentive to reduce spending, they may promote selective participation in voluntary bundled payment models.

The Big Picture

Despite the intention of bundled payment programs like CJR, hospitals may be able to take advantage of the incentive structures through patient selection and choosing more costly sites of care. Particularly in CJR, inadequate risk adjustment for patient complexity and the exclusion of outpatient joint replacement reinforces this problem. To promote the success of bundled payments and deliver more efficient care, the authors suggest that new alternative payment models should be designed with caution.

The Bottom Line

Savings during the first two years of CJR disappeared by the fourth year because of hospitals’ responses to changes in the program. Specifically, hospitals were able to drop out of the program, or they were selective about choosing patients for joint replacement and elected to perform the procedures in inpatient settings. Policy options to promote future savings from bundled payments include making these programs mandatory and adjusting benchmarking methodology to reflect more robust risk adjustment.