Problems with minimal-risk research oversight: a threat to academic freedom?
|Author:||Rasmussen, Lisa M.|
|Publication:||Name: IRB: Ethics & Human Research Publisher: Hastings Center Audience: Academic Format: Magazine/Journal Subject: Health Copyright: COPYRIGHT 2009 Hastings Center ISSN: 0193-7758|
|Issue:||Date: May-June, 2009 Source Volume: 31 Source Issue: 3|
Although criticism of a powerful bureaucratic body is expected,
your local institutional review board (IRB) might deserve a touch of
sympathy, having been subjected to withering critique in recent years
for a long list of sins. The observed faults of IRBs have been well
summarized (1) and include inconsistency, delay, grammatical pedantry,
excessive conservatism regarding legal risk, ignorance of fields
reviewed, and threats to academic freedom. (2) Minimal-risk research is
a particular area of controversy because the bureaucratic burden of
human subjects research oversight seems severely disproportional to the
potential risks of harm to research participants, as well as to the
effort required to conduct the research itself.
Many of these complaints are legitimate, and IRB review of a researcher's proposed study can engender anger, frustration, and stress. Unfortunately, such responses can impede an accurate assessment of both the cause of the problem and its potential solution. A particularly unfortunate recent development is the labeling of some legitimate complaints as issues of academic freedom when they are not.
I will not settle here the fundamental issue of whether a convincing argument exists that IRB review poses a threat to academic freedom. Instead, I evaluate the arguments in a report issued recently by a committee of the American Association of University Professors. The report, "Research on Human Subjects: Academic Freedom and the Institutional Review Board," (3) was issued by Committee A on Academic Freedom and Tenure, which is the largest and oldest of the AAUP standing committees. In the report, Committee A argues that the federal regulations governing research with humans--known as the Common Rule--grant unchecked power to IRBs in a way that poses a threat to academic freedom. It is vital to phrase complaints regarding IRBs accurately, both to avoid diluting the importance of academic freedom and to have a reasonable chance of solving the problems that do exist. I argue here that Committee A does not make a strong case that IRB oversight of human subjects research poses a threat to academic freedom, and I offer a better explanation and resolution of the problem that Committee A identified.
Assessing the alleged threat that IRBs pose to academic freedom requires a working definition of the concept of academic freedom. However, no universal definition exists. Conrad Russell points out that the notion of academic freedom began as a way of asserting the independence of the early church from the government and had mainly to do with the protection of a sphere of "independent professional judgment, within which [the government] is incompetent to meddle"--in this case, spiritual values. (4) The notion was later expanded:
This conception of academic freedom suggests that the way to protect professional autonomy is instrumentally, through liberties such as those recognized in the United States' Bill of Rights and its ancestors. No wonder, then, that academic freedom is similar to-and often conflated with-freedom of speech; but over time, the concept has shifted in meaning from a focus on the rights of institutions to the rights of individuals within institutions. Yet another interpretation of academic freedom is a consequentialist claim that granting it tends to bring about better results. (6) Which of these interpretations one subscribes to makes a dramatic difference in how one views federal involvement in research as impacting academic freedom.
Committee A does not offer an explicit definition of academic freedom, but it is reasonable to rely in the present essay on the AAUP's "1940 Statement of Principles on Academic Freedom and Tenure":
The most salient aspect of this passage in relation to human subjects research is the recognition that academic freedom is granted "for the common good," not because of an inherent right of researchers. Though this idea has not been sufficiently explored in the literature on IRB oversight, (8) this seems to place a conceptual limit on the extent of academic freedom. Consequently, there is a prima facie claim that research can be subjected to assessment regarding whether it threatens to harm the common good via harm to individuals. If this interpretation is correct, the simple assertion that government involvement in research is a threat to academic freedom proves nothing.
Granted, this is potentially the thin edge of a perilous wedge: as soon as oversight of research is justified in the name of the common good, it is not clear where the federal reach might end. However, the history of human subjects research has demonstrated that individuals can be and have been unjustly harmed by research. We must therefore acknowledge the failure of self-regulation and so accept this thin edge. But further intrusion must be fought, and fought precisely. Heedlessly claiming threats to academic freedom trivializes potential harms to human subjects, while denying that such threats exist risks a dangerous incursion into a cornerstone of academic research. Committee A is guilty of the former.
