[The need for uncertainty in clinical research. Equipoise].

When a randomized controlled trial (RCT) that has potentially therapeutic benefits for research subjects is conducted, the line between treatment and research is often difficult to draw. However, in order to protect the well-being and rights of prospective research subjects and to ensure that they are not subjected to harmful and pointless research, the distinction between normal care and scientific research must be thoroughly explained. There has to be clear-cut 'equipoise', i.e. uncertainty in physician, patient and investigator regarding the outcome of the study. Strong patient preference for the experimental arm of an RCT does not entail a right to have access to it, as by definition it represents an option of unproven safety or efficacy (equipoise). Only in a therapeutic context do patients have the right to receive the best available care and a right to discuss their treatment preferences with their physicians. A careful informed consent procedure that clarifies the differences between research and therapy, and that emphasizes the experimental nature of trial arms, should be in place prior to randomization.