Quantification of Heterogeneous Semi-Structured Patient-Reported Physical Activities Derived from a Diabetes...

Area of Science:

• Digital Health
• Exercise Physiology
• Health Informatics

Background:

• Telehealth systems can promote physical activity, but quantifying patient activity levels is challenging.
• Accurate assessment of physical activity is crucial for monitoring patient status and progress.
• The DiabMemory telehealth system collects semi-structured activity data (type, intensity, duration).

Purpose of the Study:

• To develop a method for mapping semi-structured activity data from the DiabMemory system to Metabolic Equivalents of Task (METs).
• To quantify physical activity levels using MET-minutes for telehealth patients.
• To evaluate the distribution of activity types and their associated MET scores.

Main Methods:

• Analyzed activity data from 947 telehealth patients.
• Created a mapping table linking activity type-intensity pairs to MET values using a preexisting compendium.
• Calculated physical activity in MET-minutes by combining MET scores with activity duration.

Main Results:

• Successfully mapped semi-structured activity data to MET values.
• Quantified physical activity using MET-minutes.
• Found a significant difference in MET-minutes across different activity types (p < 0.0001).

Conclusions:

• The developed method provides a quantifiable measure of physical activity for telehealth users.
• This approach enables better evaluation of physical activity changes and telehealth intervention effectiveness.
• Future work will utilize this mapping to assess DiabMemory's impact on patient activity levels.