General design issues in efficacy, equivalency and superiority trials

Critical discussion is offered of design issues pertinent to three categories of clinical trials: efficacy trials, in which an experimental intervention is compared to an inert control, and equivalency and superiority trials, in which an experimental intervention is compared to an active control. The issues discussed are sample-size determination and power, problems in designing and analyzing multicenter trials, broad versus concentrated sampling of patients, the need (or not) for a placebo control group in an active control trial, the possible appropriateness of crossing patients over from one regimen to another, various methods of randomization, the intention-to-treat principle in repeated measurements studies, and possible logical conflicts between inferences from superiority trials and inferences from equivalency trials.