The Committee's Claims
Committee A describes in a perfunctory way how IRB oversight might threaten academic freedom, leaving it to the reader to construct an argument on the basis of a few hints. This is unfortunate, as the claim could adhere to any frustration readers may have with IRBs, thus elevating their complaints from matters of procedure to matters of principle. Under this dictum, any impediment that IRB oversight poses to research can be classed as censorship. Such broad, undifferentiated claims do a disservice both to the principle of academic freedom and to IRBs. The stakes are high: the protection of both research subjects and academic freedom is fundamental to the research enterprise, yet protecting both may cause conflict. But nowhere in the report does one find an explicit, sustained argument that IRB oversight is a threat to academic freedom. Instead, there are the following hints: (9)
* Unchecked power granted to IRBs has resulted in a number of more or less familiar horror stories. This section includes a series of unelaborated anecdotes with no documenting citations (Section 1).
* No provision is made in the regulations for an appeal process when an IRB disapproves a research protocol (Section 1).
* There has been longstanding concern about the application of the federal regulations to research methodologies that present no serious risk of harm to research subjects (Introduction).
* The federal regulations instruct IRB members to form their own view of potential research risks and the importance of possible results obtained from the research, and to deny permission to conduct the research if they believe the risks are not reasonable relative to the value of the likely results. There could hardly be a more obvious potential threat to academic freedom (Section 1).
With some generosity, the following four arguments regarding the source of the threat to academic freedom can be developed from these hints: 1) IRB incompetence, as illustrated in certain "horror stories," threatens academic freedom; 2) the lack of an appeals process when an IRB disapproves a protocol threatens academic freedom; 3) federally regulated oversight of minimal risk research threatens academic freedom; and 4) academic freedom is threatened by the fact that IRB members' assessments of the balance of risks and benefits carry the authority to prohibit research.
There are several reasons why none of these arguments succeed in demonstrating a threat to academic freedom. A longer explanation of the failures is beyond the scope of this paper, but a brief outline is possible. First, while the examples of IRB incompetence or unreasonable IRB requests to researchers may well be an understatement of the problem, (10) granting this point does not commit one to equating inconvenience and hassle with abridgement of academic freedom. IRBs can function well or poorly, and which is true for a given IRB depends on many factors, not least of which are institutional support and member training. Second, though the Committee correctly points out that there is no regulatory requirement for an appeals process when an IRB disapproves a protocol, a prudent solution would be for institutions to establish a separate appeals process for researchers whose protocols were not approved. The point is that while lack of an appeals process may threaten academic freedom, it is distinct from the claim that IRB oversight in general threatens academic freedom.
The third possible threat to academic freedom that can be constructed from the Committee's report is the fact that research involving minimal risk to human subjects is overseen at all. The claim is that applying the regulatory minimal-risk standard to all research with humans is misplaced, especially when certain methodologies are involved (e.g., research on autonomous adults using surveys or interviews). According to human research regulations,
If research is of minimal risk, the argument might go, it incurs risks accepted every day without comment, and there is no justification, on grounds of protecting human subjects, for requiring IRB oversight of research.
An obvious first step in responding to the argument above is to question whether research using the methodologies the report identifies does in fact always constitute minimal risk. (12) However, one might simply grant that such research is of minimal risk and argue that it should not be overseen for a variety of reasons that have nothing to do with academic freedom. For example, an oral historian employed by an academic institution will likely be required to submit to her IRB a research proposal to interview several individuals in order to determine if the study meets the regulatory definition of human subjects research. However, a journalist working outside of an academic institution is free to conduct the same interviews without any intervention by an IRB. (13) If minimal risks are accepted in one area, one might argue, they ought to be accepted in the other, since they are by definition risks we are all exposed to daily and tolerate without comment. There is no need to appeal to academic freedom in order to make this argument.
