Real-World Accuracy of Wearable Activity Trackers for Detecting Medical Conditions: Systematic Review and Meta-Analysis.
Level 1 - systematic review of randomized trials
Systematic review and meta-analysis of diagnostic accuracy studies.
PubMed 39213525 · doi:10.2196/56972
What was done
Authors searched 10 electronic databases from inception through April 1, 2023, for studies evaluating wearable activity trackers (such as smartwatches and fitness bands) used to diagnose or detect medical conditions or events in free-living adult populations. Meta-analyses pooled diagnostic metrics including area under the curve (AUC), accuracy, sensitivity, specificity, and positive predictive value (PPV). Risk of bias was evaluated using the Joanna Briggs Institute Critical Appraisal Checklist for Diagnostic Test Accuracy Studies.
What was found
The review included 28 studies comprising 1,226,801 participants across an age range of 28.6 to 78.3 years. Conditions evaluated were COVID-19 (16 studies), atrial fibrillation (5 studies), arrhythmia or abnormal pulse (3 studies), falls (3 studies), and viral symptoms (1 study). For COVID-19 detection, pooled AUC was 80.2% (95% CI 71.0%-89.3%), accuracy was 87.5% (95% CI 81.6%-93.5%), sensitivity was 79.5% (95% CI 67.7%-91.3%), and specificity was 76.8% (95% CI 69.4%-84.1%). For atrial fibrillation, pooled PPV was 87.4% (95% CI 75.7%-99.1%), sensitivity was 94.2% (95% CI 88.7%-99.7%), and specificity was 95.3% (95% CI 91.8%-98.8%). For fall detection, pooled sensitivity was 81.9% (95% CI 75.1%-88.1%) and specificity was 62.5% (95% CI 14.4%-100%).
Why it matters
Commercial wearables demonstrate high diagnostic accuracy for atrial fibrillation and moderate discrimination for COVID-19 in free-living conditions. This highlights their promise as scalable, noninvasive screening tools, though they require further refinement before serving as standalone diagnostic instruments.
Limits
The majority of evidence is concentrated on COVID-19 (57%) and atrial fibrillation (18%), leaving other conditions underrepresented. Estimates for fall detection specificity showed extreme imprecision (95% CI 14.4%-100%). Variations in device brands, proprietary sensor algorithms, and study settings introduce notable heterogeneity.
Cited by
- supports Wearable-based algorithms tracking parameters like respiratory rate and body temperature can detect COVID-19 infection before symptoms appear.