Yonette Thomas, of Howard University, introduced the final session of the workshop that provided a look at the role of institutional review boards (IRBs) from three very different perspectives, that of a university official overseeing human research participants protection, a sociocultural anthropologist who must deal with IRBs in her work, and a research funder.
Lois Brako, of the University of Michigan, Ann Arbor, offered some observations and suggestions about the proposed changes to the Common Rule, based on her experiences with regulatory and compliance oversight at the University of Michigan. She described the university’s human research protections program as one that seeks to maximize its opportunities to be flexible and innovative, and noted ways the university works to take full advantage of the flexibility in the regulations concerning human subjects protection. The university limits the scope of the federalwide assurance by applying it only to federally funded projects. “It’s really important on a campus like ours with so many student projects,” she said. “This allows us to do some local review if we want….Again, our school has tried to take full advantage of this,” she noted.
The university also regulates only research that meets the defini-
tion of human research. In 2012, 172 of the 1,024 new submissions to the university’s health sciences and behavioral sciences institutional review board (IRB-HSBS) were classified as not regulated.1 “We don’t want to over-regulate,” she said. A large percentage of the protocols submitted to the IRB-HSBS—more than 40 percent in 2011 and 2012—are exemptions by IRB staff reviewers. Exemptions do not have to be granted by the IRB itself, she noted; the process can be carried out by educated staff members. The IRB-HSBS is also constantly using and streamlining its expedited review process; in 2012 nearly 40 percent of the new submissions were given expedited review. She added that the IRB-HSBS regularly uses waivers or alteration of informed consent and waivers of documentation of informed consent, particularly for the social sciences, and it has established cooperative research arrangements to avoid duplicate review.
As a result, even though the IRB-HSBS has about 1,200 projects ongoing, the workload of the full committee has been reduced, and the time it takes the committee to make decisions has been significantly decreased, Brako said. For example, the median turnaround time for an exempt determination went from five to six days in the first part of 2011 to one or two days by the end of 2012.2 The median turnaround time for expedited approvals also has dropped, she said, and was at about 14 to 15 days in the second half of 2012.3
The university’s IRB-HSBS has also conducted demonstration projects on minimal risk research. The projects cannot be federally funded or regulated by the Food and Drug Administration, and they must not hold certificates of confidentiality. One demonstration allows two-year approvals, and another created a new exemption category for secondary data analysis with identifiers. “In the first two years of our projects,” Brako said, “we saved ourselves about 1,000 project reviews….We also audited, and we found no problems.”
Comments on ANPRM Proposals
Brako offered several comments about the proposed changes to the Common Rule. The advance notice of proposed rulemaking (ANPRM), she noted, proposed to create a category of “excluded” research that would require only a one-page application and no review before research began; there would be a data security check and a random audit later (Emanuel and Menikoff, 2011). This is similar, Brako said, to a current University
1For details, see http://www.hrpp.umich.edu/Indicators_Report_January_2013_Final.pdf [March 2013].
of Michigan IRB-HSBS process with a short application reviewed by an IRB staff specialist. “So we’re already in a very streamlined mode, with a one- to two-day turnaround time.”
Brako suggested that “registered” might be a better term than “excluded” because “as an institution we’re responsible for connecting up to federally funded projects. We really want to see these projects registered.” She also supported the idea of allowing investigators to determine their own exemption status using standardized tools such as decisions trees or exemption wizards. She said there is some advantage to maintaining institutional screening processes to validate these exemptions but believes it would be preferable to validate them before the initiation of the research. Finally, she said, IRBs should continue to review ethical concerns related to the protection of privacy and confidentiality, but they should be able to rely on institutional resources, such as information technology experts, for the evaluation of data security.
Brako said she sees great advantage to expanding the exemption categories, and she made several suggestions for new exemption categories. These included social networking, the human testing of technology, the analysis of secondary data with identifiers, minimal risk deception research, the collection of data from videos and other recordings, and group characteristics from surveys and interview.
