Utilization of high-cost prostate cancer therapies by oncologists in the merit-based incentive payment system (MIPS).

Abstract

73 Background: TheMerit-Based Incentive Payment System (MIPS) is currently the only federally mandated value-based payment model for oncologists. MIPS uses cost and quality metrics to assign oncologists an overall MIPS score, where higher scores correspond to greater payments for practices. The extent to which MIPS delineates between oncologists administering high-cost anti-cancer therapeutics with unproven benefit is unknown. In this study, we investigated administration of low-value prostate cancer therapies by differently scored oncologists. Methods: We identified MIPS scores for oncologists using publicly available data for the year 2019. We cross referenced these scores with the 20% restricted sample of Medicare fee-for-service claims for 2019 using provider National Provider Identifications (NPIs). We then identified patients who were treated by these oncologists for prostate cancer using ICD-10 diagnosis codes for prostate cancer. Use of Sipuleucel-T (Provenge) and Radium Ra 223 dichloride (Xofigo), two low-value treatments with unproven benefit in prostate cancer, was identified using HCPCS codes in the claims file. The top 25 and bottom 25 percentile of oncologists based on their MIPS scores were identified using summary statistics. Odds ratios (OR) and corresponding confidence intervals (CI) were generated for Sipuleucel-T and Radium Ra 223 dichloride use by top 25 and bottom 25 percentile MIPS scores. Results: We identified 533,966 claims for prostate cancer treated by oncologists with a MIPS score in 2019. Of these claims, there were 100 instances of Sipuleucel-T use and 308 instances of Radium Ra 223 dichloride use (500 instances of Sipuleucel-T and 1540 instances of Radium Ra 223 dichloride if scaled to national use). Oncologists in the lowest 25% of MIPS scores less likely administered Sipuleucel-T (OR 0.25, 95% CI 0.12, 0.54) while those in the highest 25% did not show a significant trend (OR 1.36, 95% CI 0.90, 2.07). Oncologists in the lowest 25% of MIPS scores more likely administered Radium Ra 223 dichloride (OR 1.53, 95% CI 1.21, 1.95) while those in the highest 25% less likely administered the drug (OR 0.66, 95% CI 0.49, 0.88). Conclusions: Low MIPS scoring oncologists were less likely to utilize Sipuleucel-T and more likely to use Radium Ra 223 dichloride to treat prostate cancer. Our findings suggest that the MIPS program may be ineffective at measuring and incentivizing high value care among oncologists. Specialty-specific recalibration of cost and quality measures is needed to ensure that policy efforts to promote value-based care do not compromise healthcare quality and outcomes. [Table: see text]