Smartwatch and wearable detection of atrial fibrillation: Current evidence and clinical implications

Consumer wearables have made atrial fibrillation (AF) detectable at population scale. Since 2019, smartwatch and smartphone studies enrolling thousands of participants have shown that photoplethysmography-based algorithms reliably generate AF notifications in real-world use, and that such a notification carries a high positive predictive value when benchmarked against a gold-standard electrocardiogram patch. This review follows the evidence along the clinical chain, from detection through notification and confirmation of diagnosis to treatment and outcome, and the strength of that evidence diminishes at each link. Population-scale studies established feasibility but did not measure sensitivity. Diagnostic accuracy is heterogeneous and conditional: it depends on the reference standard, the degree of supervision, the population studied, and the handling of inconclusive recordings. Randomised trials show that wearable-enabled pathways can accelerate diagnosis, a process outcome; yet none has shown that wearable-detected AF screening reduces stroke, systemic embolism or death. The treatment evidence that would justify acting on a detected arrhythmia comes from implanted-device populations, not consumer-wearable cohorts, and shows that anticoagulating device-detected AF trades a reduction in ischaemic stroke against an increase in major bleeding. Implementation raises further unresolved problems: confirmation-pathway heterogeneity, workflow burden, adherence, equity and overdiagnosis. Consumer wearables can detect AF at population scale, yet whether this improves hard clinical outcomes remains unresolved. Ongoing randomised trials must answer this critical question.