In Brako’s view, new categories should be created for expedited review. An expert panel composed of social scientists and other researchers, IRB members and chairs, IRB administrators, and nonscientific IRB members should be formed to determine what the new categories should be, she added, and the panel should update the list regularly. The list from the Council on Governmental Relations comments on the ANPRM, which suggested new expedited categories for studies of Internet behavior, functional magnetic resonance imaging at standard exposure levels, the establishment of registries for future research purposes, and occupational health activities, such as walking, deep breathing, and mild exercise, could provide a starting point for this work, she noted.4 Brako also suggested letting IRBs use expedited procedures for any other activities they determine to be of minimal risk.
On the subject of the elimination of continuing review (Emanuel and Menikoff, 2011), she said she strongly supports this for qualifying minimal risk studies, but she added that the change should be accompanied by clear guidance and examples of what IRBs would no longer be required to do.
On the issue of a single IRB for multisite research, Brako said that she
4The list is available at http://www.regulations.gov/#!documentDetail;D=HHS-OPHS-2011-0005-0656 [March 2013].
supports a movement to reduce duplicate reviews by multiple IRBs, but that it should not be mandated for all cases. There are reasons to have the flexibility to use more than one IRB, she said.
Brako said she generally agrees with the idea of clarifying and harmonizing regulatory requirements and agency guidance across the agencies. Inconsistencies in guidance from various agencies—she mentioned the Food and Drug Administration, Department of Defense, Department of Justice, Department of Education, Environmental Protection Agency, and the National Science Foundation in particular—weaken human subjects protection by distracting researchers and IRBs from more important considerations. They impede research by slowing the review process and by confusing and intimidating researchers. However, she cautioned against harmonizing by creating a “one-size-fits-all” approach to different types of research, which often results in unfavorable cost–benefit ratios. Instead, she would like to see a single, multiagency regulatory standard that calibrates its provisions to the nature and magnitude of risk that it addresses.
Brako offered comments on the proposed changes to informed consent. “I think just about anyone would agree, that the process of informed consent could be improved,” she said. Specifically, consent documents can be greatly simplified if they focus on descriptions of the research activities, the risks imposed by the experimental component, the potential benefits of the research, and the burdens (e.g., financial, time commitments, alterations in medical care) imposed by participation in the project. Mainly, she said, what is needed for the social and behavioral sciences is flexibility.
Proposed Changes That Would Increase Burden
Some of the proposed changes in the ANPRM are likely to increase the burden for institutions, and Brako identified several that would likely prove particularly burdensome at the University of Michigan. Requiring federal oversight of all human subjects research at an institution that receives federal funding would be overly burdensome, she said. This change would cause her institution to lose a lot of the flexibility that it has taken advantage of in dealing with human subjects protection. Similarly, mandating institutional data security and information protections whenever data are collected, generated, stored, or used would also be burdensome. “I think there are many ways that we can deal with this data issue in a different way,” Brako said.
Expanding the meaning of “human subject” to include biospecimens without identifiers within the ANPRM provisions related to informed risk and requiring written consent for the research use of de-identified specimens would also increase the burden, Brako added. The proposal to require records of adverse events and unanticipated problems to be
submitted and stored in a central database is premature, she said. There is much to learn and understand before such a requirement is instituted, in her view.
In summary, Brako said that while she supports several proposed changes in the ANPRM that clearly reduce burden, she also believes that in some cases clear and concise guidance could accomplish more than changing the regulations in terms of assisting investigators, institutions, and IRBs to better protect research participants. She said that she is particularly concerned that some of the proposed changes seem to shift, rather than reduce, the burden.
Rena Lederman, of Princeton University, discussed how the proposed changes to the Common Rule—and, in particular, the creation of an “excused” category of research that has only informational risks—would be likely to affect anthropologists and other ethnographers. Lederman described the distinctive features of participant observation, which she described as “my discipline’s way of understanding human experience.” This approach to research is quite different from the approach in other areas of science, she explained, which makes it a poor fit for the Common Rule in many ways.
