Predictors of sinus rhythm 6 weeks after cardioversion of atrial fibrillation: a pre-planned post hoc analysis of the X-VeRT trial.

Abstract

Using a pre-planned post hoc analysis of patients included in X-VeRT, we evaluated predictors of sinus rhythm at 6 weeks after planned cardioversion. Receiver operating characteristic curves and logistic regression models were used to evaluate continuous and categorical variables as predictors of sinus rhythm 6 at weeks from cardioversion (end of study). The primary analysis was performed in successfully cardioverted patients with an evaluable electrocardiogram at end of study. A second analysis evaluated additional patients who spontaneously restored sinus rhythm before planned cardioversion. Of the 1504 patients with atrial fibrillation of >48 h or of unknown duration who were randomly assigned to either rivaroxaban or vitamin K antagonist, 1039 (64.6 ± 10.3 years, 73.4% male) underwent planned cardioversion and were included in this study. Patients receiving early cardioversion (i.e. between 1 and 5 days from hospitalization) had a 67% higher probability to have sinus rhythm at end of study than those who received delayed cardioversion (i.e. between 21 and 56 days from hospitalization) [odds ratio (OR) 1.67, confidence interval (CI) 1.27-2.18; P < 0.0001]. In a multivariate analysis of 17 baseline variables, patients with a CHADS2 score of 0 were 33% less likely to be in sinus rhythm than those with a CHADS2 score ≥2 (OR 0.66, CI 0.47-0.94; P = 0.0225). In the secondary analysis, spontaneous restoration of sinus rhythm was also found to predict sinus rhythm at end of study (OR 8.62, CI 1.54-48.16; P = 0.0142). In X-VeRT, early cardioversion and high CHADS2 scores predicted sinus rhythm at 6 weeks from cardioversion.