Arrhythmias & electrophysiology
Risk-Guided Atrial Fibrillation Screening With Artificial Intelligence–Enabled Electrocardiogram Models: A VITAL-AF Trial Analysis
Primary result · AF diagnosis rate, top decile of CH-AIResult — go to the resultDifference 2.32 /100 person-years 95% CI 0.01–4.63Authors' conclusion
Use of ECG-based AI and clinical factors identified individuals at particularly high risk for AF who may benefit from screening.
Study context
- Population
- Patients aged ≥65 years without prevalent atrial fibrillation (AF) at 16 primary care practices (Massachusetts General Hospital), with a 12-lead electrocardiogram (ECG) within 3 years before enrollmentof 30,630 without prevalent AF
- Participants
- 16,937 patients
- Comparison
- AF screening with single-lead ECG vs control, across AI-ECG and clinical AF risk deciles
- Primary endpoint
- AF screening effect: difference in incident AF diagnosis rate (per 100 person-years), screening vs control, across AF risk deciles
- Follow-up
- 2 years
Methodological detailscluster-randomized, 16 primary care practices
- Cluster-randomized
- 16 primary care practices
Primary result
AF diagnosis rate, top decile of CH-AI
Difference
2.32 /100 person-years
95% CI 0.01–4.63
P < 0.05
By group
Screeningvs Control
difference 2.32 /100 person-years (95% CI 0.01–4.63); P < 0.05
As reported in the abstract
Further findings
| Outcome | CHARGE-AF | ECG-AI | CH-AI |
|---|---|---|---|
| Discrimination of 2-year AF risk (AUROC; average precision) | |||
| AUROC | 0.711 (95% CI 0.671-0.749) | 0.784 (95% CI 0.743-0.819) | 0.788 (95% CI 0.754-0.824) |
| Average precision | 0.0952 (95% CI 0.0836-0.112) | 0.132 (95% CI 0.113-0.157) | 0.133 (95% CI 0.117-0.159) |
- Discrimination of 2-year AF risk (AUROC; average precision): Cohorts of Heart and Aging Research in Genomic Epidemiology-AF (CHARGE-AF) clinical score; ECG-AI = artificial intelligence (AI) model on 12-lead ECG alone; CH-AI = ECG-AI combined with CHARGE-AF
Original publication
Risk-Guided Atrial Fibrillation Screening With Artificial Intelligence–Enabled Electrocardiogram Models: A VITAL-AF Trial Analysis. J Am Coll Cardiol. Published 14 April 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →