Decreasing Medication Errors at Discharge: A Quality Improvement Project

Purpose/Objectives: Medication reconciliation is part of the National Patient Safety Goal and suboptimal medication reconciliation at discharge has been found to be associated with 28% of potentially avoidable readmissions within 30 days (Tong et al, 2017).
In our institution, PRN (as needed) medication errors made up 27% of the medication errors in After Visit Summaries (AVS).
The aim of our project was to increase the percentage of PRN medications that meet our medication reconciliation bundle criteria from 57% to 64% in the audited inpatient After Visit Summaries by the end of the academic year. We will increase this percentage by 3% every three months for the following year, with a goal to sustain 75% for another year.

Design/Methods: A multidisciplinary committee was created including resident physicians, nursing, information technology, pharmacy, and other departments. The IHI model of improvement was used in the evaluation and implementation of all interventions. Evaluation of the current system was done using a Pareto chart for medications errors, process mapping of the medication reconciliation process in the inpatient units, and an affinity diagram to summarize the gaps identified. A key driver diagram (figure 1) was created and Plan-Do-Study-Act (PDSA) cycles were used for implementation.
Interventions were targeted towards our top error-prone medications based on the Pareto chart. Interventions included implementation of a discharge order set for patients needing dual analgesics with prefilled instructions ensuring the 5 rights and alternating instructions, timely completion of admission medication reconciliation, and optimizing medication prescription instructions in the electronic medical record (EMR).
Measures: The outcome measure was the percentage of PRN medications meeting bundle criteria. Process measures included completion rate of admission medication reconciliation and percentage of accuracy for specific medications based on the intervention.
Data were collected on a monthly basis and displayed using control charts to assess trends.

Results: Baseline data on the accuracy of PRN medications over 19 months were reviewed and showed only 57% accuracy on the AVS. After multiple PDSA cycles over 5 months, we had a shift from a baseline of 57% to 75% (figure 2). The last 3 months showed an average accuracy of 80% which surpassed our goal of 64%.
The process measures showed that there was a reduction of 15 errors/50 audited charts to 0 errors for the analgesics. MiraLAX interventions also demonstrated similar improvement from 80% to 90% accuracy. We saw an increase in admission medication reconciliation rate from a baseline of 73% of patients to 96% by the end of the PDSA cycles.

Conclusion/Discussion: We were able to attain accurate discharge medication reconciliation rates for PRN medications beyond our predicted goal using QI methods and collaboration. Our next steps include sustaining current gains while implementing interventions on other PRN medications and outpatient scheduled medication.