Single-lead electrocardiograms and artificial intelligence for managing cardiac rhythm irregularities in general...

Cardiac rhythm irregularities, including atrial fibrillation, atrial premature beats and ventricular premature beats, are the reason for about 33% of electrocardiogram tests performed by general practitioners. Artificial intelligence has shown diagnostic accuracy in electrocardiogram analysis equivalent to that of cardiologists, and single-lead electrocardiograms offer simplified usability.

To assess the acceptability and utility of a simplified single-lead electrocardiogram device in general practice, integrated with artificial intelligence-based interpretation and a clinical decision support tree derived from artificial intelligence analysis.

A decision tree based on artificial intelligence-interpreted electrocardiography was developed for managing cardiac irregularities, and validated by expert cardiologists. Between October 2023 and March 2024, a survey was conducted among 1102 general practitioners in France to assess all the study objectives.

Of the 221 general practitioners who responded to the survey, 44% owned an electrocardiogram device, and 58% routinely performed electrocardiograms when rhythm irregularities are suspected. Seventy-seven percent believed that a single-lead electrocardiogram device would simplify electrocardiogram use in practice. A key factor for general practitioners that would enhance electrocardiogram use in their practice is publication of official guidelines. Seventy-two percent of general practitioners reported that they would use electrocardiograms more often if artificial intelligence were available for interpretation, and 57% of general practitioners considered artificial intelligence to be a diagnostic aid, rather than a fully autonomous system. The relevance of the decision tree was rated at 8 out of 10.

This general practitioner survey suggests that a single-lead electrocardiogram tool, integrating artificial intelligence-based trace interpretation with a decision tree derived from artificial intelligence analysis, could streamline the management of cardiac rhythm irregularities in general practice. However, its adoption requires the development of guidelines by scientific societies and approval from the national insurance system for reimbursement. Further research is needed to evaluate its feasibility and clinical relevance in real-world settings.