Fourth, notwithstanding the Committee's assertion, the source of the threat to academic freedom via oversight by one's colleagues is far from clear, especially since researchers undergo peer review for research funding and when submitting their manuscripts for publication. Thus, the heart of the Committee's concern with oversight by colleagues who are members of the IRB seems to be the mere fact that IRBs have the authority to prohibit research. It is hard to know what to make of this concern. The reason for the IRB's existence is to protect research participants, which means that an IRB must have the authority to prohibit research if the protection is to have force. Committee A seems to recognize that both oversight simpliciter and oversight with the authority to prohibit research are legitimate features of human subjects research. "It is arguable," the report says, "that the IRB system that was developed over the years has been effective in protecting the subjects of biomedical research that would otherwise impose a serious risk of harm on them." (14) Similarly, only six years earlier, the AAUP published the report, "Institutional Review Boards and Social Sciences Research," in which it claimed that
Thus, oversight--even with the power to deny research--does not seem to be the Committee's main objection to IRB review. Perhaps the claim is slightly different. In yet another publication, the AAUP explains that the "mere existence of the requirement that IRBs evaluate the risks of the research in relationship to its importance can have an inhibiting effect on the work of scholars." (16) The difference between peer review and IRB review, it suggests, is that behind IRB review is "the weight of the government and the specter of the official control of opinion." (17) This would be a powerful argument if review worked in the way implied in this passage. However, except under very specific circumstances, (18) it is not "the government" that reviews research protocols. Although the federal government has established legal rules for conducting research with humans, it is an institution's IRB, not the federal government, that determines whether an activity meets the regulatory definition of human subjects research and then approves or disapproves the research protocol. Although the federal Office for Human Research Protections (OHRP) ensures that IRBs function properly, keep appropriate records, maintain a quorum at meetings, and comply with other regulatory provisions, the "specter of the official control of opinion" is a hyperbolic claim. The fact that the federal government sets policy does not mean that policy is a mechanism for censoring researchers. Moreover, it may not be either desirable or even possible to extend the defense of academic freedom to the degree that it protects researchers from self-censorship (if that is how we understand "an inhibiting effect"). (19)
The Committee's Recommendations
Committee A makes two recommendations that purport to solve the problems identified in its report. First, the Committee suggests that institutions make use of the existing option not to subject research that is not federally funded to IRB oversight. Second, it argues for a change in federal regulations so that some federally-funded research specifically singled out in the regulations as not eligible for exemption be made exempt:
Obviously taking oversight away from federally-mandated bodies would solve the alleged academic freedom problems, but at what risk? Committee A recognizes that breach of confidentiality and its concomitant harm is still a possibility in research without federally-mandated oversight. The Committee's solution to this risk is to suggest that alternative means of protecting confidentiality-such as departmental review of research protocols in lieu of IRB review-would suffice. This is not the only proposed alternative review model, but it is an obvious candidate, so it is worth discussing the problems with such a model.
Even if one remains convinced that research oversight poses a threat to academic freedom, departmental review arguably poses a greater threat than the IRB model. For all their flaws and ponderousness, IRBs are collectively uninterested in specifics. It would be difficult for one person with an agenda of destruction to leverage the power of the IRB to that end. On the other hand, departments may be more prone to personal conflicts or competition than are IRBs, whose members are less likely to work together, engage in similar research, or see each other more than occasionally. Anyone whose tenure or advancement has been held hostage to the whim or dislike of even one colleague can vouch for the power seated in departments. Departmental review can squelch new methods, too, because departments often have a dominant program of research that may exclude a junior faculty member's innovations.
There are other problems with departmental review that are pedantic, yet real. First, who in a department would have the responsibility to review interdisciplinary research or research unrelated to the department's academic discipline? Second, we have only the Committee's assertion that departments know how to keep data confidential. There is a difference between the principles of blind trust and "trust but verify." Regulations would undoubtedly mandate the latter, requiring a new layer of bureaucracy to ensure that all departments (in addition to existing IRBs) follow best practices to protect the confidential information of research subjects. Third, unlike IRBs, which operate under the auspices of the university, departments are not capable of offering legal representation to principal investigators in the case of a lawsuit or an attempt to subpoena confidential information. (21) Fourth, unlike departmental review, IRB review would protect institutions. Though it is never popular to make this point, we live in a litigious society, and that warrants taking precautions. In the area of research, institutions want IRB review of protocols because that affords whatever legal protection such oversight can convey. The point is not to dodge under a legal cover, but rather to be able to demonstrate publicly that one was duly diligent in protecting human subjects of research. Fifth, if the current mechanism of IRB review is thought to be time-consuming, the alternative may be worse. Department members at most institutions are already beleaguered with meetings, committee work, student mentoring, etc. (not to mention their own teaching and research). There is no reason to assume that their reviews will be more timely or less prone to general human incompetence, and reason to suspect that departmental review will not provide other important advantages that IRB review does. Sixth, a deeper point is that IRB review is not simply about reviewing the treatment of confidentially collected information in research. As sociologist Jim Thomas points out in a defense of IRBs, "They address the fundamental processes of inquiry." These include "data acquisition, management, sharing, and ownership," "mentor/trainee relationships," "publication practices and responsible authorship," and general research misconduct. (22)
Finally, IRB membership offers a diversity of perspectives (not least those generated by gender and ethnicity differences) that might be difficult to match within a department. There is also an important regulatory requirement that IRBs must have at least one member who is not affiliated with the institution (typically referred to as a community member). The most profound question about a research protocol frequently comes from the nonaffiliated IRB member, and it is exactly this institutional distance that would be lost if a departmental model of review were implemented. Perhaps more importantly, diversity of membership provides a bulwark against the cronyism that is possible within departments.
