Patient Selection After Mandatory Bundled Payments for Hip and Knee Replacement

Background: On April 1, 2016, the Centers for Medicare & Medicaid Services (CMS) introduced bundled-payment programs for hip replacement and knee replacement (HKR) in selected metropolitan statistical areas (MSAs) to decrease the costs and cost variability of HKR and to increase the quality of care. Early program analyses showed cost savings; however, studies also demonstrated a trend toward the selection of healthier patients for HKR performed under the bundled system. We compared the characteristics of patients who underwent HKR before implementation of the bundled-payment system (pre-policy) with those of patients who underwent HKR after implementation (post-policy).

Methods: Patients who underwent HKR from 2015 to 2016 were identified from Medicare inpatient claims files. After matching for MSA characteristics, we used a difference-in-difference design to evaluate changes in patient case mix from pre-policy to post-policy by comparing Medicare beneficiaries receiving HKR in bundled MSAs (bMSAs) with those receiving HKR in non-bundled MSAs (nbMSAs). The main characteristics of interest were race, dual eligibility (for Medicare and Medicaid), tobacco use, obesity, presence of diabetes with or without complications, and Charlson Comorbidity Index (CCI) value. We also evaluated pre-policy to post-policy changes in patient case mix by comparing Medicare beneficiaries in bMSAs who underwent HKR compared with those who underwent hip hemiarthroplasty. Hip hemiarthroplasty was used as a control to determine whether there were changes in access to HKR.

Results: We found significant differences in the unadjusted baseline characteristics between the bMSA and nbMSA cohorts, both for unmatched and matched samples. We found no significant post-policy changes in the characteristics of patients undergoing HKR. Patients undergoing hemiarthroplasty had significantly higher CCI values than did those undergoing HKR in bMSAs post-policy, although the difference was small (0.36-point higher CCI value; p < 0.01). Patients undergoing hemiarthroplasty were also 2.4% more likely to have diabetes mellitus without complications compared with those who underwent HRK post-policy (p < 0.01).

Conclusions: In contrast to previous investigators, we found little to no significant change in the characteristics (including race, dual eligibility, tobacco use, obesity, presence of diabetes with or without complications, and CCI value) of Medicare beneficiaries who underwent HKR after the initiation of the CMS mandatory bundled-payment policy.