Statistical problems in the design of antiarrhythmic drug trials.

Although many aspects of design, monitoring, and analysis in trials of antiarrhythmic drugs are similar to those of other intervention studies, several features are different. These include decisions on how to identify the subjects, timing of the intervention with respect to the disease process, time of randomization, methods for measurement of outcome, and interim monitoring for harm or benefit. The resolution of these issues depends almost entirely on the specification of the question of interest.