To carry out their fieldwork, Lederman explained, anthropologists immerse themselves in the lives of the people they study, communicating in their language and staying for long stretches of time. The settings are not controlled and thus not particularly suitable for systematic hypothesis testing. “Our goal,” she said, “has generally been to put ourselves in social situations controlled by the people whose circumstances we’re interested in understanding, who are not therefore our subjects, but more accurately our hosts, our interlocutors, our instructors, our consultants, and increasingly our collaborators.” Unlike researchers in most of the social and behavioral sciences, who do not form close relationships with their subjects, ethnographers form relatively “thick” relationships with their research participants. She explained that the researchers are being observed by the participants at the same time that the participants are observing them. These full and complex interactions with research participants are a major source of the understanding that ethnographers derive from their research.
At the same time, these investigator–participant interactions do not lend themselves to the sort of informed consent envisioned by the Common Rule, Lederman said. A consent agreement works like a contract, she said, spelling out the relationship between investigator and
participant “in terms of a socially thin set of expectations and agreements.” Such an agreement is not attuned to the relationship-thick ethical demands of ethnographical fieldwork, she observed.
Problems with Proposed Data Privacy Protections
One of the key changes proposed in the ANPRM would be to allow projects that pose only “informational” risks to be excused from undergoing prior IRB reviews of their research plans, as long as special data privacy protections are instituted before the data analysis is begun, Lederman noted. The ANPRM also proposes to classify the risks posed by most anthropological research as informational. Consequently, Lederman said, it might appear that the new “excused” category would be a boon for anthropologists. Because the new rules place most anthropological research in the informational risk category, she explained, anthropologists would no longer have to submit study designs to IRBs. However, in her view, the current Common Rule’s demand for scientifically rigorous project designs—which, by the nature of ethnographical work, are essentially impossible to specify—undermines the training of students. This provision “encourages the rest of us to offer misleadingly formalized accounts of our prospective work,” Lederman said.
However, she continued, the special data privacy protections envisioned by the ANPRM, based on the Health Insurance Portability and Accountability Act privacy standards, would be equally ill-suited for ethnographical fieldwork. Although anthropologists do traditionally shield the identities of their research participants through the use of pseudonyms, useful ethnographic descriptions require a significant amount of contextual specificity. “De-identifying or anonymizing ethnographic data would render them unusable,” she said.
Furthermore, she added, while data anonymization may be feasible for such relationship-thin research as surveys and experiments, it is theoretically impossible for fieldwork if genuine anonymization means that the investigators themselves are unable to re-identify research participants. The value of ethnographical research lies in the details amassed by researchers in their long-term interactions with the research participants. By the very nature of the work, researchers know the identities of the participants, she explained. Thus de-identifying data before beginning post-fieldwork analysis, as proposed in the ANPRM, would be unworkable for sociocultural anthropologists and other ethnographers.
The effective ethical oversight of sociocultural anthropology, ethnographic sociology, and many other areas of the social and behavioral sciences requires an approach that is not so narrowly grounded in the ethical issues that arise in the biomedical sciences. One useful start, she observed,
would be to tighten up the definition of what is being regulated so as to refocus IRB review more explicitly on biomedical and physical risk. Doing so would enable IRBs to give potentially higher-risk projects the attention they deserve, she suggested. Then alternative approaches could be developed to more adequately address the actual ethical dilemmas of sociocultural anthropology fieldwork and related forms of scholarship: approaches that promote rather than undermine these sciences.
Lederman suggested forming a national commission to develop a rationale and framework for promoting ethical conduct in those areas of scholarship poorly served by or newly excused from the Common Rule. Such a commission might be made up of experts in fields inadequately served by the existing system of oversight, together with legal scholars and philosophers knowledgeable about those fields. To avoid reinventing the wheel, she explained, the commission should be instructed to draw together and build on existing knowledge.
Cheryl Crawford Watson, of the National Institute of Justice, described human subjects protection in research funded by the Department of Justice (DOJ). Because of the sensitive nature of much of that research, which often focuses on criminals and illegal activities, research is subject to additional regulations protecting confidentiality, above and beyond the Common Rule. This creates additional issues for the IRBs reviewing the research, Watson observed.
