Key safety parameters of sotalol were examined in 1,288 patients entered into recent controlled trials of ventricular (85% of patients) or supraventricular arrhythmias (15%). Most patients were middle-aged male Caucasians with significant heart disease. The most serious adverse event was proarrhythmia, occurring in 56 patients (4.3%). Of these, 27 had hemodynamic compromise due to malignant ventricular arrhythmias. Most had a history of sustained ventricular tachycardia, myocardial infarction, congestive heart failure (CHF) or cardiomyopathy, or a combination of these. The other 29 had nonsevere events; 38% continued taking sotalol. Proarrhythmia was manifested by torsades de pointes in 24 of the 56 patients. No universal causal relation was found with commonly associated factors such as bradycardia, hypokalemia and long QT interval. The mean QT and QTc at baseline within 1 week of a severe proarrhythmic event were greater than those of patients not having proarrhythmia. Nineteen patients (1%) discontinued therapy with sotalol because of drug-related CHF. Predisposing conditions included low initial baseline ejection fraction, history of CHF, cardiomyopathy or cardiomegaly, or both, male gender and age greater than 65 years. Heart failure usually occurred within 7 to 30 days of initiating therapy. The most common reason for premature discontinuation of the drug in patients treated for sustained ventricular tachycardia was ineffectiveness (39%), whereas adverse effects were the most common reasons among patients treated for complex ventricular ectopy (21%). Dyspnea and bradycardia were the most common cardiovascular effects, and fatigue, dizziness and asthenia the most common noncardiac, adverse effects. Although frequently reported, these adverse effects resulted in discontinuation of only 1 to 4% of the patients at risk.(ABSTRACT TRUNCATED AT 250 WORDS)