Regulatory Changes Affecting IRBs and Researchers
Let's face it. Getting IRB approval for research projects falls somewhere on a continuum from "downright painful" to "just get it over with."
The experiences of researchers applying for IRB approval vary considerably, but very few applicants are likely to tell you they find it to be a pleasant experience. There are many legitimate reasons for this: Getting IRB approval is always time consuming; it can be embarrassing when a board asks for changes in procedures; sometimes, a board's judgments can seem arbitrary and ill-informed; and, the bottom line, experienced researchers simply don't like to be told what to do when conducting their research, particularly by individuals who may be unfamiliar with practices within their discipline.
IRBs operate under federal guidelines, but control of decisions to approve research projects is intentionally local. Thus, the approval process and the character and quality of decisions will depend on many factors: the composition of each local IRB, the quantity and types of proposals submitted, the amount of institutional resources provided, and so forth.
You might have observed that the IRB approval process at your institution has become more burdensome over the past few years. Let me explain briefly why this has occurred and what changes we can expect in the future.
THINKING GLOBALLY, ACTING LOCALLY
Over the past few years, a number of factors have come together at the federal level, placing increased pressure on IRBs to tighten their procedures.
All research with human participants has come under increased federal government scrutiny, resulting in more stringent approval procedures at the level of local IRBs. In 1994, in response to concerns that the rights of human research participants were not being sufficiently protected (stemming largely from increases in the scope of biomedical research and clinical trials), then-President Clinton notified all government agencies to carefully review their procedures to ensure compliance with federal guidelines for conducting research.
A year later, the president assembled the National Bioethics Advisory Commission (NBAC) which examined the system for protecting human research participants and identified potential improvements. NBAC's December 2000 draft report included recommendations for a number of sweeping changes in both the structure and process of IRBs. Public comments were solicited and a final published version of the report, Ethical and Policy Issues in Research Involving Human Participants, is scheduled for release this month. More information about the report can be found at the NBAC website; www.bioethics.gov.
Many of the NBAC recommendations simply augment IRB procedures that are already part of the federal guidelines. Others suggest completely new policies. For example, the NBAC's first recommendation is to increase protection of human participants by requiring IRB review of all research involving human subjects regardless of where or by whom it is conducted. As it stands now, research conducted by private individuals, corporations, and other organizations not receiving federal research funds (e.g., physicians, psychotherapists, biotechnology companies, business and industry, some universities) as well as research conducted by some federal departments and agencies is not required to undergo IRB scrutiny and is not regulated by the federal guidelines.
In addition, NBAC suggests that federal oversight of the protection of research participants should be assigned to a single government agency. At present, the system of protection for human subjects is spread across various research agencies within the Department of Health and Human Services (DHHS) as well as in other federal departments, including defense and even agriculture. This results in multiple, sometimes conflicting approaches to research oversight, making it burdensome to find needed information or guidance. However, it should be noted that DHHS's new Office for Human Research Protections is emerging as the lead agency in this system. (More on the new Office below.)
NBAC also recommends that a single, comprehensive federal policy, including more decision guidelines, should be created that applies to all research using human subjects. In addition, the NBAC recommends that all investigators, IRB members, and regulatory staff should be certified by demonstrating appropriate research-related competencies before conducting or reviewing research.
Current federal regulations guide the composition of IRBs in a very general way only. NBAC has recommended the establishment of criteria and standards for selection of IRB members, something that could require changes in the composition of local IRBs. These are sound recommendations for a variety of reasons related both to the quality of reviews an IRB conducts and to the comfort of investigators applying for IRB approval, and many institutions follow some or all of what is recommended already. Some examples are: One quarter of an IRB's members should be unaffiliated with the institution and/or nonscientists; and the diversity in expertise of IRB members should be commensurate with the types of research that the IRB reviews.
A well-constituted IRB will include a sufficient number of members who are knowledgeable in areas of research in which they are likely to be asked to review. When IRBs do not have sufficient and diverse membership, even expedited reviews can take weeks - when they should be able to be completed within a week or two at the most - or may be conducted by individuals not familiar with the research area. For projects requiring full-committee review, federal regulations currently stipulate that an IRB cannot approve projects unless at least half the members are present, including at least one community member, but the number and diversity in competencies of members attending these meetings is extremely important to quality review as well. When there is a diverse mix of IRB members, the discussions are more extensive and more fruitful, and potential risks to participants are less likely to be overlooked.
COMPONENTS OF REVIEW
The two main components of an IRB review are a risk: benefits analysis, and review of informed consent procedures.
Clearly, a risk:benefits analysis requires individuals skilled in research methodology as well as individuals familiar with the research area. It is my belief (and this belief is supported in the federal guidelines for IRBs) that poorly designed research will produce little or no benefits - period - and, in fact, can be harmful. For those who would argue this point, I recommend that you read Robert Rosenthal's excellent article (Rosenthal, 1994); perhaps it will change your mind. Assessing the four categories of potential risks (physical, social, psychological, and economic) requires the right mix of individuals for the proposal under review, but it always requires the presence of people who understand the phenomenology of the research subject's experience.
