Addressing low-risk comparative effectiveness research in proposed changes to US federal regulations governing research.

ON JULY 25, 2011, THE DEPARTMENT OF HEALTH and Human Services published an advance notice of proposed rule making, outlining potential changes to the federal regulations overseeing human subjects research. These regulations, known since 1991 as “the Common Rule,” have been effective in ensuring that most human research in this country is reviewed prospectively by an institutional review board (IRB) charged with determining that risks of proposed research are minimized and acceptable and also ensuring that individual participants provide informed consent for most types of research before participation. Some of the proposed changes suggested in the Advance Notice of Proposed Rulemaking (ANPRM) are designed to expand or deepen federal protections, whereas others are intended to reduce the oversight burden on investigators and IRBs for lower-risk research. These proposed changes, however, do not suggest any specific regulatory changes that would affect comparative effectiveness research— specifically, the increasing body of work in comparative clinical effectiveness research (CCER) that, as described by the Patient Centered Outcomes Research Institute, compares the relative effectiveness and safety of alternative preventive, diagnostic, or treatment options. Such CCER studies often involve comparing widely used, US Food and Drug Administration (FDA)–approved therapies or diagnostic tools, with the goal of assisting patients and clinicians in making health care decisions. Other highpriority CCER studies assess the health effects of clinical practices that have been widely adopted by clinicians, despite limited evidence about the risks and benefits. As highlighted in a federal call for grant applications, significant advances in CCER will depend on reducing the intensity and burden of oversight for human research participants in prospective clinical studies that compare the benefits and risks of interventions in common clinical use. The majority of the proposed changes in the ANPRM are sensitive to this concern that regulatory protections should not be overly burdensome. These changes seek to identify categories of research requiring less oversight, thereby “reducing burden, delay, and ambiguity” while allowing IRBs to focus on studies that “could seriously harm subjects.” A significant proposed change is to “excuse” from IRB review altogether all survey, interview, and focus group research as well as, potentially, what the ANPRM describes as “social and behavioral research . . . involv[ing] . . . benign interventions . . . for which prior review does little to increase protections to subjects”—eg, some low-risk behavioral and social science interventional research. The ANPRM notes that IRBs “have a tendency to overestimate the magnitude and probability of reasonably foreseeable risks”; thus, eliminating these studies, which typically pose minimal risk, from IRB review allows committees to concentrate on categories of research posing greater ethical concern. As noted in the ANPRM, this is consistent with recommendations from the Institute of Medicine that ‘‘The degree of scrutiny, the extent of continuing oversight, and the safety monitoring procedures for research proposals should be calibrated to a study’s degree of risk. Minimal risk studies should be handled diligently, but expeditiously, while studies involving high risk should receive the extra time and attention they require.’’ The failure of the ANPRM to address CCER specifically in its discussion of expedited or what the notice proposes calling “excused” research serves to perpetuate the view that all clinical research, of which clinical comparative effectiveness studies are a variant, involves more than minimal risk. In so many types of CCER, however, studies generally pose no or only minimal additional risks or burdens to patients over what patients would experience in clinical care. Although comparative effectiveness or patient-centered outcome studies using retrospective data analysis already are subject to minimal oversight, it seems that many prospec-