Given the recent controversy of the SUPPORT study and the upcoming HHS Public Meeting on risks of standard of care research, we thought it would be an appropriate time to review the regulatory requirements regarding research risk.
The federal human subjects regulations (45 CFR 46 and 21 CFR 56) do not define the term “risk.” The regulations only define “minimal risk” as “…the probability and magnitude of harm or discomfort anticipated in the research are not greater in and of themselves than those ordinarily encountered in daily life or during the performance of routine physical or psychological examinations or tests.” [46/56.102(i)] Thus, we can infer that “risk” is the probability and magnitude of harm or discomfort anticipated in the research. The question raised by the current controversy is, “What harms or discomforts should be included in our evaluation of research risks?” For the answer to this question, we can look to the phrase “in and of themselves” in the definition. The inclusion of this phrase would seem to mean that it is the risks introduced into the situation by the research that is the focus, not the overall risks faced by subjects.
All subjects face some risk when they participate in research – life is not risk-free. In order to evaluate research risk we must differentiate between the risks the subjects face independent of the research and the risks introduced by the research. If we are studying cancer treatments, all of the subjects face the risks associated with cancer and cancer treatments. These are not research risks. The risks of the research depend on whether the research design introduces new risks or increases the magnitude or probability of the existing risks. In theory, a research study could be comparing two forms of standard cancer treatment with equal side effects and equal likelihood of success. In this case, the research is the random assignment to one treatment or the other, not the treatments themselves. As this does not really involve much likelihood of harm to subjects, the study could possibly be a minimal risk study.
Another factor involved in evaluating research risk is the requirement that subjects be informed about the reasonably foreseeable risks of the research. Again the question is: what risks do subjects need to be informed about? Following the analysis above, it is the risks introduced by the research, not the risks the subjects would be facing without the research. That is not to say that subjects should not be informed about non-research risks, but that these should be clearly differentiated from the research risks.
While not addressing the specific issues of the current controversy about standard of care research, it is essential that IRBs and investigators fully understand the regulatory requirements regarding research risk and how it applies to the review and conduct of research.