Systematic screening for atrial fibrillation with non-invasive devices: a systematic review and meta-analysis

Background

Systematic screening individuals with non-invasive devices may improve diagnosis of atrial fibrillation (AF) and reduce adverse clinical events. We systematically reviewed the existing literature to determine the yield of new AF diagnosis associated with systematic AF screening, the relative increase in yield of new AF diagnosis with systematic screening compared to usual care, and the effect of systematic AF screening on clinical outcomes compared with usual care.

Methods

The Medline, Embase, Web of Science and Cochrane Library databases were searched from inception through 1st February 2025 for prospective cohort studies or randomised clinical trials (RCTs) of systematic AF screening with the outcome of incidence of previously undiagnosed AF from screening. Incidence rates (IR) and relative risks were calculated and random effects meta-analysis performed to synthesise rates of AF in prospective cohort studies and RCTs, as well as outcomes in RCTs.

Findings

From 3806 unique records we included 32 studies representing 735,542 participants from 8 RCTs and 24 prospective cohorts. The diagnosis rate for incident AF in prospective cohorts was 2.75% (95% CI 1.87–3.62), and the pooled relative risk in RCTs was 2.22 (95% CI 1.41–3.50). The use of age and NT-proBNP (IR 4.36%, 95% CI 3.77–5.08) or AF risk score classification (4.79%, 95% CI 3.62–6.29) led to higher new AF diagnosis yields than age alone (0.93%, 95% CI 0.28–2.99). Pooled data from RCTs did not demonstrate an effect of screening on death (RR 1.01, 95% CI 0.97–1.05), cardiovascular hospitalisation (1.00, 95% CI 0.97–1.03), stroke (0.95, 95% CI 0.87–1.04) or bleeding (1.08, 95% CI 0.91–1.29).

Interpretation

Systematic screening for AF using non-invasive devices is associated with increased diagnosis of AF, but not reduced adverse clinical events. Screening studies of AF utilising alternative risk stratifications and outcome measures are required.