Accuracy of Apple Watch Heart Rate Measurements in Patients with Obstructive Sleep Apnea and Atrial Fibrillation: A Prospective Observational Study (Preprint)

BACKGROUND Patients with obstructive sleep apnea (OSA) are at higher risk for atrial fibrillation (AF). Consumer wearable heart rate (HR) sensors may be a means for passive HR monitoring in patients with AF. OBJECTIVE To assess the Apple Watch’s agreement with telemetry in measuring HR in patients with OSA in AF. METHODS Patients with OSA in AF were prospectively recruited prior to cardioversion/ablation procedures. HR was sampled every 10 seconds for 60 seconds using telemetry and an Apple Watch concomitantly. Agreement between devices was assessed using modified Bland-Altman analysis accounting for repeated measures per patient, mixed-effects regression modeling with the patient as a random effect while controlling for rapid ventricular response, and Lin's concordance correlation coefficient. RESULTS 20 patients (mean 66 +/- 6.5 years, 85% male, 3 (15%) with RVR) participated in this study, yielding 134 HR observations per device. The Apple Watch had 95% of differences fall within 18.5 beats per minute (bpm) above and 18.3 bpm below telemetry measurements, with a mean bias of 0.26 bpm. After controlling for RVR, 95% of differences fell between -14.3 and 14.5 bpm with a mean bias of 0.23 bpm. Lin’s concordance correlation coefficient was 0.88 (95% CI 0.85-0.91), suggesting acceptable agreement between the Apple Watch and telemetry. CONCLUSIONS In OSA patients in AF, the Apple Watch provided acceptable agreement with HR measurements by telemetry. Further studies with larger sample populations and wider range of HR are needed to confirm these findings.