JMIR Publications

ABSTRACT

Background:

Atrial fibrillation (AF) is the most common recurrent arrhythmia in clinical practice, with most clinical events occurring outside the hospital. Low detection and non-adherence to guidelines are the primary obstacles to AF management. Photoplethysmography (PPG) is a novel technology developed for AF screening. However, there has been limited validation of PPG-based smart devices for the detection of AF and/or the underlying clinical factors impacting detection.

Objective:

To explore the feasibility of PPG-based smart devices for the detection of AF in real-world settings.

Methods:

Subjects were recruited from September 14, 2018 to October 16, 2018,who aged 18 years or older (n = 361), were recruited to screen for AF with active measurement, which initiated by the users, from PPG-based smart wearable devices (i.e., a smart band or smart watches; PRO AF PPG algorithm developed by Huawei Device Co. Ltd). Of these, 200 subjects were also automatically and periodically monitored for 14 days with a smart band. The baseline diagnosis of “suspected” AF was confirmed by electrocardiogram (ECG) and physical examination. The sensitivity and accuracy of PPG-based smart devices for monitoring AF were evaluated.

Results:

A total of 2,353 active measurement signals and 23,864 periodic measurement signals were recorded. 11 subjects were confirmed with persistent AF, and 20 were confirmed with paroxysmal AF. Smart devices demonstrated a greater than 91% predictive ability for AF. The sensitivity and specificity of devices in detecting AF among active recording of the 361 subjects were 100% and above 99%, respectively. For subjects with persistent AF, 127 (97.0%) active measurements and 2,240 (99.2%) periodic measurements presented as AF by the algorithm. For subjects with paroxysmal AF, 36 (17%) active measurements and 717 (19.8%) periodic measurements were identified as AF by the algorithm. All persistent AF could be detected as “AF episodes” by the PPG algorithm on the first monitoring day, while 14 (70%) patients with paroxysmal AF demonstrated “AF episodes” within the first six days. The average time to detect paroxysmal AF was two days [(IQR) 1.25–5.75] by active measuring and one day [(IQR) 1.00–2.00] by periodic measurement (P =.10). The first detection time of AF burden < 50%/24 hours was four days by active measuring and two days by periodic measuring. The first detection time of AF burden > 50%/24 hours was one day for both active and periodic measurement (active measurement P =.02, periodic measurement P =.03).

Conclusions:

PPG-based smart devices demonstrated good AF predictive ability in both active and periodic measurements. However, AF type could impact detection, resulting in increased monitoring time. Clinical Trial: This study was registered in the Chinese Clinical Trial Registry of the International Clinical Trials Registry Platform of the World Health Organization (ChiCTR-OOC-17014138).