Millions of Medicare beneficiaries have new prescription drug benefits starting in January 2006, under Medicare Part D enacted by the Medicare Prescription Drug, Improvement, and Modernization Act. Nearly all Part D plans involve substantial and new types of patient cost-sharing, e.g. coverage gaps;however, little is known about potential adverse clinical consequences, or the financial consequences overall and specifically for patients. We propose to evaluate the effects of Part D-related cost-sharing on drug use, clinical events, and direct medical costs between 2005-2009, using a quasi-experimental pre-post design with concurrent controls, within Kaiser Permanente's Northern and Southern California (KPNC &KPSC) regions. Within this integrated delivery system, there is an ongoing natural experiment among 700,699 subjects age 65+ years old. Nearly half of these subjects (47%) face increases in their drug cost-sharing as part of KP's Medicare Advantage (MA) Part D plan in 2006. The remaining 53% of subjects, with employer-supplemented plans (controls), have lower drug cost-sharing levels with no significant changes from 2005-2006. The levels and types of cost-sharing within the KP Part D plans differ by region, permitting three main comparisons: 1) a basic Part D MA plan with high copayments and a coverage gap vs. employer-supplemented plans with bwer copayments and no coverage gap, both in KPNC;2) an enhanced Part D plan with high copayments and generic-only coverage instead of a gap vs. employer-supplemented plans, both in KPSC;and 3) basic vs. enhanced Part D MA plans, e.g. no coverage vs. generic-only coverage during the gap period. We will use repeated measures methods to test the hypotheses that 1) plans with higher levels of cost-sharing are associated with lower drug use and higher rates of adverse clinical outcomes (ED visits, hospitalizations, and deaths);and 2) higher cost-sharing plans are associated with lower annual pharmacy and total medical costs, but higher non-pharmacy and patient out-of-pocket costs. We will investigate these outcomes within the Overall Population, and within select Vulnerable Populations, e.g. patients with chronic diseases or low socioeconomic status. We will adjust for relevant factors, such as comorbidity levels and prior hospitalizations. This study will be the first to examine the clinical and economic effects of new Medicare Part D plans, using comprehensive automated data from before and after the introduction of Part D, for a well- defined population.
Showing the most recent 10 out of 14 publications
