OBJECTIVES : Atrial fibrillation (AF) is a prevalent arrhythmia associated with substantial morbidity attributable largely to a 5-fold increased risk of stroke. Anticoagulation can prevent strokes in individuals with AF, but can cause bleeding. Population-based screening for AF is endorsed by certain guidelines but can be performed using many approaches, including newly developed smart watches. Yet whether AF screening is clinically effective is unclear.

METHODS : We constructed a microsimulation model in C++ to project the lifetime trajectories of a large cohort reflecting the subset of the US population meeting proposed age-based criteria for AF screening (age ≥ 65). We assessed several strategies involving a clinician-prescribed smart watch for AF screening and compared this to other contemporary strategies including combinations of pulse palpation, electrocardiography, and short-term event monitoring. We additionally assessed the effect of varying screening start age and frequency. The primary outcome was clinical effectiveness as defined by an increase in quality-adjusted life years.

RESULTS : Several smart watch strategies dominated the scenario of no screening, as did some other screening strategies (QALYs gained 30-130 per 10,000 persons). A smart watch strategy utilizing PPG technology and confirmed by both watch ECG and patch monitor appeared to be the most effective strategy. Deployment of the best performing strategy resulted in 130 QALYs gained and a reduction in stroke incidence rate by 0.1 per 1000 person-years, despite an increase in major bleeding by 0.3 per 1000 person-years.

CONCLUSIONS : Contemporary screening strategies utilizing combinations of traditional and smart watch modalities can improve outcomes in AF. Future guidelines should reflect the relative effectiveness of specific screening modalities.