Wearable devices for detection of cardiac arrest: a prospective proof-of-concept study

Background Out-of-hospital cardiac arrest remains a leading cause of mortality worldwide. Many events are unwitnessed, delaying cardiopulmonary resuscitation and defibrillation. Wearable devices that measure heart rate using photoplethysmography (PPG) could enable automated cardiac arrest detection, alerting nearby responders, thereby improving cardiac arrest outcomes. Objectives To provide a proof-of-concept evaluation of a smartwatch and paired smartphone application to detect rapid ventricular arrhythmias (ventricular tachycardia [VT] or ventricular fibrillation [VF]) or loss of pulse induced with programmed ventricular stimulation during electrophysiology studies. Methods Sustained ventricular arrhythmias were induced in patients under controlled conditions using a validated electrophysiology study protocol. Participants wore an Apple Watch SE paired with a smartphone application programmed to generate alerts for heart rates >210 bpm or <30 bpm. Alerts were compared with gold-standard invasive electrophysiology laboratory recordings. Results Sustained arrhythmias were induced in 2 of 3 patients, resulting in three events: one VT episode (3 s, 274 bpm) and two VF episodes (22 s, mean 333 bpm; 13 s, mean 341 bpm). No events were detected by the smartwatch (0/3 events detected). Conclusions In this proof-of-concept study, standard PPG-based heart rate measurements on a commercial smartwatch failed to detect induced ventricular arrhythmias. This likely reflects the dependence of PPG on pulsatile peripheral blood flow, which may be reduced or absent during ventricular tachyarrhythmias. PPG-based heart rate detection alone appears insufficient. Monitoring raw PPG waveforms for abrupt variation in the signal amplitude or incorporating a continuous electrocardiogram with a second electrode on another part of the body may enable reliable cardiac arrest detection and early intervention.