The effects of cost sharing in the Medicare Part D benefit on branded and generic oral anti-diabetic adherence
Sacks, Naomi Cohen
MetadataShow full item record
This dissertation includes three papers on oral anti-diabetic (OAD) medication adherence and the Medicare Part D benefit. The data source for all analyses is the IMS Health Longitudinal Prescription database, which contains patient de-identified prescription data from computerized pharmacy records and reflects over 60% of the prescriptions filled in the United States. The analytic samples culled from this database consist of elderly diabetes patients with Medicare Part D standard (non-LIS), low-income-subsidy (LIS), or commercial prescription medication coverage. I compare non-LIS patients, whose coverage contains a gap (also known as the "donut hole"), to patients with LIS or commercial coverage and no gap. The first paper tests the hypothesis that diabetes patients with non-LIS coverage are selective in their use of oral anti-diabetic (OAD) medications, and less likely to be adherent to branded-only class OADs. I use logistic regression to model the effects of this coverage on within-class OAD adherence, comparing non-LIS to propensity-matched LIS and commercially covered patients. I find evidence that the Part D benefit is associated with cost-related non-adherence in branded-only, but not primarily generic, class OADs. In the second paper, I examine the effects of increasing generic medication use on concurrent adherence evaluations. I use the same logistic model to compare three recognized but different concurrent adherence outcome measures. The measures generate estimates of the effects of Part D coverage on adherence that differ in significance and magnitude and that vary further, depending on the distribution of prices in the underlying patient sample. In the third paper, I test the hypothesis that patients decrease more costly branded OAD use in anticipation of the Part D coverage gap. Using a difference-in-difference design, I use GEE via proc genmod to estimate the effects of the gap on within-class OAD adherence over the coverage year. Compared to LIS and commercially covered, non-LIS patients have increasingly lower odds of adherence to branded-only class OADs starting in the first month of the coverage year. These findings provide evidence of forward-looking behavior, and suggest further that deductibles and co-payments, Part D cost sharing components in addition to the gap, adversely affect adherence.
Thesis (Ph.D.)--Boston University PLEASE NOTE: Boston University Libraries did not receive an Authorization To Manage form for this thesis or dissertation. It is therefore not openly accessible, though it may be available by request. If you are the author or principal advisor of this work and would like to request open access for it, please contact us at firstname.lastname@example.org. Thank you.