Comparing the Effectiveness of Alerts and Dynamically Annotated Visualizations (DAVs) in Improving Clinical Decision...

Area of Science:

• Clinical Informatics
• Human-Computer Interaction
• Medical Decision Support Systems

Background:

• Electronic health record alerts often have high false-alarm rates, leading to disregard.
• Visualizations and improved guidance understandability can enhance decision-making.
• Existing decision support systems face challenges with interruptions and alert triggers.

Purpose of the Study:

• To compare the effectiveness of dynamically annotated visualizations (DAVs) versus interrupting pop-up alerts.
• To assess the impact of real-time decision support on reducing inappropriate diagnostic imaging orders.
• To evaluate physician preferences and perceptions of understandability, transparency, and relevance.

Main Methods:

• A between-subject design was used in a simulated environment with 11 patients.
• Physicians were randomly assigned to receive either alerts or DAVs as decision support.
• Secondary measures included self-reported understandability, algorithm transparency, and clinical relevance.

Main Results:

• Fewer inappropriate diagnostic imaging tests were ordered with DAVs (18%) compared to alerts (34%).
• The DAV system was rated significantly higher for understandability, transparency, and relevance.
• Physician preference favored DAVs across all tested patient scenarios.

Conclusions:

• Dynamically annotated visualizations are more effective than traditional alerts in reducing inappropriate imaging orders.
• DAVs offer a superior user experience, particularly in complex or ambiguous clinical scenarios.
• This visualization approach may improve decision support in settings with high false-alarm rates or where interruptions are detrimental.