Meta-analysis of interventions for medication adherence to antihypertensives

To identify methods targeted at improving adherence to antihypertensives and determine their effect on adherence using meta-analytic techniques.

A literature search from 1970 to December 2000 using MEDLINE, International Pharmaceutical Abstracts, PsychLit, ERIC, and EMBASE was performed using the terms compliance, adherence, and medication. Randomized articles with an intervention directed at a patient/caregiver, a comparator group, and a minimum of 10 subjects in each intervention group were identified by 3 independent reviewers. Articles that did not report sample size data or adequate results of the intervention were excluded. Sixteen citations focusing on antihypertensive adherence were identified. Of the 16 citations, 6 studied either more than one intervention in the same population or different interventions in different patient populations, yielding 24 cohorts with 2446 patients.

Fifty-eight percent of the methods focused on behavioral interventions (BIs), 29% studied the effect of a combination of behavioral and educational interventions (BEIs), and 13% utilized educational interventions (EIs) alone. Overall, the study groups were nonhomogenous (Q = 183.92; p < 0.001). However, when the groups were separated by the intervention type, the BIs were homogenous (Q = 1.19; p = 1.00) with an overall effect size (ES) of 0.04 (95% CI -0.01 to -0.09), indicating a trend toward improved adherence. Fifty percent of the BIs were performed in the physician's office; however, setting did not influence the intervention's impact (p = 0.13). Within the BIs, no single intervention improved adherence over others.

Based on the interventions included in this meta-analysis, there is no single intervention that improves adherence to antihypertensives over others; therefore, a patient-specific approach should be modeled.