None of this is to suggest that department members are incapable of conducting any of these tasks, particularly since IRB members come from academic departments. However, there is more to research oversight than simply ensuring that confidential information will be kept "locked up." Yet Committee A's solution to the problems it identifies seems to leave us with only two choices: eliminate IRB review of minimal-risk research, or charge academic departments with some of the IRB's duties and eliminate the rest. Assuming the foregoing arguments have shown that neither option is satisfactory, are there other possible review mechanisms that would speed the process of protocol review, eliminate possible threats to academic freedom, and protect human subjects against the possibility that novel strategies might expose them to risk?
There may be higher-order solutions to the problems Committee A identified. Some universities already have separate social science IRBs equipped to evaluate the true risks of particular research. Expedited and exempted research can be conducted rapidly if the institution provides the type of resources needed to support a robust IRB office that would include such a body.
A proposed "oversight triage system" for review of quality improvement (QI) research might work for the social science and humanities research that Committee A is concerned about (obviously the details would differ depending on discipline, but a similar type of system would work in both of these research contexts). The proposed QI triage system would involve an existing institutional body quickly determining whether activities that collect health care information "for purposes beyond responding to the needs of individual patients" meet the regulatory definition of human subjects research or are exempt from the regulations. (23) Because triage is meant to be rapid, a few IRB members could be charged with the responsibility of conducting review of protocols within a short period of time. This would mean that social scientists and humanities faculty would either need to be IRB members themselves or at least help educate members about the typical methods used in their research.
Another alternative is what I call a "template review," combined with a researcher's agreement to use and be held accountable to the terms of the template. Disciplines at the national level might formulate templates to guide very common research approaches. For example, a research template for oral historians could stipulate that the researcher will interview individuals, record their answers, refer them to counselors if the questions have provoked strong emotions, procure consent forms, lock the transcripts securely, and identify what will happen to the transcripts at the close of research. IRBs at individual institutions would review the template once and approve it (or even decide to accept any templates from given professional societies). Thus, a researcher would simply submit a form to the IRB stating her agreement to abide by the format of the template. Upon receipt of the form, the IRB would approve the protocol.
What Committee A identified are ongoing problems with the IRB review system that can cause serious inconvenience for researchers, but do not constitute threats to academic freedom. To find the most effective solutions to these problems, we must first accurately diagnose them and then remedy them accordingly. These remedies may well differ from the Committee's recommendations.
(1.) Emanuel EJ, Wood A, Fleischman A, et al. Oversight of human participants research: Identifying problems to evaluate reform proposals. Annals of Internal Medicine 2004:141(4):282-291.
(2.) American Association of University Professors, Committee A. Research on human subjects: Academic freedom and the institutional review board, 2006, http://www.aaup.org/AAUP/comm/rep/A/humansubs.htm.
(3.) See ref. 2, American Association of University Professors 2006.
(4.) Russell C. Academic Freedom. London, U.K.: Routledge, 1993,
(5.) See ref. 4, Russell 1993, p. 4.
(6.) See Streiffer R, Academic freedom and academic-industry relationships in biotechnology. Kennedy Institute of Ethics Journal 2006;16(2):129-149.
(7.) American Association of University Professors. 1940 statement of principles on academic freedom and tenure, http://www.aaup.org/ AAUP/pubsres/policydocs/contents/1940statement.htm.
(8.) See ref. 6, Streiffer 2006, for discussion of academic freedom in a related context.
(9.) See ref. 2, American Association of University Professors 2006.
(10.) Emanuel et al., for example, cite a long list of structural, procedural, and assessment problems with IRBs, none of which are alleged to be a threat to academic freedom. See ref. 1, Emanuel et al. 2004.
