(Management Case Study) Implementation of a Medication Alert Optimization Committee at a Large Health System

Purpose: In the organization’s previous electronic health record (EHR), prescribers and pharmacists received differing alerts, with pharmacists receiving more alerts than prescribers. Additionally, the previous EHR did not contain the capability to review alert data and alert acceptance rates. There was no formal process to curate alert types and evaluate their usage. While preparing to transition to a new EHR, it was determined that a process was needed to evaluate medication-related alerts. This management case study describes the process to develop a medication alert optimization committee at a large health system with over 15 hospitals and 44 outpatient health centers.

Methods: In the previous EHR, there were four types of medication alerts: drug-drug interactions, duplicate therapy, drug-disease interactions, and customized medication alerts. Drug-drug interactions in the previous EHR fired for a warning severity of contraindicated and included a total of 9,794 warnings. Duplicate therapy warnings were turned on for exact drug duplicates but did not fire within a medication class, such as opioids. Drug-disease warnings did not fire. Customized medication alerts were built on an as-needed basis. With this variation, an EHR medication alert optimization committee (EMAC) was developed to support the new EHR transition. The committee consisted of subject matter experts from medication safety, pharmacy informatics, physician informatics, pharmacy operations, and clinical pharmacy specialists (CPS). The committee began meeting in March 2022 on a monthly cadence and reports to a newly developed clinical decision support (CDS) structure at the health system. The primary goal of establishing EMAC was to review and make recommendations for medication-related alerts including drug-drug interactions, drug-disease precautions, dose warnings, duplicate therapy warnings, and customized medication alerts. The alert review also included which alert severity types would display and to whom. Finally, all custom medication alert requests required certain criteria to be met. All medication-related alert review and recommendations were then presented to the system CDS committee for final approval and subsequent build into the new EHR.

Results: In order to review the relevant medication alerts, EMAC took a systematic approach to review by alert type. For drug-drug interactions, the recommendation was to fire active alerts if the alert carried a warning severity of severe or contraindicated, leading to a total of 4,322 drug-drug interactions and a decrease of more than 50% from the previous EHR. For duplicate therapy alerts, EMAC reviewed the health system’s high alert medication list and recommended only firing duplicate therapy for select high alert medications, leading to 14 medication classes that would actively fire a duplicate therapy alert. Drug-disease precautions were reviewed in detail by a CPS, who recommended modification of 318 warnings in order to reduce nuisance alerts. Pediatric dose-range checking was also customized by a CPS review. Finally, EMAC reviewed settings for pharmacists and prescribers and recommended that all caregivers see the same alerts. All of the medication-alert recommendations made by EMAC were accepted at the CDS committee. Ongoing monitoring of medication alerts were continued through EMAC post EHR implementation, including review of alerts per order and alert acceptance rate. Alert data will be utilized to reduce alert fatigue and nuisance alerts.

Conclusion: The implementation of EMAC indicated a successful, systematic review of medication alerts prior to EHR go-live. The implementation also demonstrated a greater than 50% reduction in drug-drug interaction alerts, thereby decreasing the risk of alert fatigue and nuisance alerts. Key factors for the success of EMAC included ensuring a multidisciplinary approach with subject matter experts from all relevant areas, establishing a reporting structure for approval and implementation of the alert recommendations, and involving system IT analysts for informed decision making. Future directions will include ongoing alert monitoring and data review for continuous alert optimization.