Frequency and outcomes of arrhythmias and conduction system abnormalities using real-world handheld single-lead ECGs: A VITAL-AF trial analysis

Background Use of consumer handheld single-lead electrocardiograms (SL ECG) is increasingly common, yet frequency and prognostic implications of abnormal SL ECG findings remain unclear. Purpose To quantify the frequency and outcomes of cardiologist-diagnosed abnormal findings on real-world SL ECGs performed systematically among participants of an atrial fibrillation (AF) screening trial. Methods We analyzed tracings obtained prospectively from primary care patients aged ≥65 years who underwent screening with 30s second SL ECG using the KardiaMobile 1L device (AliveCor, Inc). All tracings were manually interpreted by board-certified study cardiologists and findings assigned into clinically relevant categories of abnormality based on prior literature. Associations between each category and 6.5-year all-cause mortality (maximum available follow-up) were assessed using Kaplan–Meier curves and multivariable Cox proportional hazards models with robust variance estimation clustered by patient (as some patients had multiple SL ECG). The model was adjusted for age, sex, diabetes, heart failure, coronary artery disease, and myocardial infarction. Patients with exclusively normal sinus rhythm (NSR) across all screening SL ECGs served as the reference group. Results We analyzed 16,496 SL ECGs from 14,403 patients. 41% of patients screened for AF had at least one abnormal finding (Table; Figure 1a, 1b). In multivariable-adjusted models, atrial fibrillation or flutter (HR 1.80, 95% CI 1.50–2.17), conduction system disease (HR 1.24, 95% CI 1.04–1.49), supraventricular arrhythmia (HR 1.23, 95% CI 1.02–1.49), and ventricular arrhythmia (HR 1.26, 95% CI 1.03–1.54) were each associated with higher mortality compared with NSR (Figure 1c). Tracings marked as unreadable were associated with nearly twofold higher mortality (HR 1.97, 95% CI 1.64–2.37). Conclusion In a large primary care screening trial population, nearly half of older adults had SL ECG evidence of arrhythmia or conduction system disease. Several non-AF-related abnormal findings were associated with greater mortality, highlighting a potentially broader prognostic utility of SL ECG. Future work is warranted to identify optimal methods to integrate and react to abnormal SL ECG findings observed in routine care.TableFigure