(11.) Department of Health and Human Services. Protection of Human Subjects. 45 CFR 46.102w.
(12.) Examples given in contrast with this claim that also meet the description of research advanced by the report, include Humphreys L. Tearoom Trade: Impersonal Sex in Public Places. Chicago, IL: Aldine Publishing Company, 1970; and Thomas J, When cyber-research goes awry: The ethics of the Rimm "cyberporn" study. The Information Society 1996:12(2):189-197, http://venus.soci.niu.edu/~cudigest/rimm/ go.jt2.
(13.) It is worth noting the many levels of oversight that journalists nevertheless experience from their editors, publishers, and risk management departments.
(14.) See ref. 2, American Association of University Professors 2006.
(15.) American Association of University Professors. Institutional review boards and social science research, 2000, http://www.aaup.org/ AAUP/comm/rep/A/protecting.htm.
(16.) American Association of University Professors. Protecting human beings: Institutional review boards and social science research. Academe 2001^7:3, http://findarticles.com/p/articles/mi_qa3860/ is_200105/ai_n8939770/pg_1.
(17.) See ref. 16, American Association of University Professors 2001.
(18.) See Department of Health and Human Services. Protection of Human Subjects 45 CFR 46, subpart D, for the exception in federal regulations that would allow direct federal review of protocols for some research involving children. This seems to be a fairly rare occurrence; Wendler and Varma point out that as of 2005, only 17 protocols have ever been recommended for such reviews. See Wendler D, Varma S. Minimal risk in pediatric research. Journal of Pediatrics 2006;149:855-861.
(19.) One ought to bear in mind here other possible instances of self-censorship, such as the reluctance to publish criticisms of one's department colleagues. These are probably not best construed as academic freedom issues but rather, for example, choices to weigh collegial relations more heavily than the need to publish a particular argument.
(20.) See ref. 2, American Association of University Professors 2006, section 2.
(21.) This point is made by Bruce Bridgeman, IRB chair at the University of California, Santa Cruz, in a response letter to the report (Bridgeman B. Letter to the editor. Academe 2006;92(1):6.) The AAUP report's authors responded in the same column. The possibility of legal appropriation of research records is also discussed in reference to the Brajuha case. See ref. 12, Thomas 1996; Thomas J. "Big brother" or allies? In defense of IRBs and RCR. SSSI Notes 2001, http://venus.soci. niu.edu/~jthomas/ethics/sssieth.html. It is not clear how strong a point this is. Perhaps a university could choose to offer legal representation to researchers in alternative models as well, but it may not be obliged to do so.
(22.) See ref. 21, Thomas 2001.
(23.) Grady C. Quality improvement and ethical oversight. Annals of Internal Medicine 2007;146(9):677-678, at 678.
* Lisa M. Rasmussen, PhD, is Assistant Professor, Department of Philosophy, University of North Carolina Charlotte, Charlotte, NC.
Lisa M. Rasmussen, "Problems with Minimal-Risk Research Oversight: A Threat to Academic Freedom?" IRB: Ethics & Human Research 31, no. 3 (2009): 11-16.
So great a power of ideas, facing so great a power of the sword, could only survive if it established a very firm taboo on State interference in its own affairs in any shape or form. The Church, then, believed it could only preserve its liberty by preserving its liberties. (5)
Institutions of higher education are conducted for the common good and not to further the interest of either the individual teacher or the institution as a whole. The common good depends upon the free search for truth and its free exposition. Academic freedom is essential to these purposes and applies to both teaching and research. Freedom in research is fundamental to the advancement of truth ... Teachers are entitled to full freedom in research and in the publication of the results, subject to the adequate performance of their other academic duties. (7)
Minimal risk means that 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. (11)
a university's effort to ensure that all researchers comply with its human-subject regulations does not offend academic freedom.... the aim of reducing risks to human research subjects does not itself endanger academic freedom, and its abandonment would yield nothing positive for the freedom of research. (15)
[W]hat we recommend is ... that research on autonomous adults whose methodology consists entirely in collecting data by surveys, conducting interviews, or observing behavior in public places, be exempt from the requirement of IRB review-straightforwardly exempt, with no provisos, and no requirement of IRB approval of the exemption. (20)
|Gale Copyright:||Copyright 2009 Gale, Cengage Learning. All rights reserved.|