Predicting atrial fibrillation from ECG characteristics

Abstract Background/Introduction Atrial fibrillation (AF) is the most common cardiac arrythmia, and demographic shifts in age continues to increase the prevalence of atrial fibrillation on a global scale. With manifestations of short and intermittent duration, identifying and diagnosing AF often proves to be a challenge. As a consequence, patients who develop AF are at risk of experiencing stroke and other cardiac diseases such as heart failure (HF) before appropriate treatment is initialized. Despite the potential benefits of early AF detection, risk profiling based on electrocardiographic (ECG) findings is currently lacking. Several smaller studies have identified ECG characteristics associated with increased risk of AF, but these characteristics have not been individually evaluated in large-scale nationwide cohorts, and no evidence exists on the additive effect of these ECG characteristics. Purpose We aimed to evaluate the risk of new-onset AF, HF, and stroke according to five ECG characteristics previously linked to AF, in order to identify potential predictors of AF. Methods We used national registers and the Danish Nationwide ECG Cohort to include patients with ECG measurements taken between 2010 and 2022. Patients were grouped according to how many of the following ECG characteristics were present in their baseline ECG: Elevated heart rate (>80), prolonged PR-interval (≥196 ms for women and ≥204 ms for men), biphasic P-waves (positive and negative amplitude within same p-wave), abnormal P terminal force (≥ 6000 mV×ms), and prolonged P-wave duration(≥120 ms). Patients were followed for 2 years to evaluate the risk of atrial fibrillation, heart failure (HF), and stroke, according to baseline ECG characteristics. Results We included 1,505,474 patients of whom 594,702 had exactly one ECG characteristic, 197,367 had two characteristics, 44,983 had three characteristics, 6,390 had four characteristics, and 451 had all five characteristics on their baseline ECG. The risk of developing AF or HF increased with the number of ECG characteristics present on the baseline ECG. Patients with two ECG characteristics had a 2-fold increased risk of AF or HF compared to patients with no ECG characteristics (HR: 2.11 95%CI: 2.05-2.16). Patients with two or more ECG characteristics present on their baseline ECG exceeded a 5% 2-year absolute risk of developing either AF or HF. Among patients with only one ECG characteristic on their baseline ECG, elevated heart rate and prolonged PR interval were the factors most associated with an increased risk of AF compared to patients with none of the ECG characteristics on their baseline ECG. Conclusions The absolute and relative risk of atrial fibrillation, heart failure, and stroke increased with the number of ECG characteristics present on the baseline ECG. We suggest that the ECG characteristics evaluated in this study could be used in screening scenarios for atrial fibrillation to identify high risk patients.Absolute and relative 2-year risksRisk acc. to specific ECG characteristic