Patterns and predictors of medication discrepancies in primary care.

RATIONALE, AIMS AND OBJECTIVES Identifying medication discrepancies across transitions of care is a common patient safety problem. Research examining relations between medication discrepancies and adherence, however, is limited. The objective of this investigation is to explore the relations between adherence and patient-provider medication discrepancies, and to test the hypothesis that non-adherence would be associated with medication discrepancies. METHODS Three hundred twenty-eight outpatients completed a current medication list and measures of health literacy, adherence, perceived physical functioning and subjective well-being. Patient lists were compared with active medications in the electronic medical record. Multivariate analyses identified demographic, clinical and patient-reported variables associated with discrepancies involving prescribed daily medications. RESULTS Despite high rates of self-reported adherence, patients reported taking fewer medications than the number of active medications in their medical record (3.79 vs. 4.83, P < 0.001). We identified one or more discrepancies in most records (294/328 or 89.6%). Identified discrepancies were completely reconciled in only 21.1% of patients with discrepancies. Discrepancies were associated with lower health literacy, poorer physical health status and subjective well-being, and poorer adherence to the regimen patients believed they had been prescribed. Multivariate analysis indicated that the number of medical record-reported medications and subjective well-being independently predicted the presence of discrepancies. CONCLUSIONS Findings suggest a complex relation between treatment adherence and medication discrepancies in which patient well-being and regimen complexity work in tandem to create discordance between patient and provider medication plans. Simplifying regimens when possible and attending to patient life satisfaction may improve adherence to a regimen constructed jointly between patient and provider.