1165 Improving the Accuracy of Health Maintenance Reminders in the Electronic Medical Record to Promote Adherence to Colonoscopy Surveillance Guidelines

Abstract

INTRODUCTION: Health Maintenance (HM) reminders in the Epic electronic medical record (EMR) offer an opportunity to increase adherence to recommended colonoscopy surveillance intervals. However, HM reminders must be updated by providers with specific knowledge of surveillance guidelines, and providers must be comfortable using the HM functionality. This quality improvement initiative aimed to improve the rate at which gastroenterologists accurately update HM colonoscopy surveillance recommendations. METHODS: 1,602 colonoscopies performed by 7 gastroenterologists at a single community hospital were tracked prospectively between April 2018–February 2019. For each colonoscopy, the gastroenterologist’s post-procedure surveillance recommendation was compared to the HM surveillance recommendation in the patient’s Epic EMR. Accuracy was defined as concordance between these two recommendations. After obtaining baseline data, group and individual performance rates were disseminated and the gastroenterologists underwent an educational intervention to improve familiarity with the HM functionality. Monthly HM accuracy rates were tracked for the subsequent 6 months. Accuracy rates in the post-intervention month and the last month of follow-up were compared to baseline using the student’s T-test. RESULTS: The baseline HM accuracy rate for the group was 69% overall and 56% for high-risk patients (defined as a recommended surveillance interval of ≤5 years). For the month immediately following the intervention, the overall accuracy rate rose significantly compared to baseline (86% vs 69% overall, P ≤ 0.001; 80% vs 56% for high-risk patients, P ≤ 0.001). However, for the last month of follow-up the overall accuracy rate had decreased back to baseline (70% vs 69%, P = 0.646). The high-risk accuracy rate in the last month remained significantly higher than baseline but had also declined (65% vs 56%, P = .008). Notably, 82% of discordant HM reminders recommended a longer surveillance interval than was clinically indicated. CONCLUSION: Sharing performance data and provider education can significantly improve the rate at which gastroenterologists accurately update colonoscopy HM surveillance intervals in the EMR. However, a single intervention is not sufficient to sustain this improvement over time. Repeated intervention is needed to ensure that HM surveillance intervals remain accurate, particularly since the majority of discordant HM reminders recommend a longer surveillance interval than is clinically indicated.