Atrial fibrillation is the most frequently occurring cardiac arrhythmia, with 1.0-1.5 million cases annually. It is a risk factor for congestive heart failure, and stroke, 75,000 cases of the latter occurring annually in patients with atrial fibrillation. The safety of the most widely used antiarrhythmic agent for this group of patients, quinidine, has been called into question. This study seeks to determine whether two other agents, amiodarone and sotalol, are safe and effective treatments for patients with atrial fibrillation.

Primary Hypothesis: The primary objective is to compare the effects of amiodarone, sotalol, and placebo in maintaining sinus rhythm in patients with atrial fibrillation converted to sinus rhythm.

Secondary Hypotheses: To compare the three therapies in regard to: 1. Frequency of episodes of major and minor strokes. 2. Frequency of episodes of major and minor bleeds. 3. Frequency of sudden death, cardiac mortality, and total mortality. 4. Frequency of life-threatening pro-arrhythmic reactions. 5. Frequency of episodes of congestive heart failure. 6. Frequency of side effects necessitating discontinuation of therapy. 7. Frequency and mean duration of hospitalization directly related to atrial fibrillation or flutter. 8. Mean change in maximal exercise capacity on treadmill during atrial fibrillation or flutter versus sinus rhythm. 9. Time to the development of sinus rhythm from randomization to day 28 of the study. 10. Mean duration of the intervals between occurrences of atrial fibrillation or flutter after day 28. 11. The mean ventricular response documented on electrocardiogram (ECG) recordings during occurrences of atrial fibrillation or flutter after day 28. 12. Changes in health-related quality of life as measured by the SF-36 and an atrial fibrillation quality of life questionnaire. 13. Time to first occurrence of atrial fibrillation or flutter after day 28 or cessation of treatment due to adverse drug reactions after randomization.

Intervention: Patients are randomized to amiodarone (400mg bid for 14 days, 400mg qam and d200mg qhs for 14 days, 300mg qd for 48 weeks, then 200mg qd), sotalol 80mg bid for 7 days and 160mg bid thereafter) or placebo.

Primary Outcomes: The time from day 28 of randomization to first occurrence of atrial fibrillation or flutter. Failure time will be set at 0 days for patients who fail to cardiovert at day 28.

Study Abstract: Atrial fibrillation is the most frequently occurring cardiac arrhythmia, with 1.0-1.5 million cases annually. It is a risk factor for congestive heart failure, and stroke, 75,000 cases of the latter occurring annually in patients with atrial fibrillation. The safety of the most widely used antiarrhythmic agent for this group of patients, quinidine, has been called into question. This study seeks to determine whether two other agents, amiodarone and sotalol, are safe and effective treatments for patients with atrial fibrillation. All patients will have atrial fibrillation continuously for greater than 72 hours. Background medications will include warfarin for anticoagulation and digoxin plus diltiazem or verapamil for heart rate control. If warfarin is contraindicated, left atrial thrombus must be excluded by transesophageal echo (TEE) and aspirin 325 mg QD may be used. Patients will be randomly assigned to receive sotalol (80 mg bid for 7 days and 160 mg bid thereafter), amiodarone (400 mg bid for 14 days, 400 mg qam and 200 mg qhs for 14 days, 300 mg qd for 48 weeks, then 200 mg qd) or placebo. Treatment assignment will be stratified by participating hospital, whether the patient has ischemic heart disease and whether the patient is symptomatic. After randomization, patients will stay on drugs for rate control until sinus rhythm is restored and on anticoagulation until two months after sinus rhythm has been restored. After four weeks, patients remaining in atrial fibrillation will undergo DC cardioversion. Those patients not on warfarin must undergo another TEE within 48 hours prior to cardioversion. Patients will have their heart rhythm monitored transtelephonically every week and occurrences of atrial fibrillation or flutter will be documented twice within 24 hours. In the case of documented atrial fibrillation or flutter occurrence, the patient will be re-anticoagulated and at appropriate time subjected to a further DC cardioversion to restore sinus rhythm. Patients in sinus rhythm will be followed until the end of the study. Patients relapsing into AF will be followed a minimum of one year or until relapse, whichever is later. Assuming 35% of patients on placebo, 50% on sotalol, and 60% on amiodarone remain in normal sinus rhythm at the end of one year, a sample size of 706 patients, 279 on amiodarone, 279 on sotalol, and 148 on placebo (85% power and two-sided overall alpha level of 0.05 for the set of three pairwise comparisons) will be needed for these group differences to be statistically significant.

Inclusion Criteria:

• Patients who have atrial fibrillation continuously for greater than 72 hours.

Exclusion Criteria: