Objective: We aim to clarify the circumstances in which randomized, controlled trials should be designated as minimal risk, allowing institutional review boards to approve their conduct with a waiver of informed consent if obtaining informed consent is not feasible.
Methods: An ethical analysis of the minimal risk standard as applied to randomized, controlled trials was conducted.
Conclusions: In determining whether an randomized, controlled trial should be designated as minimal risk, the potential sources of risk that must be considered are as follows: physical risk from study treatments, the loss of individualized care, risk from nontherapeutic components of the research protocol, and the psychological impact of participation, particularly if the research takes place without informed consent in an emergency setting. The risks of research participation should be considered in comparison with the risk of nonparticipation; e.g., the risks specific to research participation should be considered separately from the risks inherent in treatment of the potential research participant's underling condition. Participation in an randomized, controlled trial may pose no more than minimal risk when: 1) genuine clinical equipoise exists; 2) all of the treatment options included in the research study fall within the current standard of care; 3) there is no currently available treatment with a more favorable risk-benefit profile than the treatments included in the study; 4) the nontherapeutic components of the research are safely under the minimal risk threshold; and 5) the research protocol provides sufficient latitude for treating physicians to individualize care when appropriate. The potential for research participation to have a negative psychological impact on participants or their families should be considered in risk assessment. The requirement for informed consent should only be waived to the extent necessary, and opportunities for the research participant or surrogate to decide whether to participate in the research should be maximized.