NEJM -- Cardioversion from Atrial Fibrillation without Prolonged Anticoagulation with Use of Transesophageal Echocardiography to Exclude the Presence of Atrial Thrombi

Volume 328:750-755		March 18, 1993		Number 11

Cardioversion from Atrial Fibrillation without Prolonged Anticoagulation with Use of Transesophageal Echocardiography to Exclude the Presence of Atrial Thrombi
Warren J. Manning, David I. Silverman, Stephen Gordon, Harlan M. Krumholz, and Pamela S. Douglas

ABSTRACT Background Because atrial thrombi are poorly detected by conventionalnoninvasive techniques such as transthoracic echocardiography,patients with prolonged atrial fibrillation usually receiveseveral weeks of oral anticoagulation therapy before cardioversionis attempted. We wondered whether transesophageal echocardiography,an accurate method of detecting atrial thrombi, would allowearly cardioversion to be performed safely if no thrombi wereidentified.

Methods A total of 669 consecutive patients admitted with thediagnosis of atrial fibrillation were screened. Patients wereexcluded if they were receiving long-term anticoagulation, ifthe duration of atrial fibrillation was two days or less, ifthey were not candidates for cardioversion, or if transesophagealechocardiography was contraindicated. Of 119 qualifying patients,94 agreed to participate; the average duration of atrial fibrillationwas 4.5 weeks. Participating patients underwent transthoracicechocardiography and transesophageal echocardiography followedby cardioversion if no thrombi were seen. Short-term anticoagulationwith heparin was used in 80 patients before cardioversion, and60 patients received warfarin for one month after cardioversion.

Results Fourteen atrial thrombi were identified in 12 patients(13 percent), and 12 of the 14 thrombi were visualized onlyon transesophageal echocardiography. Cardioversion was deferredin all 12 patients. Two of these 12 patients died suddenly;4 of the 10 surviving patients underwent uneventful cardioversionafter prolonged oral anticoagulation. Seventy-eight of the 82patients without thrombi underwent successful cardioversionto sinus rhythm (47 by means of antiarrhythmic drugs and 31by electrical cardioversion), all without long-term oral anticoagulation.None of these patients (95 percent confidence interval, 0 to4.6 percent) had an embolic event.

Conclusions In patients with atrial fibrillation of unknownor prolonged duration who are not receiving long-term anticoagulation,atrial thrombi are detected by transesophageal echocardiographyin only a small minority (13 percent in our study). Our preliminarydata suggest that if transesophageal echocardiography excludesthe presence of thrombi, early cardioversion can be performedsafely without the need for prolonged oral anticoagulation beforethe procedure.

Atrial fibrillation is one of the most common types of arrhythmia,occurring in up to 4 percent of patients over 60 years of age¹^,²^,³.This arrhythmia is characterized by a lack of organized atrialelectrical and mechanical activity, and the resulting bloodstasis favors the formation of atrial thrombi. Atrial fibrillationhas been identified as the rhythm responsible for more thanhalf of all instances of systemic thromboembolism from the heart⁴;the brain is the primary target⁵^,⁶. In addition, atrial fibrillationis associated with a decrease in cardiac output because of theloss of the atrial contribution to ventricular filling⁷.

Cardioversion is performed in patients with atrial fibrillationin an effort to improve cardiac function, relieve symptoms,and decrease the risk of thrombus formation⁸. Unfortunately,successful cardioversion is associated with a 5 to 7 percentincidence of embolism among patients who have not received anticoagulanttherapy⁹^,¹⁰^,¹¹^,¹²^,¹³^,¹⁴. Because atrial thrombi are poorly detectedby conventional noninvasive techniques, such as transthoracicechocardiography, the standard care of patients with atrialfibrillation of unknown or prolonged duration (longer than twodays) usually includes several weeks of oral anticoagulationbefore cardioversion is attempted⁸^,¹⁵. Anticoagulation decreasesthe risk of an embolic event after cardioversion to less than1.6 percent,¹⁰^,¹²^,¹⁶