Real-World Accuracy of Wearable Activity Trackers for Detecting Medical Conditions: Systematic Review and...

Area of Science:

• Digital Health
• Biomedical Engineering
• Medical Diagnostics

Background:

• Wearable activity trackers (e.g., fitness bands, smartwatches) monitor physiological data, offering potential for disease detection.
• The diagnostic accuracy of these devices for specific medical conditions remains under investigation.

Purpose of the Study:

• To systematically review and perform a meta-analysis on the use of wearable activity trackers for disease and medical event detection.
• To evaluate the diagnostic performance metrics including accuracy, sensitivity, and specificity.

Main Methods:

• Searched 10 electronic databases for studies up to April 1, 2023, involving adults using wearables for free-living condition diagnostics.
• Conducted meta-analyses to assess diagnostic accuracy, area under the curve, sensitivity, specificity, and positive predictive value.
• Performed subgroup analyses based on device type (e.g., Fitbit, Oura ring) and assessed risk of bias.

Main Results:

• Included 28 studies with over 1.2 million participants, focusing on COVID-19 (57%), atrial fibrillation (18%), arrhythmia (11%), and falls (11%).
• For COVID-19 detection, pooled accuracy was 87.5% (AUC 80.2%).
• For atrial fibrillation detection, pooled sensitivity was 94.2% and specificity was 95.3%. Fall detection showed 81.9% sensitivity and 62.5% specificity.

Conclusions:

• Wearable activity trackers demonstrate significant potential for disease detection, particularly for atrial fibrillation and COVID-19.
• While encouraging, further advancements are necessary to enhance the precision and clinical applicability of these wearable diagnostic tools.