JMIR Publications

ABSTRACT

Background:

Atrial fibrillation (AF) is a common cause of stroke and timely diagnosis is critical for secondary prevention. Little is known about smartwatches for AF detection among stroke survivors.

Objective:

We aimed to examine accuracy, usability, and adherence to a smartwatch-based AF monitoring system designed by older stroke survivors and their caregivers.

Methods:

Pulsewatch is a randomized-controlled trial in which stroke survivors received either a smartwatch-smartphone dyad for AF detection (Pulsewatch system) plus an ECG patch or the patch alone for 14 days to assess accuracy and usability of the system (Phase I). Participants were subsequently re-randomized to potentially 30 additional days of system use to examine adherence to watch wear (Phase II). Participants were at least 50-years-old, survived an ischemic stroke, and had no major contraindications to oral anticoagulants. Accuracy for AF detection was determined by comparing to cardiologist-overread ECG patch, and usability was assessed with the System Usability Scale. Adherence was operationalized as daily watch wear time over the 30-day monitoring period.

Results:

120 participants were enrolled (mean age 65, 41% female, 87% white). The Pulsewatch system demonstrated 92.9% accuracy (95% confidence interval: 85.3%–97.4%) for AF detection. Mean usability score was 65/100 and on average, participants wore the watch on 21.2 (SD 8.3) of 30 days.

Conclusions:

Our findings demonstrate that a smartwatch system designed by and for stroke survivors is a viable option for long term arrhythmia detection among older adults at risk for atrial fibrillation, though it may benefit from strategies to enhance adherence to watch wear. Clinical Trial: clinicaltrials.gov registration number NCT03761394 (https://www.clinicaltrials.gov/ct2/show/NCT03761394)