Use of low-value cancer treatments in Medicare Advantage versus traditional Medicare.

Abstract

11075 Background: Medicare Advantage (MA) has steadily grown during the past decade, covering over half of Medicare beneficiaries in 2024. Because MA plans receive capitated payments for Medicare beneficiaries, they have the financial incentives to reduce service utilization and costs. MA plans may discourage enrollees from using low-value services, such as medically unnecessary care and expensive treatments for which low-cost alternatives are available. No prior study examining the use of low-value services between MA and TM has evaluated the use of low-value cancer treatments specifically. We compared use of low-value cancer treatments between MA and TM. Methods: Using national Medicare data, we performed retrospective analyses of beneficiaries who had a new cancer diagnosis between 2016 and 2021, and who were at risk of receiving one of the following low-value treatments: growth factors (GCSF) for patients receiving low-risk chemotherapy; denosumab for castration sensitive prostate cancer (CSPC); nab-paclitaxel instead of paclitaxel for breast or lung cancers; adding bevacizumab to carboplatin + paclitaxel for ovarian cancer; and branded drugs or biologics for which generic or biosimilar versions existed. We estimated linear regression models for each cohort/outcome separately. The key explanatory variable was MA enrollment (vs. TM). We used inverse probability of treatment weights to balance characteristics between MA and TM: patient age, sex, race, Medicare and Medicaid dual-eligibility status, cancer metastasis, a frailty index, and summary health-risk scores, area-level socio-economic and health care environment variables. We included year indicators to adjust for temporal trends and oncology practice indicators to control for practice-specific prescribing patterns. Results: Receipt of any low-value cancer treatment was 1.7 percentage points lower in MA than in TM (34.2% versus 35.9%; p<0.001). MA enrollees had lower utilization than TM for GCSF (7.3% versus 8.9%; p<0.001), denosumab (26.4% versus 33.1%; p<0.001), nab-paclitaxel (7.9% versus 8.7%; p<0.05), addition of bevacizumab for ovarian cancer (8.3% versus 10.5%, p=0.001), and biologics with biosimilar alternatives (66.8% vs. 68.5%; p<0.001). Receipt of branded drugs did not significantly differ between MA and TM. Conclusions: For Medicare beneficiaries at risk of receiving a low-value cancer treatment, MA enrollees were less likely to receive low-value cancer treatments than TM beneficiaries. These reductions in low value services may prevent avoidable toxicity and lower treatment cost to the patient and health care system. Increased efforts are needed to identify approaches that MA plans use to reduce low-value cancer treatments, and explore ways to promote those approaches in both MA and TM.