Cardiology ·
AI-ECG-Guided Monitoring Modestly Increased Detection of Atrial Findings
In the single-center, open-label LATENT randomized trial, 14,665 adults without previous atrial fibrillation or flutter were assigned to AI-ECG-guided care or usual care. The strategy modestly increased detection of a composite of atrial fibrillation, flutter, and prespecified pre-AF findings, but results were statistically fragile and did not significantly increase AF/flutter detection alone.
- Trial randomized 14,665 adults without prior AF or AFl.
- Composite atrial findings were detected in 0.74% versus 0.48%.
- The primary result was sensitive to reclassification of one event.
- AF/AFl detection alone did not differ significantly.
The LATENT trial randomized 14,665 adults aged 40 to 85 years without previous atrial fibrillation (AF) or atrial flutter (AFl) to an AI-ECG-guided strategy or usual care. Participants identified as high risk by AI analysis of a sinus-rhythm ECG could be offered up to seven days of Holter monitoring. High-risk intervention participants who were monitored were matched with controls who received identical monitoring; cardiologists interpreting results were blinded to AI findings and group assignment. The primary outcome was newly detected AF, AFl, or prespecified pre-AF atrial findings within 90 days.
The composite outcome occurred in 54 of 7,337 intervention participants (0.74%) and 35 of 7,328 controls (0.48%; odds ratio 1.54, 95% CI 1.01–2.37; P=0.046). The result was sensitive to reclassification of one event, which changed the P value to 0.057. The intervention effect differed by baseline AI-ECG risk: among high-risk participants, the outcome occurred in 32.1% versus 2.6%; among low-risk participants, rates were 0.4% versus 0.5%. AF/AFl detection alone and antiarrhythmic treatment initiation did not differ significantly between groups.
The findings suggest that AI-ECG risk stratification may help target rhythm monitoring, but the modest, fragile composite-outcome result does not establish improved patient outcomes or support routine adoption. Monitoring-related adverse events, including skin reactions and discomfort, were reported by 22.9% of monitored intervention participants and 18.8% of matched controls.
In practice
For adults aged 40 to 85 years without previous AF or AFl, this single-center trial suggests that AI-ECG risk stratification can target monitoring toward people more likely to have AF, AFl, or prespecified pre-AF findings detected. The overall increase in the composite outcome was small and statistically fragile, and AF/AFl detection alone and antiarrhythmic treatment initiation did not differ significantly. The study does not show improved patient outcomes, so it does not establish a basis for routine adoption of AI-guided monitoring.
AI-generated clinical interpretation · Not editor-reviewed
Is this summary clinically accurate?
Help fellow clinicians: your rating sends inaccurate summaries straight to our editors.
Sign in to rate this summary →Source
Circulation: Leveraging AI-ECG Technology for Early Notification and Tracking of AF Development (LATENT): A Randomized Controlled Trial ↗This is an automated AI-condensed summary that has not yet been reviewed by an editor. Always consult the full item at the original source.