Two recent studies demonstrate that the number of elective procedures does not vary significantly in bundled payment programs when compared to controls. Wilson, de Brantes and Conway analyzed1 the Medicare Bundled Payments for Care Improvement initiative to determine if the volume of Lower Joint Replacement (LJR) procedures increased under the model. Their analysis was in response to an editorial3 in the Journal of the American Medical Association that raised questions about elective procedure volume and bundled payments in an accompanying JAMA study.2
The JAMA study demonstrated that mean Medicare episode payments declined $1166 more in LJR bundle cases than in matched controls with no diminution in quality. Of note to the editorialist, the JAMA study reported that the mean volume of LJR procedures per quarter in bundle hospitals increased from 61.5 to 64.6 from baseline to the study period while in non-bundle hospitals the mean volume went from 59.6 to 59.2. The editorial erroneously claimed that this represented a notable increase in the volume of procedures in the bundle group.
The crucial point is that the difference in volume change between the two groups was not statistically significant (difference-in-differences estimate, 3.38; 95% CI, −2.4 to 9.2*). Importantly, the Wilson study confirmed this finding.
Because the total number of LJRs is increasing across the U.S., the Wilson study looked at whether the increases are greater where the providers were participating in a bundled payment model. They employed a sophisticated analysis using Medicare claims data from 2010-2015, adjusting for trends in regional demographic and market characteristics and their effects on total numbers of LJR procedures, such as regional shifts in numbers of Medicare beneficiaries, hospital mergers, and opening of new joint replacement centers.
Their analysis demonstrated that the increase in the number of LJR procedures – which was not statistically significant according to both the JAMA study and the Wilson study – was largely attributable to a combination of 1) growth in the Medicare population or the number of hospitals in the region and 2) a concentration of surgical units due to mergers or other market events, such as the opening of new joint replacement centers.
In sum, two recent studies demonstrate no increase in volume of procedures in bundled payment programs compared to controls. While future studies should continue to closely watch procedural volume where bundles are adopted, all elective surgical programs, whether or not they participate in bundled payment, should have a rigorous approach to assessing the appropriateness of surgical procedures offered to patients.
* For non-statisticians, the CI, or confidence interval, means that the ‘difference in differences estimate’, which in this case was 3.38, could have been anywhere inside the high and low ends of the CI (−2.4 to 9.2) and it would not have reached statistical significance.
Dummit LA, Kahvecioglu D, Marrufo G, et al. Association between hospital participation in a Medicare bundled payment initiative and payments and quality outcomes for lower extremity joint replacement episodes. JAMA. 2016;316(12):1267-1278.
Fisher ES. Medicare’s Bundled Payment Program for Joint Replacement: Promise and Peril? JAMA. 2016;316(12):1262-1264.