DETECT-1C: CardioWatch detects non-induced cardiac arrest with 100% sensitivity
A prospective clinical validation study published in The Lancet Regional Health – Europe evaluated automated cardiac arrest detection using photoplethysmography signals from the Corsano CardioWatch 287-2.
The study included 44 intensive care patients following the clinical decision to withdraw life-sustaining treatment because of a poor prognosis. Patients wore the CardioWatch while continuous invasive blood pressure and ECG provided the reference measurements. All cardiac arrests were non-shockable, including asystole and pulseless electrical activity—the rhythms responsible for most out-of-hospital cardiac arrests.
The algorithm was refined in two training cohorts and then evaluated in an independent test cohort of 23 patients. In this test cohort, it achieved:
-
100% sensitivity, detecting all 23 cardiac arrests
-
96% positive predictive value
-
Only one false-positive alert during 172.5 hours of monitoring
This is the first prospective study to validate automated wrist-based detection in true, non-induced cardiac arrest and specifically in non-shockable rhythms. Previous wearable studies had focused mainly on simulated or medically induced cardiac arrest.
Corsano Health has been a long-term technology and research partner in the DETECT programme for the past five years. Corsano provided the CardioWatch platform and contributed PPG data infrastructure, software, algorithm development and study methodology. Corsano’s Kambiz Ebrahimkheil and Peter Stas are co-authors of the publication.
The results represent an important step towards a wearable system that could automatically recognize cardiac arrest and activate the emergency response chain. Further studies are required to validate sensitivity and false-positive rates during daily life and across more diverse patient populations.
Reference: Edgar R, Jansen CE, Pol LR, et al. Automated cardiac arrest detection using wrist-derived photoplethysmography during withdrawal of life-sustaining treatment: a prospective clinical validation study. The Lancet Regional Health – Europe. 2026;67:101791. https://doi.org/10.1016/j.lanepe.2026.101791
