Innovative strategies for trial design.

Most clinical trial designs used in evaluating therapies require large numbers of subjects. Childhood rheumatic diseases are rare and large numbers of subjects are not available. We have largely been unsuccessful in developing evidence based therapeutics for pediatric rheumatology. Because these diseases are rare, it is important for each clinical trial to enroll all possible subjects, and it is important to extract as much information as possible from each subject. Placebo controlled randomized studies are sometimes seen as unacceptable to physicians and patients. We need study designs that are more acceptable and that will increase accrual. Some clinical trial designs, especially those that involve crossover treatment assignments, can increase the amount of information that each subject contributes. These types of studies can only be used for treatments that are reversible. For treatments that may permanently affect the disease course, I propose the randomized placebo phase design as one way of increasing study acceptability and accrual.