Quality of Treatment Selection for Medicare Beneficiaries With Cancer

Abstract

PURPOSE The Medicare part D Low-Income Subsidy (LIS) improves access to oral cancer drugs, but provides no assistance for clinician-administered/part B drugs. This analysis assessed the association between LIS participation and receipt of optimal cancer treatment. METHODS We investigated initial systemic therapy using SEER-Medicare data (2015-2017) and National Comprehensive Cancer Network (NCCN) Evidence Blocks (EB) as the standard for treatment recommendations. We included cancer clinical scenarios wherein (1) ≥one treatment was optimal (higher efficacy and safety scores) versus other treatments; (2) identifiable in SEER-Medicare (eg, not defined by clinical data unavailable in registry data or claims); and (3) both EB and ASCO Value Framework agreed regarding optimal treatment. We fit logistic regression models to assess the association between receipt of systemic therapy ( v no therapy) and patient and provider characteristics. Contingent on receipt of treatment, we modeled the likelihood of receiving a treatment ranked (by EB scores) within the highest or lowest quartile for that cancer type. RESULTS Nine thousand two hundred and ninety patients were included across 11 clinical scenarios. Fifty-seven percent (5,336) of patients received any systemic therapy and 43% (3,954) received no systemic therapy. Compared with non-LIS participants, LIS participants were less likely to receive any systemic therapy versus no systemic therapy (odds ratio, 0.64 [95% CI, 0.57 to 0.72]). Contingent on receiving systemic therapy, LIS participants received treatment ranked within the worst quartile 24.8% of the time, compared with 21.9% of non-LIS patients (adjusted prevalence difference, 4.3% [95% CI, 0.5 to 8.2]). CONCLUSION LIS participants were less likely to receive systemic therapy at all and were more likely to receive treatments that receive low NCCN EB scores.