Cost implications of patient spending on heart failure medications in the US Medicare program

Objectives The aim of this study was to model the associations between patient spending on heart failure (HF) medications and Medicare and all-payer expenditures on health care services for participants in the Medicare prescription drug (Part D) program. Methods Correlational analysis of pooled 2011-12 data from the Medicare Current Beneficiary Survey. Analysis was restricted to community-dwelling beneficiaries with self-reported HF at baseline, continuous Part D coverage, and no Low-Income Subsidy (LIS). The main predictor was mean patient expenditure on a HF-related prescription per 30-day supply. The outcomes were all-payer and Medicare-specific payments for inpatient and total health care services during the observation year. Key findings Mean patient drug expenditure was not statistically associated with Medicare or all-payer inpatient payments or (after covariate adjustment) with total health care payments. However, patient expenditure was statistically associated with total Medicare payments, e(gamma) = 1.022, 95% CI [1.004 to 1.041]. Marginal effects analysis predicted an average rise in total Medicare payments of $190.32, 95% CI [$40.54 to $341.10], for each additional $1 of patient spending per prescription, P = 0.013. Given an average 2.4 HF-indicated drug classes per participant and assuming 12.2 copays per year, a hypothetical $1 increase in prescription copay predicted a net loss to Medicare of $160.90 per participant. Conclusion Prescription drug spending by Medicare beneficiaries with HF was not associated with higher inpatient or all-payer costs. A modest association between patient drug spending and total Medicare costs was observed, but longitudinal and cost-effectiveness analyses are needed to support causal inference.