Improving Patient Safety through Medical Alert Management: An Automated Decision Tool to Reduce Alert Fatigue.

Area of Science:

• Health Informatics
• Clinical Decision Support
• Patient Safety

Background:

• Electronic medical records (EMRs) incorporate drug safety alerts to minimize adverse drug events.
• Alert fatigue, characterized by user desensitization to frequent alerts, poses a significant risk to patient safety.
• Existing alert systems often generate a high volume of non-critical alerts, contributing to alert fatigue.

Purpose of the Study:

• To mitigate alert fatigue by developing a predictive method for filtering superfluous drug safety alerts.
• To improve the effectiveness of drug safety alerts in electronic medical records.

Main Methods:

• A predictive model was designed to identify and filter alerts based on historical override rates, override rate variance, and sample size.
• The method was retrospectively tested and validated using a dataset from a large pediatric network.
• Alert filtering accuracy was evaluated based on selected parameter values.

Main Results:

• The implemented alert filtering achieved 91-96% accuracy in identifying superfluous alerts, varying with parameter selection.
• The system successfully reduced the number of non-critical alerts presented to users.
• This filtering allowed for a more focused review of potentially critical drug safety alerts.

Conclusions:

• Predictive filtering of drug safety alerts is an effective strategy to combat alert fatigue in EMR systems.
• Reducing alert fatigue can enhance clinical workflow and allow healthcare providers to better respond to critical safety notifications.
• This approach holds the potential to decrease the incidence of adverse drug events and improve overall patient safety.