This Week in Cardiology: AI Adjudication, Bleeding Risk, and Digital Care

From the Journal of the American College of Cardiology, a trial examined the utility of an AI-based adjudication system for major adverse cardiac events in patients with myocardial infarction. In this study involving 5,661 patients, the Auto-MACE system demonstrated high agreement ranging from 88% to 97% with human committee review for confident events, though agreement for all events was lower at 76% to 86%. The authors describe these findings as suggesting that while AI may assist in review, it is not currently a replacement for human adjudication due to the variability in agreement for less confident cases.

Meanwhile, researchers in the Journal of the American Heart Association explored prognostic markers following percutaneous coronary intervention. A post hoc analysis of the FAVOR III China trial evaluated angiographic microcirculatory resistance in 3,404 patients. Over a period of 36.0 months, overall major adverse cardiac event risks appeared similar between groups, yet a post-PCI resistance value of 250 or greater identified a specific subgroup with significantly increased three-year risk. The authors note that while this metric highlights a high-risk population, the overall prognostic significance remains insufficiently established to change standard practice immediately.

Elsewhere this week, attention turned to bleeding risks in atrial fibrillation. In European heart journal. Cardiovascular Imaging, a randomized controlled trial of 1,841 patients found that shorter pretreatment fibrin clot lysis time was associated with a significantly higher risk of major and clinically relevant non-major bleeding. The association was strongest when comparing the shortest versus longest lysis time quartiles. Clinicians are advised to interpret these observational findings cautiously, as the data do not establish causality between clot lysis speed and bleeding outcomes.

We also saw research in European journal of preventive cardiology regarding digital interventions for self-care. In a randomized controlled trial of 294 community-dwelling adults with coronary artery disease, a remotely delivered i-CARE intervention led to significant improvements in self-care maintenance, confidence, and blood pressure compared to standard care. These results are presented as preliminary evidence for digital self-care support. Finally, in Circulation, a Phase 2 randomized, double-blind, placebo-controlled trial of 241 symptomatic adults with frequent premature atrial contractions showed that memantine reduced PAC burden and new-onset atrial fibrillation compared with placebo over six weeks. Safety was acceptable with no drug-related serious adverse events, though the authors emphasize that these findings require confirmation in larger trials.