[Skip to Navigation]
Original Investigation
February 24, 2021

Screening for Atrial Fibrillation in the Older Population: A Randomized Clinical Trial

Author Affiliations
  • 1Hurvitz Brain Sciences Program, Sunnybrook Research Institute, and Division of Neurology, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada
  • 2Division of Neurology, Department of Medicine, University of Toronto, Toronto, Ontario, Canada
  • 3Clinic and Policlinic for Cardiology, University Hospital, Leipzig, Germany
  • 4Department of Cardiology, University Medical Center Göttingen, Göttingen, Germany
  • 5DZHK (German Centre for Cardiovascular Research), partner site Göttingen, Göttingen, Germany
  • 6Department of General Practice, University Medical Center Göttingen, Göttingen, Germany
  • 7Department of General Practice and Primary Care, University Medical Center Hamburg-Eppendorf, Hamburg-Eppendorf, Germany
  • 8Libin Cardiovascular Institute, University of Calgary, Calgary, Alberta, Canada
  • 9Women’s College Hospital, Department of Family and Community Medicine, University of Toronto, Toronto, Ontario, Canada
  • 10Population Health Research Institute, McMaster University, Hamilton, Ontario, Canada
  • 11Clinic for Cardiology and Pneumology, University Medicine Göttingen, Göttingen, Germany
  • 12LMC Manna Research, Burlington, Ontario, Canada
  • 13Southlake Regional Health Centre, Newmarket, Ontario, Canada
  • 14Camrose Primary Care Network, Camrose, Alberta, Canada
JAMA Cardiol. 2021;6(5):558-567. doi:10.1001/jamacardio.2021.0038
Key Points

Question  Will screening older individuals for atrial fibrillation with a wearable electrocardiographic monitor be feasible, detect a high rate of atrial fibrillation, and lead to anticoagulation for most patients?

Findings  In a randomized clinical trial of 856 participants aged 75 years or older with hypertension from outpatient primary care practices, new atrial fibrillation was detected in 5.3% of the screening group vs 0.5% of the control group. Median atrial fibrillation duration on continuous electrocardiographic monitoring was 6.3 hours, and anticoagulation was prescribed to 75.0% of the participants with screen-detected atrial fibrillation.

Meaning  In this trial, a wearable electrocardiogram-based screening intervention increased atrial fibrillation detection 10-fold and prompted anticoagulation in most cases; this strategy warrants evaluation to prevent stroke.


Importance  Atrial fibrillation (AF) is a major cause of preventable strokes. Screening asymptomatic individuals for AF may increase anticoagulant use for stroke prevention.

Objective  To evaluate 2 home-based AF screening interventions.

Design, Setting, and Participants  This multicenter randomized clinical trial recruited individuals from primary care practices aged 75 years or older with hypertension and without known AF. From April 5, 2015, to March 26, 2019, 856 participants were enrolled from 48 practices.

Interventions  The control group received standard care (routine clinical follow-up plus a pulse check and heart auscultation at baseline and 6 months). The screening group received a 2-week continuous electrocardiographic (cECG) patch monitor to wear at baseline and at 3 months, in addition to standard care. The screening group also received automated home blood pressure (BP) machines with oscillometric AF screening capability to use twice-daily during the cECG monitoring periods.

Main Outcomes and Measures  With intention-to-screen analysis, the primary outcome was AF detected by cECG monitoring or clinically within 6 months. Secondary outcomes included anticoagulant use, device adherence, and AF detection by BP monitors.

Results  Of the 856 participants, 487 were women (56.9%); mean (SD) age was 80.0 (4.0) years. Median cECG wear time was 27.4 of 28 days (interquartile range [IQR], 18.4-28.0 days). In the primary analysis, AF was detected in 23 of 434 participants (5.3%) in the screening group vs 2 of 422 (0.5%) in the control group (relative risk, 11.2; 95% CI, 2.7-47.1; P = .001; absolute difference, 4.8%; 95% CI, 2.6%-7.0%; P < .001; number needed to screen, 21). Of those with cECG-detected AF, median total time spent in AF was 6.3 hours (IQR, 4.2-14.0 hours; range 1.3 hours-28 days), and median duration of the longest AF episode was 5.7 hours (IQR, 2.9-12.9 hours). Anticoagulation was initiated in 15 of 20 patients (75.0%) with cECG-detected AF. By 6 months, anticoagulant therapy had been prescribed for 18 of 434 participants (4.1%) in the screening group vs 4 of 422 (0.9%) in the control group (relative risk, 4.4; 95% CI, 1.5-12.8; P = .007; absolute difference, 3.2%; 95% CI, 1.1%-5.3%; P = .003). Twice-daily AF screening using the home BP monitor had a sensitivity of 35.0% (95% CI, 15.4%-59.2%), specificity of 81.0% (95% CI, 76.7%-84.8%), positive predictive value of 8.9% (95% CI, 4.9%-15.5%), and negative predictive value of 95.9% (95% CI, 94.5%-97.0%). Adverse skin reactions requiring premature discontinuation of cECG monitoring occurred in 5 of 434 participants (1.2%).

Conclusions and Relevance  In this randomized clinical trial, among older community-dwelling individuals with hypertension, AF screening with a wearable cECG monitor was well tolerated, increased AF detection 10-fold, and prompted initiation of anticoagulant therapy in most cases. Compared with continuous ECG, intermittent oscillometric screening with a BP monitor was an inferior strategy for detecting paroxysmal AF. Large trials with hard clinical outcomes are now needed to evaluate the potential benefits and harms of AF screening.

Trial Registration  ClinicalTrials.gov Identifier: NCT02392754