Intervention to Improve Residents\u27 Medication Reconciliation Accuracy at Hospitalization Discharge

Background: Medication reconciliation is an effective tool to reduce discharge medication errors, which are a major patient safety concern. At the Washington DC VA medical center, medication reconciliation is performed largely by internal medicine residents. Previously, education on medication reconciliation has been successfully provided to residents at our institution. However recently, the curriculum, in addition to formal training, has largely been abandoned. Objective: We aimed to improve the accuracy of medication reconciliation at discharge via an educational intervention for internal medicine residents. Methods: The intervention was run from August 2017 to December 2017 and delivered to internal medicine residents from four academic institutions rotating at the Washington DC VA medical center. The educational intervention featured a pocket card outlining the discharge medication reconciliation process, a discharge instruction guide in the orientation booklet, and once weekly or biweekly educational seminars delivered at resident conference. The seminars were led by a chief resident or a faculty member and focused on teaching both quality improvement principles and our institution-specific medication reconciliation process. An effort was made for the seminars to be interactive including time for residents to practice medication reconciliation or assess their own performance using a self-evaluation rubric. The intervention was refined over the course of three PDSA cycles. The accuracy of 50 post-intervention discharge medication lists was compared to that of 50 pre-intervention lists. Accuracy was assessed using a rubric to review the medication list on the patient’s discharge instructions compared to the data in the progress note on the discharge date. Results: The patients included in the pre-intervention chart review were discharged on average with 11 medications while the patients in the post-intervention group received 10. Between the two groups, the number of duplicate medications (26% vs. 9%), extraneous medications (20% vs. 14%), medications sorted by disease or indication (64% vs. 72%), and necessary medications that were omitted (26% vs. 14%). All improved via the intervention. Conclusion: Our educational intervention targeting internal medicine residents is effective in improving the accuracy of medication reconciliation at hospital discharge. However, the challenge remains in making this intervention multidisciplinary, time-effective, and sustainable. We hope that the success of this QI project will help promote future institutional support and stakeholder participation in continuous efforts to ensure patient safety via accurate medication reconciliation