Artificial Intelligence-Enabled Electrocardiogram Screening for Detection of Low Ejection Fraction in Routine Care: A Systematic Review and Meta-Analysis of Randomized Controlled Trials.
Ali Abdelhaleem Omar Ahmed, Omer Abdalhaleem Omer Ahmed, Rana Abdelhaleem Omar Ahmed et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis evaluated randomised controlled trials of AI-enabled ECG screening for low ejection fraction in adults without known heart failure undergoing routine ECGs. Three trials, including 37,467 participants, compared screening with usual care. After adjustment for clustering, AI-ECG screening increased detection of new low EF (≤ 50%) confirmed by echocardiography (RR 1.36, 95% CI 1.12 to 1.64; p = 0.002), with a number needed to screen of 204. Echocardiography referral rates did not differ significantly between groups.
Key findings
- AI-ECG screening increased new low EF detection versus usual care (RR 1.36, 95% CI 1.12 to 1.64; p = 0.002; I 2 = 0%).
- The number needed to screen to detect one additional case was 204.
- Echocardiography referral rates were similar (RR 1.02, 95% CI 0.96 to 1.09; p = 0.55; I 2 = 36%).
- Across studies, the AI tool’s positive predictive value ranged from approximately 20% to 34%.
Why it matters
Low ejection fraction may be present before heart failure is recognised, while symptoms can be non-specific. These trial findings suggest that AI analysis of routine ECGs can improve detection without a substantial increase in echocardiography referrals. The modest positive predictive values also underscore that screening results require clinical confirmation rather than being treated as diagnoses.
What to keep in mind
Only three RCTs were included, and the positive predictive value varied across studies. The authors note that validation across diverse AI platforms is still warranted.