Using AI-Enhanced Electrocardiograms to Guide Heart Failure Prevention Treatments A Retrospective Multicenter Study on Preventing Left Ventricular Dysfunction

What this Trial Is About

Researchers are evaluating whether artificial intelligence-enabled electrocardiography AI-ECG can identify people at high risk for left ventricular dysfunction and if early use of guideline-directed medical therapies can reduce this risk. This multicenter retrospective study uses data from Taiwans healthcare system collected between 2016 and 2024, focusing on patients with preserved left ventricular systolic function at baseline. The study applies AI-ECG risk stratification and emulates a target trial to assess therapy impacts on heart function decline. The study compares patients starting guideline-directed medical therapiessuch as angiotensin converting enzyme inhibitors ACEi, angiotensin II receptor blockers ARB, beta-blockers, and sodium-glucose cotransporter 2 SGLT2 inhibitorsagainst those not receiving these treatments. AI-ECG models developed from over 50,000 paired ECG and echocardiogram recordings identify high-risk individuals despite normal initial ejection fraction. The study evaluates the association between early preventive medication use and incident left ventricular dysfunction. Participants must have had both an ECG and echocardiogram within 90 days and preserved heart pumping function initially. Researchers will review medical records, medication prescriptions, ECGs, and echocardiography results. The main measurement is the rate of left ventricular ejection fraction dropping to 40% or below during up to 10 years of follow-up. Safety and outcomes will be monitored using real-world data, with analyses exploring robustness across clinical settings and patient profiles.

AI-Enabled Electrocardiogram-Guided Guideline-Directed Medical Therapy on Incident Left Ventricular Dysfunction: A Target Trial Emulation Study