Design, results, and interpretation of randomized, controlled trials in congestive heart failure and left ventricular dysfunction

This article discusses how randomized trials can be designed in patients with heart failure and left ventricular dysfunction to provide reliable answers regarding various interventions. Most previous trials have been limited by their small size, short duration, and restricted entry of highly selected patients, leading to unconvincing answers. The effects of treatment on surrogate end points (e.g., exercise tolerance, arrhythmia suppression) do not necessarily correlate with the ability of an intervention to reduce mortality or morbidity. Therefore, promising results obtained on surrogate end points should be complemented by data from further studies using clinical outcomes. Recent experiences from major trials in heart failure and other cardiovascular diseases indicate that if such trials are large, they are likely to provide reliable information on clinically important outcomes even if only limited data are collected on individual subjects. The results of some of the larger trials are summarized, and their implications for both study design and clinical applications are discussed.