To examine factors associated with oversupply and undersupply of antihypertensive medication, and examine evidence for medication acquisition as distinct from self-reported adherence.
Analysis of pharmacy refill records, medical charts, and in-person interviews.
Five hundred sixty-two male veterans with hypertension enrolled in a randomized controlled trial to improve BP control.
Patients were classified as having undersupply (<0.80), appropriate supply (≥0.80 and ≤1.20), or oversupply (>1.20) of antihypertensive medication in the 90 days before trial enrollment based on the ReComp algorithm. Determination of BP control was based on clinic measurements at enrollment. Demographic, clinical, psychosocial, and behavioral factors relevant to medication-taking behavior and BP were assessed at enrollment.
Twenty-three percent of the patients had undersupply, 47% had appropriate supply, and 30% had oversupply of antihypertensive medication. Multinomial logistic regression revealed that using fewer classes of antihypertensive medications and greater perceived adherence barriers were independently associated with greater likelihood of undersupply. Current employment was associated with decreased likelihood of oversupply, and greater comorbidity and being married were associated with increased likelihood of oversupply. Agreement between ReComp and self-reported adherence was poor (κ = 0.19, P < 0.001). Undersupply, oversupply, and self-reported nonadherence were all independently associated with decreased likelihood of BP control after adjusting for each other and patient factors.
Antihypertensive oversupply was common and may arise from different circumstances than undersupply. Measures of medication acquisition and self-reported adherence appear to provide distinct, complementary information about patients' medication-taking behavior.
From the *Department of Population Health Sciences, University of Wisconsin, Madison, Wisconsin; †Health Services Research and Development Northwest Center of Excellence, Veterans Affairs Puget Sound Health Care System, Seattle, Washington; ‡Department of Medicine, University of Washington, Seattle, Washington; §Center for Health Services Research in Primary Care, Veterans Affairs Medical Center, Durham NC; and Departments of ∥Medicine and **Psychiatry and Behavioral Sciences and ††Center for Aging and Human Development, Duke University, Durham, North Carolina.
Supported by the Department of Veterans Affairs, Veterans Health Administration, HSR&D Service, investigator initiative grant 20-034 (to H.B.) and an Established-Investigator award from the American Heart Association (to H.B.). Dr. Thorpe was supported by a post-doctoral fellowship from the Department of Veterans Affairs, Office of Academic Affairs. Dr. Bryson is supported by a VA Career Development Award (RCD 03-177).
This work was also presented as a poster at the 2008 AcademyHealth Annual Research Meeting, June 8–10, 2008, Washington, DC.
The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
Reprints: Carolyn Thorpe, PhD, MPH, Health Innovation Program, Department of Population Health Sciences, University of Wisconsin, E5/724 CSC, 600 Highland Ave, Madison, WI 53792-7685. E-mail: firstname.lastname@example.org.