By Joshua D. Moss, MD

Associate Professor of Clinical Medicine, Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco

Dr. Moss reports he is a consultant for Biosense Webster and Abbott.

SYNOPSIS: In patients with risk factors for atrial fibrillation, screening with a self-applied wearable ECG patch resulted in significantly increased rates of new atrial fibrillation diagnoses within four months, along with greater use of anticoagulants and healthcare resources.

SOURCE: Steinhubl SR, et al. Effect of a home-based wearable continuous ECG monitoring patch on detection of undiagnosed atrial fibrillation: The mSToPS randomized clinical trial. JAMA 2018;320:146-155.

Current screening for atrial fibrillation (AF) in high-risk populations generally is limited to auscultation, pulse palpation, and “spot” 12-lead ECGs during routine visits. Steinhubl et al sought to study the effect of a more aggressive but practical method of screening, using a self-applied, two-week, continuous ECG monitoring patch at home during routine activities.

The study population consisted of participants in a single large national health insurance plan, with eligible patients chosen from more than 1 million candidates based on risk factors for AF: age ≥ 75 years of age, or men > 55 years of age or women > 65 years of age with one or more comorbidities (including prior stroke, heart failure, or the combination of diabetes and hypertension, among others). Patients with any current or prior diagnosis of atrial arrhythmia who already were on anticoagulation therapy or who had an implantable pacemaker or defibrillator were excluded. More than 100,000 eligible patients were contacted, with most eventual enrollees contacted by email. Individuals who chose to enroll were consented remotely. A total of 2,659 patients were randomized: 1,366 received an ECG patch and instructions for self-application within two weeks (immediate group), and 1,293 received their patches four months later (delayed group). Additionally, 5,318 matched observational controls who were eligible for the study but not contacted for participation in the randomized trial were identified, two for each patient randomized.

The primary endpoint in the intention-to-treat analysis of randomized patients was incidence of newly diagnosed AF (defined as ≥ 30 seconds of AF, flutter detected by device, or a new clinical diagnosis recorded in claims data at the end of the initial four-month monitoring period). In the immediate monitoring group, 908 of 1,366 participants wore an ECG patch, and incidence of new AF was 3.9%. In the delayed monitoring group, incidence of new AF was 0.9% in the first four months (before those participants received their patch). In the observational study with one-year follow-up, new AF was detected at a rate of 6.7 per 100 person-years in the actively monitored cohort (the two arms of the randomized trial) vs. 2.6 per 100 person-years in the matched observational controls.

Patients who were actively monitored were more likely to start both anticoagulation and antiarrhythmic medications. Further, there were more office visits as well as cardioversion and ablation procedures for these patients. However, the actively monitored group experienced a slightly lower incidence of hospitalizations or ED visits.


Improvements in digital technology for cardiac rhythm monitoring and AF diagnosis have made wearable devices that patients can use to send data to their physicians, or even self-diagnose, more accessible. Data from the mSToPS trial corroborate the work of other investigators who have evaluated more frequent or continuous monitoring and found a higher incidence of AF than would have been realized otherwise, including the REHEARSE-AF and CRYSTAL-AF studies. Uniquely, patients in the mSToPS trial were approached mostly via email, consented remotely, and applied and removed their own monitoring devices. In an ongoing trial, researchers are enrolling patients to use Apple Watch-based photoplethysmography to monitor for AF.

The increased rate of AF detection is not surprising. However, the absolute difference in detection rates between those monitored (for a median total monitoring time of about 25 days over a four-month period) and those not monitored still is impressive (considering 458 of 1,366 patients randomized to immediate monitoring never wore a patch). Also unsurprising is the resultant increase in healthcare resource use in the actively monitored cohort, with an increase in office visits, more prescriptions for anticoagulation therapy, and additional cardioversions and ablation procedures. The results may not be completely generalizable to a broader population. Patients who were invited to participate and enrolled were more likely to have been invited by email rather than direct mail, were slightly younger, more often male, and exhibited less hypertension and diabetes but more obesity and sleep apnea than those who declined. Additionally, patients who participated in randomization but never wore a monitor had some different characteristics than those who wore the patch. The larger question raised is whether more aggressive screening for AF in asymptomatic patients will translate to real long-term health benefits, and at what cost. The primary goal of detection for many patients would be stroke prevention via anticoagulation, but such a benefit has not yet been demonstrated. Additionally, more anticoagulation inevitably will lead to more bleeding events. The recently published NAVIGATE ESUS study ended early because patients empirically anticoagulated after a presumed embolic stroke without a clear source experienced more bleeding events and no apparent change in recurrent stroke risk at 11 months’ follow-up. There are secondary benefits to earlier AF detection, such as a higher likelihood of aggressive risk factor modification, but also other ill effects, such as anxiety for some and complications of therapy for others. Whether patients should screen and diagnose themselves with AF is the subject of active debate in the cardiology and EP communities. However, one thing is certain: Our methods for educating patients about AF and all the potential benefits and risks of early diagnosis and treatment must evolve at the same pace as the technology for detection.