ID: 4349493 SINUS RHYTHM ECG DURING BLANKING PERIOD PREDICTS ZERO RECURRENCE IN PERSISTENT AF: A MACHINE LEARNING APPROACH

Background: Clinical atrial fibrillation (AF) progression from atrial high-rate episodes (AHREs) poses significant clinical risks.Objective: This study aimed to develop a scoring system to predict clinical AF progression.Methods: Patients with documented AHRE but no prior clinical AF were enrolled.A simplified scoring system was developed to predict subsequent clinical AF progression using multivariable logistic regression.Results: Among 182 patients with documented AHRE but no prior clinical AF, 22 (12%) developed clinical AF, confirmed by 12-lead electrocardiogram.Multivariate analysis identified several key predictors for the progression of AHRE to clinical AF (Figure A): CHA 2 DS 2 _VAS score .5 (odds ratio [OR]: 2.811, 95% confidence interval [CI]: 1.114-7.094),left atrial (LA) diameter .46mm (OR: 3.894, 95% CI: 1.499-10.114),and the duration of AHRE, which were strongly correlated with increased progression risk.Based on these factors, a scoring model was developed, assigning points as follows: CHA 2 DS 2 _-VAS score .5 (1 point), LA diameter .46mm (1 point), AHRE duration .6 minutes (1 point), .6 hours (2 points), and .24hours (3 points).The accumulating score (range 0-5) achieved an area under the receiver-operating characteristic curve (AUC) of 0.78 (95% CI: 0.65-0.89),which was superior to CHA 2 DS 2 _VAS score alone (AUC: 0.639, 95% CI: 0.503-0.774)(Figure B) and an accumulated score of 4 was associated with a 50% risk of clinical AF progression (Figure C).Conclusion: This simplified scoring system demonstrated high predictive value, effectively stratifying patients by AF progression risk and offering a reliable clinically applicable tool for identifying high-risk individuals.