Predicting Unsuccessful Electrical Cardioversion for Acute Atrial Fibrillation (from the AF-CVS Score)

Electrical cardioversion (ECV) is the standard treatment for acute atrial fibrillation (AF), but identification of patients with increased risk of ECV failure or early AF recurrence is of importance for rational clinical decision-making. The objective of this study was to derive and validate a clinical risk stratification tool for identifying patients at high risk for unsuccessful outcome after ECV for acute AF. Data on 2,868 patients undergoing 5,713 ECVs of acute AF in 3 Finnish hospitals from 2003 through 2010 (the FinCV study data) were included in the analysis. Patients from western (n = 3,716 cardioversions) and eastern (n = 1,997 cardioversions) hospital regions were used as derivation and validation datasets. The composite of cardioversion failure and recurrence of AF within 30 days after ECV was recorded. A clinical scoring system was created using logistic regression analyses with a repeated-measures model in the derivation data set. A multivariate analysis for prediction of the composite end point resulted in identification of 5 clinical variables for increased risk: Age (odds ratio [OR] 1.31, confidence interval [CI] 1.13 to 1.52), not the First AF (OR 1.55, CI 1.19 to 2.02), Cardiac failure (OR 1.52, CI 1.08 to 2.13), Vascular disease (OR 1.38, CI 1.11 to 1.71), and Short interval from previous AF episode (within 1 month before ECV, OR 2.31, CI 1.83 to 2.91) [hence, the acronym, AF-CVS]. The c-index for the AF-CVS score was 0.67 (95% CI 0.65 to 0.69) with Hosmer-Lemeshow p value 0.84. With high (>5) scores (i.e., 12% to 16% of the patients), the rate of composite end point was ∼40% in both cohorts, and among low-risk patients (score <3), the composite end point rate was ∼10%. In conclusion, the risk of ECV failure and early recurrence of AF can be predicted with simple patient and disease characteristics.