Study Shows Hospitals Are Not Cherry-Picking the Healthiest Medicare Patients for TJA

    When a bundled payment model for total joint arthroplasty procedures in Medicare patients was introduced by the Centers for Medicare and Medicaid Services in 2016, there was a fear that providers would “cherry-pick” healthier patients at lower risk of complications to maximize their reimbursement.

    It didn’t happen: A study from The Johns Hopkins University shows no significant decrease in access to total hip arthroplasty (THA) and total knee arthroplasty (TKA) for sicker patients, racial/ethnic minorities, or others at a higher risk of complication or death during the first year of the Comprehensive Care for Joint Replacement (CJR) program.

    The CJR program mandates bundled payments for elective THA and TKA procedures in Medicare patients. Rolled out in 67 randomly selected metropolitan statistical areas, the program involves the use of predetermined pricing that coveres physician and hospital fees and all related care, from hospital admission through 90 days after surgery.

    RELATED: CMS Proposes Changes to the Comprehensive Care for Joint Replacement Model

    The program was designed to decrease costs and cost variability while increasing the quality of care for THA and TKA. However, some physicians expressed concern that bundled payments might lead hospitals to preferentially select healthier patients at lower risk of complications or death. Similarly, hospitals might avoid patients with comorbidities or other characteristics associated with increased risk of complications or death, including black race and lower socioeconomic status.

    Using a sample of Medicare claims from 2015 to 2016, the researchers evaluated possible changes in patient characteristics after the start of the bundled payment program. The study included a matched set of more than 12,000 episodes of THA and TKA and follow-up care for patients in areas where bundled payments were introduced and 20,000 episodes in other areas.

    After adjustment for age and sex, there were no significant differences in the characteristics of patients undergoing THA and TKA in bundled payment areas compared with areas with no change in reimbursement. Further analysis controlled for important risk factors for complications after THA and TKA, including general health, smoking, or diabetes, and showed no change in the proportions of patient who were black or low-income (based on Medicaid eligibility).

    For comparison, the researchers examined the characteristics of patients undergoing hip hemiarthroplasty. The results showed small but significant increases in the rate of comorbidity after the introduction of bundled payments. “Although small, these changes suggest that some surgeons may prefer hemiarthroplasty rather than total hip replacement in less-healthy patients to avoid treating such patients under a bundled-payment program,” the researchers write. However, there was no change in the numbers of black or low-income patients selected for hemiarthroplasty.

    The study helps to address concerns that the bundled payment program might lead hospitals to select healthier patients for THA and TKA while denying access for those in need. “[W]e did not find that bundled payments were associated with decreased access to [THA and TKA] for patients who had more comorbidities, were members of a minority racial group, or were of lower socioeconomic status,” the researchers write. They note some limitations of their study, including reliance on claims data and a relatively short follow-up period.

    Within those limitations, the study provides reassurance that the CCJR program has not led to cherry-picking or other major changes in patient selection based on risk criteria. The researchers call for further studies to allay concerns that bundled payments might reduce access to THA and TKA for higher-risk patients.


    Humbyrd CJ, Wu SS, Trujillo AJ, Socal MP, Anderson GF. Patient Selection After Mandatory Bundled Payments for Hip and Knee Replacement: Limited Evidence of Lemon-Dropping or Cherry-Picking. J Bone Joint Surg Am. 2020 Feb 19;102(4):325-331. doi: 10.2106/JBJS.19.00756.