Approval of informed consent procedures is also best served by a mix of diverse individuals. Here, individuals skilled in the subtleties of the psychological pressures inherent in research participation are clearly needed, but so are nonscientists and community members - just to keep us honest.
NBAC's recommendations would allow more flexibility for investigators to obtain and document informed consent and for IRBs to waive informed consent. The recommendations also would eliminate the need for continuing reviews of minimal risk research. They also would allow for a single IRB to review research conducted at multiple sites.
If implemented, the NBAC recommendations will bring about many changes in the structure and procedures of IRBs; some will be expected to increase the workload of IRB members; others should decrease it. But, clearly, the NBAC recommendations are calling for more training and accountability for IRBs, institutions, and investigators.
The immediate future of the NBAC report is somewhat indefinite, only because the panel was convened under the Clinton Administration, and its charter expires in October 2001. NBAC will transmit its recommendations to the National Science and Technology Council, which also may or may not continue to exist under the Bush Administration. But the report represents the most comprehensive assessment of the human subjects protection system to date, and NBAC's recommendations are likely to have a significant impact on the policy discussions around these issues.
Some important changes in research oversight have already occurred. In the Fall of 2000, a new organizational structure - the Office for Human Research Protections (OHRP) - was created within DHHS to increase the visibility and status of regulatory oversight. Many of the issues raised below are explained in depth at the OHRP website: ohrp.osophs.dhhs.gov
. Readers are encouraged to visit that site to learn more about the changes in federal policies and requirements affecting IRBs.
Among the changes instituted by OHRP is that all institutions engaging in research supported by federal funds must submit new federal-wide assurances (FWA) of compliance with federal guidelines, and satisfy new educational and other requirements to continue eligibility to receive government grants. It should be noted that as part of their FWA, research institutions agree to oversee funded and nonfunded research in the same manner, following the federal guidelines. Therefore, new requirements for individuals seeking federal funds will also affect nonfunded researchers at these institutions.
In addition, all investigators conducting NIH-sponsored research using human participants are now required to document that they have received training in the protection of human subjects. Even though the policies for protecting human participants have been strengthened, the requirements for human subjects investigators and IRB members remain less stringent than those of many other regulatory compliance boards, such as those overseeing radiation safety, biosafety, and animal research.
At the same time, concerns about scientific misconduct led the Office for Research Integrity (ORI; also part of DHHS), which oversees institutional investigations of misconduct in research, to issue an extensive and controversial Policy on Instruction in the Responsible Conduct of Research (RCR); for more information about RCR, go to www.ori.hhs.gov/.
Implementation of RCR would require education of all researchers, IRB members, and institutional regulatory staff in nine core areas, including: data acquisition and sharing, mentor responsibilities, responsible authorship, peer review, scientific misconduct, and conflict of interest, as well as ethical guidelines for research with human and animal subjects.
The goal of the policy is to ensure that all persons engaged in research and all regulatory personnel involved in local oversight would receive training in relevant areas of the research and communication process. ORI believes that such education would result in fewer incidents of scientific misconduct.
The policy is not without its critics. The extent of the education involved has been considered impractical or excessive by some scientists and research institutions. Originally, RCR was to be implemented in research institutions by October 2001, but the deadline was deferred until October 2003 to give institutions more time to comply.
At the present time, implementation of RCR has been suspended because of Congressional concerns about statutory procedures. ORI continues to review the policy and to develop RCR education programs, but final resolution awaits the appointment of a new Assistant Secretary for Health.
Despite suspension of the policy, a number of government agencies, such as the Center for Disease Control and NIH, have already begun to implement RCR for their intramural researchers. It is not yet known whether the Bush administration will favor nationwide implementation, but clearly, the direction of all of these programs and activities seems to be toward more government oversight of human subject protections and more stringent requirements for researchers and IRBs.
Most IRB members, particularly in universities and hospitals, are unpaid volunteers who generously donate time from their busy schedules to provide a valuable service. As requirements for selection and training of IRB members become more onerous, it is going to be more difficult than ever to recruit and retain excellent people. The downside for researchers is that research may become more expensive for all of us if we are forced to move to a system of "professional" IRBs. Such IRBs exist already, and it can cost from $300 to $3500 for each proposal review-and there is no guarantee that a professional IRB would conduct better or more expeditious reviews.
Rosenthal, R. (1994). Science and ethics in conducting, analyzing, and reporting
psychological research. Psychological Science, 5(3), 127-134.
CHRISTINE HANSEN received her PhD from Michigan State University in personality and social psychology. She is the co-author of Myers and Hansen's Experimental Psychology (5th Edition), published by Brooks/Cole, and her research has been supported by grants from the National Science Foundation and NIH. Currently, she is an Associate Professor at Oakland University and serves as the chair of their IRB.