Research reveals smartwatches can really detect heart disease and Samsung proves the most accurate
Study confirms smartwatches detect heart disease with high accuracy, but experts suggest using caution
The Thaiger key takeaways
- A meta-analysis of 26 studies involving 17,349 participants found smartwatches can accurately detect atrial fibrillation (AF) with up to 97% accuracy.
- Samsung and Amazfit ranked highest for diagnostic precision, slightly ahead of Apple and Withings.
- Despite impressive results, researchers warn of false positives, especially among healthy users, and stress the need for professional medical follow-up.
Twenty-six studies have confirmed that smartwatches are highly accurate in detecting irregular heartbeats and atrial fibrillation (AF), with Samsung and Amazfit ranking the most reliable. However, researchers warn users to remain cautious due to the risk of false positives.
Atrial fibrillation (AF) is the most common type of cardiac arrhythmia and a major risk factor for stroke and heart disease. Early screening and diagnosis are therefore extremely important.
Today, smartwatches have become easily accessible to the general public. Several brands have even received approval from the US Food and Drug Administration (FDA) to use AF detection features. Still, many consumers remain unsure about how accurate these wrist-worn technologies truly are.
To answer this question, researchers conducted a systematic literature review and meta-analysis, compiling data from 26 related studies involving a total of 17,349 participants. The aim was to evaluate the overall accuracy of smartwatches in detecting AF. The research was published in this journal from the JACC
The findings showed outstanding accuracy in AF detection
Two key statistics stand out: Sensitivity was 95% (average 94.81%), meaning that if a person truly has AF, the smartwatch has a 95% chance of detecting it. Specificity was 97% (average 96.12%), meaning that if a person does not have AF, the device has a 97% chance of correctly identifying them as normal. The Area Under the Curve (AUC), a measure of overall performance, was 0.97 (the closer to 1.0, the better).
Which technology performs better? (ECG vs PPG)
Modern smartwatches use two main methods for detecting AF:
-
ECG (Electrocardiogram): measures electrical signals directly from the heart (e.g., Apple Watch, Samsung, Withings).
-
PPG (Photoplethysmography): uses green light sensors to track blood flow at the wrist (used in many brands).
Interestingly, the analysis found both technologies deliver similar diagnostic accuracy:
-
PPG-based: AUC 0.98
-
ECG-based: AUC 0.97
Which smartwatch brand is the most accurate?
Among brands with sufficient data, researchers found:
-
Samsung – highest diagnostic accuracy (AUC 0.98)
-
Amazfit – very close behind (AUC 0.98)
-
Apple Watch – AUC 0.97
-
Withings – AUC 0.97
However, when researchers adjusted results to account for sample size effects through meta-regression, the differences between brands were found to be not statistically significant.
Important cautions for users
- False positives in real-world settings
The studies had an average AF prevalence of 11%, much higher than in the general population. In low-risk groups—such as younger, healthy smartwatch users—the chance of false positives increases significantly. - Consequences of false positives
False alerts can cause anxiety, lead to unnecessary medical tests, or place an extra burden on healthcare services. - Effect of sample size on accuracy
Larger studies tended to report better accuracy, meaning variations might result more from study design than the devices themselves. - Short-term AF episodes
Smartwatches may detect brief, subclinical AF episodes, but the medical community is still uncertain whether these short occurrences significantly increase stroke risk or require anticoagulant treatment. - Recording quality
Some participants were excluded from studies because of poor recording quality, suggesting that movement or improper device use may still affect accuracy in real life. - Publication bias
Statistical analysis indicated possible publication bias, meaning studies showing poor accuracy might not have been published, slightly inflating the overall positive picture.
Latest Thailand News
Follow The Thaiger on Google News:
