Toward alert triage: scalable qualitative coding framework for analyzing alert notes from the Telehealth Intervention...

Area of Science:

• Digital Health
• Health Informatics
• Clinical Data Analysis

Background:

• Telehealth generates vast amounts of data, risking clinician burnout and missed critical patient information.
• Efficient data triage is essential for successful telehealth program implementation and patient safety.

Purpose of the Study:

• To develop an automated classification framework for nurses' notes to create a training dataset for an alert triage system.
• To enhance the efficiency and accuracy of data management in telehealth programs.

Main Methods:

• A coding framework and regular expression-based keyword matching were developed using 24,931 alert notes.
• The automated alert triaging model was evaluated for scalability and performance on 800 sampled alert notes.

Main Results:

• 92% of alert notes (22,717/24,579) were successfully classified into 17 codes.
• The automated analysis achieved an average precision of 0.86 and recall of 0.90.

Conclusions:

• The developed approach is feasible and scalable for community-based telehealth, particularly for older adults.
• Coded alert notes can improve predictive modeling and false alert triage, advancing remote health monitoring systems.