Aims Despite exclusion of left atrial thrombi by transoesophageal echocardiography, cardioversion-related thromboembolism has been reported in atrial fibrillation or flutter. To define a low-risk group for cardioversion without previous anticoagulation, patients were selected for immediate cardioversion if there were no thrombi, no echo spontaneous contrast and the outflow velocity of the left atrial appendage was greater than 0·25 m . s 1 on transoesophageal echocardiography.
Methods and ResultsTwo hundred and forty-two consecutive patients referred for cardioversion of atrial fibrillation or flutter with a duration of more than 2 days and no anticoagulation therapy were examined with transoesophageal echocardiography. After the transoesophageal echocardiography examination, patients who were eligible for immediate cardioversion were anticoagulated with low molecular weight heparin (dalteparin) subcutaneously, together with warfarin prior to cardioversion. Dalteparin treatment was continued until the patient had reached therapeutic prothrombin values.Based on the transoesophageal echocardiographic findings the patients were divided into two groups: immediate cardioversion, group A, with a mean age of 62 13 years (n=162); or conventional warfarin treatment before cardioversion, group B, with a mean age of 67 10 years (P<0·05) (n=80). In group A, lone atrial fibrillation or flutter was more common (53%; 95% CI: 45-61) compared to group B (34%; 95% CI: 23-44, P<0·05), while heart disease was more common in group B (45%; 95% CI: 34-56) compared to group A (31%; 95% CI: 24-39, P<0·05). Echocardiography revealed thrombi in 5% (95% CI: 2·6-8) of the patients, left atrial size was larger, fractional shortening lower, and a higher proportion had impaired left ventricular function in group B. No thromboembolic event occurred at or after cardioversion in any of the patients; however, before planned cardioversion one transitory ischaemic attack, one lethal stroke and one cardiac death occurred in three of the patients with thrombi despite warfarin therapy. One-month follow-up maintenance of sinus rhythm was 75% in group A compared to 45% in group B (P<0·01).Conclusion After using our transoesophageal echocardiographic exclusion criteria (no thrombi, no spontaneous echo contrast and left atrial appendage outflow velocity d0·25 m . s 1 ) cardioversion can safely be performed in 2/3 of patients with atrial fibrillation or flutter without previous anticoagulation therapy. These patients maintained sinus rhythm significantly better after 1 month compared to patients with prolonged warfarin therapy before cardioversion.