Any research funded by the DOJ, Watson explained, is governed by a confidentiality statute that not only forbids the release of identifiable information by federal employees and those receiving federal funds, but that also states that such information is immune from legal processes and shall not, without the consent of the person who provided the evidence, be used in any judicial, legislative, or administrative proceedings. Researchers who receive funding from the DOJ who collect personally identifiable information must submit a privacy certificate that describes the research, promises that the researcher will comply with the confidentiality requirements, and describes in detail the procedures the research will use for protecting the confidentiality of the identifiably information collected as part of that research. The procedures for protecting the confidentiality of the information address, for example, the administrative and physical security that will be used to protect the data, who will be allowed access to the data, the information transfer agreements that will be used, and details about the final disposition of the data, including how and in what form it will be archived.
The privacy certificate also includes details about what the research participants will be told during informed consent discussions. Watson added. They must be told that the information they provide will be used only for research and statistical purposes, that the study is voluntary and they can quit at any time, what the study’s risks and benefits are, how the data will be securely maintained, and how the data will be archived or disposed of after the study.
Confidentiality can only be broken with the subject’s consent, and the identifiable data collected for the study can only be used for research purposes with the subject’s consent. “We are a law enforcement agency,” Watson explained. “Few would participate in DOJ-funded research without strong protections on that identifiable data.” The one exception is that the regulations do not apply to information collected regarding future criminal intentions.
This exception can bring a researcher into conflict with state mandatory disclosure laws, she noted. Whereas the certificate of confidentiality required by many other federal agencies allows a researcher to disclose matters, such as child abuse, reportable communicable diseases, or a subject threatening to harm someone, the privacy certificate required by DOJ does not, she noted. Thus, when mandatory disclosure laws are an issue, the researcher must get two different consent forms signed by the participants—a consent to participate and a consent to allow reporting.
IRB Interactions with the DOJ
Because of these DOJ-specific regulations, researchers go through a somewhat different process with DOJ than they would if their research were funded by, for example, the Department of Health and Human Services, Watson explained. IRBs play the same role with DOJ-funded research as they do with research funded by other federal agencies. Extramural research—that is, research funded by DOJ but performed outside the department—is reviewed by the grantee’s own IRB or by a commercial IRB. Intramural research is reviewed by the Office of Justice Programs’ IRB at DOJ.
DOJ research is governed by several federal laws and regulations, including 42 USG § 3789g, which provides for confidentiality of information. DOJ adopted the Common Rule but not its subparts, Watson noted. However, if an IRB or a grantee organization has promised or, in its federalwide assurance has said, that it would follow the subparts, the DOJ requests that it do so.
Researchers who receive DOJ funding need to understand that there are withholding conditions on those funds, Watson said. For the funds to be released, the DOJ must receive both the privacy certificate and the
Human Subjects Protection Form, which provides the grantee’s Federal-wide Assurance number and IRB number and indicates whether the IRB review has already taken place or whether it will take place at some point in the future.
One of the main areas that cause concern is attempts by principal investigators to self-exempt, she said. In some cases, the research truly should be exempt; but in other cases, the research is greater than minimal risk and the researcher has, for unknown reasons, decided that it fits within the exemption criteria.
Another common issue is that IRBs are unaware of DOJ regulations, particularly the confidentiality statute. Watson said she sees many consent forms that specify that the investigator will report child abuse, suicidal ideation, threat of harm to others, and so on. The DOJ will not accept such forms. In other cases, the IRBs are simply confused by the regulations, she added, as they are somewhat different from what IRBs usually encounter. In at least two cases, researchers have decided not to take DOJ funding—in one case after it had already been awarded—because the DOJ will not accept informed consent forms that include mandatory disclosure statements. To address these issues, the DOJ has an effort under way to compile an information packet it will send to IRBs and principal investigators.
Emanuel, E.J., and J. Menikoff. (2011). Reforming the regulations governing research with human subjects. New England Journal of Medicine 365:1145–1150.
This page intentionally left blank.