Medication reconciliation by clinical pharmacists in an outpatient family medicine clinic

Objectives: To evaluate the incidence of medication discrepancies in electronic health record (EHR) medication lists in an outpatient family medicine clinic where clinical pharmacists perform medication reconciliation, to classify and resolve the discrepancies, to identify the most common medication classes involved, and to assess the clinical importance of the discrepancies. Methods: This research was conducted at Bethesda Family Medicine Clinic in St. Paul, MN, with data collected from February 2009 to February 2010. To be included, patients had to be 18 years or older and have at least 10 medications listed in the EHR. The clinical pharmacist saw each patient before the physician, reviewed the medication list with the patient, and made corrections to the EHR medication list. When possible, comprehensive medication management (CMM) also was conducted. Results: During 1 year, 327 patients were seen for medication reconciliation. A total of 2,167 discrepancies were identified and resolved, with a mean (±SD) of 6.6 ± 4.5 total discrepancies and 3.4 ± 3.2 clinically important discrepancies per patient. The range of total discrepancies per patient was 0 to 26. The most common discrepancy category was "patient not taking medication on list" (54.1%). Overall, the source of the discrepancy usually was the patient, but it varied according to discrepancy category. The most common medication classes involved were pain medications, gastrointestinal medications, and topical medications. Of the 2,167 discrepancies, 51.1% were determined to be clinically important by the pharmacist. The pharmacist conducted CMM in 48% of patients. Conclusion: Outpatient medication reconciliation by a pharmacist identified and resolved a large number of medication discrepancies and improved the accuracy of EHR medication lists. Because more than 50% of the discrepancies were thought to be clinically important, improving the accuracy of medication lists could affect patient care.