OBJECTIVES: Nonadherence is a major reason for morbidity and mortality. The purpose of this study was to determine medication adherence of patients post myocardial infarction (MI). METHODS: Patients admitted to the University of Michigan Medical Center with diagnosis of MI were identified consecutively and prospectively from July 1999 to July 2000. Clinical data were obtained retrospectively from medical records including medications at discharge. Patients were interviewed by telephone six months after discharge. Follow-up data included a self-reported list of medications and a four-item adherence questionnaire. RESULTS: Complete information was obtained from 216 patients. The mean age was 63.4 (+12.6) years and 67.6% were male. Based upon the questionnaire, 36.1% of patients reported nonadherence. Lower adherence rate (49.1%) was obtained when comparing medications prescribed at discharge to medications documented at follow-up. Based on the questionnaire, the more medication prescribed, the greater the adherence (p=0.02). Trend tests of adherence over increasing number of medications were positive in all medication classes (beta-blockers, ACE inhibitors, anticoagulants, and lipid lowering drugs). Multivariate models show the similar trend when adjusted for age and gender. Adherence also increased with more documented comorbidities (p=0.01). When comparing medications prescribed to medications at follow-up there were no significant predictors of adherence. CONCLUSIONS: Nonadherence is frequent in patients after an MI (36%). Utilizing a medication adherence questionnaire to assess adherence is better than merely comparing medication lists at two different points in time. Better adherence to all 4 evidence-based therapies after an MI is seen in patients prescribed more medications and with more disease states. Higher adherence may be related to a recent major health event, more recent education on treatment benefits, and/or to a focus on rehabilitation. This study demonstrates the need for a larger, broader study that includes health beliefs, psychosocial assessment, and other patient factors that may influence compliance.