Volume 52 | Number 3 | June 2017

Abstract List

Bruce Stuart Ph.D., Franklin B. Hendrick Ph.D., Jing Xu M.A., J. Samantha Dougherty Ph.D.


To determine the magnitude and mechanisms of response to Medicare Part D cost sharing by low‐income subsidy () recipients using oral hypoglycemic agents (s) and statins.

Data Sources

Medicare data for a 5 percent random sample of beneficiaries with diabetes enrolled in fee‐for‐service Part D drug plans in 2008.

Study Design

We evaluated the impact of differences between generic and brand cost sharing rates among cohorts of and non‐ recipients to determine if wider price spreads increased the generic dispensing rate () and reduced total drug use and cost.

Principal Findings

We found little association between cost sharing and aggregate and statin use. In adjusted analyses, non‐ beneficiaries who paid 46 percent of total costs had 2.5 percent fewer days supply than full benefit dual eligibles who paid just 5 percent of their therapy costs. For statins, the difference in days supply between those facing the lowest and highest cost sharing was 4.6 percent. Higher cost sharing was associated with filling fewer but larger prescriptions for both generics and brands.


Higher generic and brand copays had little association with and statin use among recipients. This implies that modest changes in required cost sharing for these medicines would have very little substantive impact on generic dispensing or utilization patterns among recipients and thus would have little effect on total program spending. At the same time, any increases in out‐of‐pocket costs would be expected to shift costs and place greater financial burden on low‐income beneficiaries, particularly those in poor health.