TRANSESOPHAGEAL ECHOCARDIOGRAPHICALLY FACILITATED EARLY CARDIOVERSIONFROM ATRIAL-FIBRILLATION USING SHORT-TERM ANTICOAGULATION - FINAL RESULTS OF A PROSPECTIVE 4.5-YEAR STUDY

Objectives. We sought to validate the safety of transesophageal echoca rdiographically guided early cardioversion in conjunction with short-t erm anticoagulation as a strategy for guiding early cardioversion in h ospitalized patients with atrial fibrillation. Background. Because atr ial thrombi are poorly seen by conventional imaging techniques, severa l weeks of prophylactic anticoagulation is routinely prescribed before cardioversion. Transesophageal echocardiography is a superior test fo r identifying atrial thrombi; preliminary feasibility studies have sup ported its use to guide early cardioversion for patients in whom no th rombus is observed, but safety has not been validated in any large ser ies. Methods. All patients admitted to hospital with atrial fibrilla t ion during a 4.5-year period were screened. The inclusion criterion wa s a clinical duration of atrial fibrillation >2 days or of unknown dur ation. Patients received anticoagulation with heparin/warfarin and und erwent conventional transthoracic echocardiography followed by transes ophageal study. Patients in whom transesophageal echocardiography reve aled no atrial thrombus underwent pharmacologic or electrical cardiove rsion followed by warfarin therapy for 1 month. Cardioversion was defe rred in patients with evidence of atrial thrombi, and they received pr olonged warfarin treatment. Results. Two hundred thirty-three patients (86% of those eligible) agreed to participate, and 230 underwent tran sesophageal echocardiography. Transesophageal echocardiography identif ied 40 atrial thrombi (left atrium 34, right atrium 6) in 34 patients (15%). One hundred eighty six (95%) of 196 patients without thrombi ha d successful cardioversion to sinus rhythm, all without prolonged anti coagulation, and none (0%, 95% confidence interval 0% to 1.6%) experie nced a clinical thromboembolic event. Eighteen patients with atrial th rombi underwent uneventful cardioversion after prolonged anticoagulati on. Conclusions. Compared with smaller series that have shown only fea sibility, this large prospective and consecutive study of patients und ergoing transesophageal echocardiographically facilitated early cardio version in conjunction with short-term anticoagulation validates the s afety of this strategy. This treatment algorithm has a safety profile similar to conventional therapy and minimizes both the period of antic oagulation and the overall duration of atrial fibrillation.