• Samir Gupta, Marie E Faughnan, George A Tomlinson, and Ahmed M Bayoumi.
• Department of Medicine, University of Toronto, Toronto, Ontario M5B 1W8, Canada. guptas@smh.toronto.on.ca
• J Clin Epidemiol. 2011 Oct 1; 64 (10): 1085-94.

ObjectiveRare diseases may be difficult to study through conventional research methods, but are amenable to study through certain uncommonly used designs. We sought to explain these designs and to provide a framework to assist researchers in identifying the most appropriate design for a given research question.Study Design And SettingWe systematically searched for literature describing rare disease research frameworks, trial designs, and trials that applied them. We present the advantages and disadvantages of each approach using these published examples, and a practical framework to help researchers in selecting between design choices.ResultsWhen research participants are limited, researchers should consider using: 1) a crossover design; 2) n-of-1 trials; or 3) one of the following adaptive designs: a) a response-adaptive randomization design, b) a ranking and selection design, c) an internal pilot design,or d) a sequential design. Bayesian analysis may be applied to conventional designs, or to any of these uncommon designs. Several of these approaches may also be used in combination. The choice between methods should be guided by factors related to the intervention, disease,anticipated recruitment duration and success, and current state of knowledge about the treatment.ConclusionThese techniques may facilitate research in rare diseases.

6 keywords...

hide…