Paperdoi:10.1016/j.hlc.2026.02.016
Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias
Heart, Lung and Circulation (2026-06-26)
TL;DR
32 patients, 322 rhythms during electrophysiology studies; Apple Watch S4 and Fitbit Versa HR vs intracardiac electrograms / 12-lead ECG (accurate = within ±10%). Sinus: 100% vs 87% (p = .04); SVT: 89% vs 86%; SVT termination: 62% vs 41%; VT: 41% vs 16%; neither tracked HR changes in VF or asystole.
- Form factor: wrist (Apple Watch Series 4, Fitbit Versa)
Abstract
Background & aimSmartwatches can provide a non-invasive heart rate (HR) monitoring to aid screening for atrial fibrillation, however, the accuracy of HR detections in the context of other arrhythmias requires further investigation. We aimed to assess the accuracy of HR measurements of two commonly used wearables at onset, during and termination of various induced and paced arrhythmias.MethodPatients undergoing an electrophysiology study were fitted intraprocedurally with both the Apple Watch Series 4 (AW) and Fitbit Versa (FB). Detected HRs on the devices were recorded at initiation, during, and at termination of various paced and induced arrhythmias. True HR was confirmed via intracardiac electrograms and 12-lead electrocardiogram. Detected HRs on the AW and FB were assessed for accuracy, defined as HR with in ±10% of the true rate.ResultsA total 322 cardiac rhythms from 32 patients were analysed. Both AW and FB demonstrated reasonable HR detection during sinus rhythm (100% vs 87%; p=0.04) and supraventricular tachycardia (SVT) (89% vs 86%; p=1). Both devices performed poorly at detecting SVT termination (62% vs 41%; p=0.04) and ventricular tachycardia (41% vs 16%; p=0.007). Neither device identified changes in HR during ventricular fibrillation or asystole.ConclusionsBoth the Apple Watch and Fitbit demonstrate reasonable accuracy in HR detection during regular rhythms of supraventricular origin, however, struggle with arrhythmia terminations and ventricular arrhythmias. Care should be taken when assessing HR trends for presence of cardiac arrhythmias.
Verbatim from Europe PMC (opens in a new tab) · Licence: cc by
Results on ChatPPG leaderboards
All leaderboardsNo results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.
Tasks & signals
Tasks
Signals
Keywords
How it works
Cite this paper
Turnbull, S., Gupta, A., Garikapati, K., Bhaskaran, A., De Silva, K., Kotake, Y., Bennett, R. G., Campbell, T. G., Thomas, S. P., Chow, C. K., & Kumar, S. (2026). Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias. Heart, Lung and Circulation. https://doi.org/10.1016/j.hlc.2026.02.016
Turnbull, Samual, Anunay Gupta, Kartheek Garikapati, Ashwin Bhaskaran, K De Silva, Yasuhito Kotake, and Richard G. Bennett, et al. 2026. "Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias." Heart, Lung and Circulation. https://doi.org/10.1016/j.hlc.2026.02.016.
Turnbull, S., Gupta, A., Garikapati, K., Bhaskaran, A., De Silva, K., Kotake, Y. and Bennett, R. G. et al. (2026) 'Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias', Heart, Lung and Circulation. doi: 10.1016/j.hlc.2026.02.016.
Turnbull, Samual, et al. "Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias." Heart, Lung and Circulation, 2026. https://doi.org/10.1016/j.hlc.2026.02.016.
S. Turnbull et al., "Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias," Heart, Lung and Circulation, 2026, doi: 10.1016/j.hlc.2026.02.016.
@article{turnbull2026smartwatchbased,
title = {{Smartwatch-Based Heart Rate Detection in Atrial, Ventricular and Pulseless Arrhythmias}},
author = {Turnbull, Samual and Gupta, Anunay and Garikapati, Kartheek and Bhaskaran, Ashwin and De Silva, K. and Kotake, Yasuhito and Bennett, Richard G. and Campbell, Timothy G. and Thomas, Stuart P. and Chow, Clara K. and Kumar, Saurabh},
year = {2026},
journal = {Heart, Lung and Circulation},
doi = {10.1016/j.hlc.2026.02.016}
}Added to the ChatPPG research hub