A Metrics Taxonomy and Reporting Strategy for Rule-Based Alerts

Area of Science:

• Health Informatics
• Clinical Decision Support Systems
• Healthcare Management

Background:

• Electronic health records generate numerous rule-based alerts.
• Effective alert management is crucial for patient safety and operational efficiency.
• Existing alert taxonomies may not fully capture action-oriented aspects.

Purpose of the Study:

• To develop an action-oriented taxonomy for rule-based alerts.
• To categorize alerts based on structure, actions, and intended process outcomes.
• To establish a framework for analyzing and optimizing alert systems.

Main Methods:

• Development of a taxonomy based on 333 rule-based alerts at Kaiser Permanente Northwest (KPNW).
• Identification of 9 major and 17 overall alert classes.
• Creation of a specific metric approach for 5 alert classes, including the three most frequent.

Main Results:

• A comprehensive action-oriented alerts taxonomy was successfully developed.
• The 3 most numerous alert classes in KPNW constituted 67% (224/333) of all alerts.
• A metric approach was defined for key alert categories, facilitating targeted analysis.

Conclusions:

• The developed taxonomy provides a structured approach to understanding and managing clinical alerts.
• This classification aids in identifying high-impact alerts and optimizing their use.
• The findings support the refinement of alert systems for improved healthcare delivery.