Abstract 15268: Abnormal Single-Lead Electrocardiograms Predict Future Atrial Fibrillation Risk: The VITAL-AF Trial

Abstract

Introduction: The use of single-lead electrocardiograms (1L ECG) is increasing due to the proliferation of devices available to consumers. Automated software algorithms provide real time analyses using the 1L ECG signal, yet it is unclear whether these interpretations confer a future risk of AF. Aim: Investigate the risk of future AF by automated 1L ECG interpretation in patients who did not receive an AF diagnosis within 60 days of screening 1L ECG Methods: Patients enrolled in the VITAL-AF trial were categorized based on automated interpretation of 1L ECGs by the AliveCor KardiaMobile device: Normal = all 1L ECGs were interpreted as “Normal”, Possible AF = at least one “Possible AF”, Other = at least one “No Analysis” or “Unclassified” but no “Possible AF”. Time to incident AF was calculated with the beginning of the at-risk period defined as 60 days after the latest 1L ECG. Confirmatory testing was not required by the trial protocol, so the at-risk period was designed to exclude delayed AF diagnoses. Cox proportional hazard regression models were used to compare groups adjusting for age, body mass index, hypertension, myocardial infarction, diabetes mellitus, congestive heart failure and smoking status. Results: The study included 14,181 patients ( Normal n = 10,930, Other n=2,689, Possible AF n= 562). As expected, a report of Possible AF had the highest AF incidence (5.75% per year, 95% Confidence Interval [CI] 4.12-8.00) and was associated with greater risk of incident AF compared to Normal (adjusted Hazard Ratio [aHR] 2.73, 95% CI 1.87-3.98). Interestingly, a diagnosis of Other was also associated with a marked increased risk of incident AF (2.37, 95% CI 1.88-2.99) ( Figure) . Conclusions: Abnormal 1L ECGs with incident AF not confirmed within 60 days were still predictive of future AF compared to “Normal” tracings. Further monitoring may be warranted following abnormal 1L ECGs.