Transesophageal echocardiographic predictors for maintenance of sinus rhythm after electrical cardioversion of atrial fibrillation

The aim of this study was to assess the value of transesophageal echocardiography (TEE) in patients with atrial fibrillation in predicting restoration and maintenance of sinus rhythm after electrical cardioversion. TEE was performed in 62 patients with atrial fibrillation before their first elective cardioversion. Clinical variables evaluated were: age, gender, duration, and etiology of atrial fibrillation. TEE variables included: left atrial (LA) length, width, and size, LA annulus size, as well as presence of LA spontaneous contrast, thrombus and mitral regurgitation, LA appendage size and flow, and left ventricular function. Based on initial outcome of cardioversion, patients were grouped into patients who remained in atrial fibrillation and in whom sinus rhythm was restored. The latter group of patients was followed for 1 year, and grouped into patients who reverted to atrial fibrillation and in whom sinus rhythm was maintained. Successful cardioversion was achieved in 50 of 62 patients (81%). None of the clinical or TEE variables were related to initial outcome. At 1-year follow-up, 29 of 50 patients (58%) who underwent successful cardioversion continued to have sinus rhythm. The following variables were related to maintenance of sinus rhythm: duration of atrial fibrillation (6.7 +/- 7.3 vs 2.0 +/- 2.4 months; p <0.005); LA length (6.2 +/- 0.7 vs. 5.5 +/- 1.0 cm; p <0.008); width (5.1 +/- 0.5 vs. 4.5 +/- 0.7 cm; p <0.002); size (26.4 +/- 5.0 vs 19.8 +/- 6.5 cm(2); p <0.0005); annulus size (4.0 +/- 0.2 vs 3.7 +/- 0.3 cm; p <0.0005); presence of LA spontaneous contrast (13 [62%] vs 4 [14%]; p <0.002), and LA appendage flow (19 +/- 8 vs 36 +/- 15 cm/s; p <0.0005). In multivariate logistic regression analysis, LA annulus size, but especially LA appendage flow, were significantly associated with maintenance of sinus rhythm. Thus, in TEE-guided electrical cardioversion of atrial fibrillation, variables often used to assess thromboembolic risk may also be used to predict I-year outcome of cardioversion. (C) 1997 by Excerpta Medica, Inc.