Predicting neurally mediated syncope based on pulse arrival time: algorithm development and preliminary results

Neurally mediated syncope (NMS) is a common disorder that is triggered by orthostatic stress. The circulatory adjustments to orthostatic stress occur just prior to a sudden loss of consciousness. NMS prediction would protect patients from falls or accidents.

Based on simultaneously recorded heart rate (HR) and pulse wave during 70° head-up tilt (HUT) table testing we investigated a syncope warning system. In 14 patients with a history of suspected NMS we tested 2 algorithms based on HR and/or pulse arrival time (PAT). When the cumulative risk exceeded the threshold, which was calculated during the first 2 minutes following the posture change to upright position, a syncope prediction alarm was triggered. All syncopes (n = 7) were detected more than 16 seconds before the onset of dizziness or unconsciousness by using a prediction alarm based on HR and PAT (syncope prediction algorithm 2). No false alarm was generated in patients with negative HUT (n = 7). Syncope prediction was improved by detecting the slope of HR changes as compared with monitoring PAT changes alone (syncope prediction algorithm 1). The duration between the prediction alarm and the occurrence of syncope was 99 ± 108 seconds.

Predicting NMS is feasible by monitoring HR and the onset of the pulse wave at the periphery. This approach might improve NMS management.