Automated clinical decision support alerts in electronic health record (EHR) systems are sometimes perceived by busy members of the clinical team as unhelpful or an outright nuisance, and to the extent that such perceptions become more common and so-called “alert fatigue” sets in among users, everyone loses.
In 2020 at Vanderbilt University Medical Center, the Vanderbilt Clinical Informatics Center (VCLIC) in coordination with HealthIT organized a grassroots program to optimize clinical decision support alerts and help stem alert fatigue at VUMC. In eStar, VUMC’s clinical IT system, clinical decision support alerts and reminders are called best practice advisories, or BPAs, and the program to rid the system of suboptimal alerts was called BPA Clickbusters.
According to a new report in the Journal of the American Medical Informatics Association, during two three-month rounds of clickbusting in 2020, comprehensive reviews of 20% of the rule-based alerts in use at VUMC led to 71,227 fewer interruptive alerts being sent per week, for a total reduction in interruptive alerts of 15%.
Some alerts were eliminated outright, while others were altered in ways that affect when they do and don’t fire, or for which users. Some alerts were determined to already be optimal and were left unchanged. Recommended changes were adjudicated by the program sponsors or forwarded to relevant clinical governance committees for a decision. Helping to engage people’s sporting instincts, Clickbusters was gamified, with scoring (and prizes) based on a volunteer’s overall impact on the system.
“Beyond helping improve the alerts in our system, Clickbusters had the important added benefit of involving more eStar users in the evaluation and improvement of electronic clinical decision support at VUMC,” said Allison McCoy, PhD, assistant professor of Biomedical Informatics and scientific director of VCLIC’s Clinical Informatics Core. McCoy authored the report with VCLIC’s director, Adam Wright, PhD, professor of Biomedical Informatics, and 31 of their colleagues from across VUMC.
Clickbusters participants, who included physicians, nurses and pharmacists, signed up to review a given BPA in detail and potentially discontinue or improve it. Participants followed a 10-step process that started with the clinical logic that gave rise to the advisory and how often the advisory was disregarded in practice; in many cases, the process concluded with testing, placing into production and evaluating a new and improved BPA.
In all, 84 BPAs were reviewed, 10 were turned off, 42 were modified and 32 were left as is.
According to the report, the process used by Clickbusters “could be readily replicated at other clinical sites and applied to other functions of the EHR, such as order sets, clinical documentation tools and information displays.”
Meanwhile, the report notes, the clinical informatics team at VUMC reviews all BPAs in eStar at least once every two years.
