Background: Prior studies of the Medicare Part D coverage gap are limited in generalizability and scope.
Objective: To determine the effect of the coverage gap on drugs used for asymptomatic (antihypertensive and lipid-lowering drugs) and symptomatic (pain relievers, acid suppressants, and antidepressants) conditions in elderly patients with hypertension and hyperlipidemia.
Design: Quasi-experimental study using pre-post design and contemporaneous control group.
Setting: Medicare claims files from 2005 and 2006 for 5% random sample of Medicare beneficiaries.
Patients: Part D plan enrollees with hypertension or hyperlipidemia aged 65 years or older who had no coverage, generic-only coverage, or both brand-name and generic coverage during the gap in 2006. Patients who were fully eligible for the low-income subsidy served as the control group.
Measurements: Monthly 30-day supply prescriptions available, medication adherence, and continuous medication gaps of 30 days or more for antihypertensive or lipid-lowering drugs; monthly 30-day supply prescriptions available for pain relievers, acid suppressants, or antidepressants before and after coverage gap entry.
Results: Patients with no gap coverage had a decrease in monthly antihypertensive and lipid-lowering drug prescriptions during the coverage gap. Nonadherence also increased in this group (antihypertensives: odds ratio [OR], 1.60 [95% CI, 1.50 to 1.71]; lipid-lowering drugs: OR, 1.59 [CI, 1.50 to 1.68]). The proportion of patients with no gap coverage who had continuous medication gaps in lipid-lowering medication use and antihypertensive use increased by an absolute 7.3% (OR, 1.38 [CI, 1.29 to 1.46]) and 3.2% (OR, 1.35 [CI, 1.25 to 1.45]), respectively, because of the coverage gap. Decreases in use were smaller for pain relievers and antidepressants and larger for acid suppressants in patients with no gap coverage. Patients with generic-only coverage had decreased use of cardiovascular medications but no change in use of drugs for symptomatic conditions. No measures changed in the brand-name and generic coverage groups. Results of sensitivity analyses were consistent with the main findings.
Limitation: Because this study was nonrandomized, unobserved differences may still exist between study groups.
Conclusion: The Part D coverage gap was associated with decreased use of medications for hypertension and hyperlipidemia in patients with no gap coverage and generic-only gap coverage. The proposed phasing out of the gap by 2020 will benefit such patients; however, use of low-value medications may also increase.
Primary funding source: Penn-Pfizer Alliance and American Heart Association.