Fig. 7: Clinical value of electromechanical assessment in patients at early stages of atrial fibrillation progression.

a Schematic flow-chart and study outcomes. *, † one patient in each of these groups also had left ventricular ejection fraction (LVEF) < 50%. b Prevalence of each of the CHA₂Ds₂VASc variables. c Comparisons of electromechanical dissociation (EMD) values between patients with successful and failed pharmacological cardioversion (PhCV). The square and whiskers show mean and standard deviation. Two-sided Student’s t test was used to assess differences. d Receiver operating characteristic curves (ROC) of four models for predicting acute PhCV. e Odds ratio and Confidence Intervals95% for each variable shown in the selected models (in d). Wald’s test was used to assess statistical significance. f Integrated discrimination improvement (IDI) upon addition to the AF classification: the electrical model (DF leads II and V2, N = 70), the clinical model (body mass index [BMI] and left atrial diameter [LAd], N = 80), and the EMD (N = 58). Bars and whiskers show IDI and standard error. The statistical significance of IDI was assessed with a Z test. g Quantification of EMD in patients with paroxysmal (PxAF) and persistent AF (PsAF). A mixed model based on Generalized Estimating Equations (GEE) was used to test for significant differences and interactions. h ROC curves of the models in (d) (but AF classification) to predict acute PhCV in the subset of patients with PsAF. Data normality was assessed with the Shapiro–Wilk test. Source data are provided as a Source Data file. TIA: transient ischemic attack. TEE transesophageal echocardiography, TTE transthoracic echocardiography.