irbs, mission creep, and prior restraint

I am beginning a new thread here to avoid threadjacking the other conversation going on about the relationship between COI and IRBs.

Fabio writes:

How far do we let IRBs go before we actively resist? If the IRB decided they needed to see my medical records, should I just give it to them?


IRBs were initially designed to manage risk related medical research. Now they try to govern social science research as well (which is ok in my view). But they also find it fit to judge whether something is real science, whether people can be paid by researchers, and all sorts of other activities. …. The institution of the IRB has grown enormously in ways that were not anticipated by anyone.

Here’s my view: the IRB system has grown way beyond its proper role, to the point that it routinely and actively exercises prior restraint over appropriate, legitimate research for which responsibility should be borne and accepted by the investigator. This is largely because IRB members and administrators tend to think extremely expansively about the concepts of “risk” and “harm.” I have faced major delays with the IRB here at UNC over the years because they thought that people who wrote letters to the editor with the intent to publish could face embarrassment if their words were disclosed; and, more recently, because survey respondents might be embarrassed if their opinions about sensitive topics (e.g., abstinence-only sex education and same-sex marriage) were to get out. My response on the latter: if they’re embarrassed about their opinions they should consider adopting new ones! I do not think that making people feel sad or uncomfortable ought to constitute a risk of harm.

Analytically, I think the likely explanation for this expansion is institutional isomorphism: generations of academics are trained to think about harm to subjects as the third rail of scientific work. Universities engage a CYA strategy to manage perceived legal risks, which gives rise to a growing bureaucracy and field of IRBs. I think it’s extremely likely that the COI process will expand similarly.

AAUP’s position on this is, in my view, appropriate:

we recommend that research whose methodology consists entirely of collecting data by surveys, conducting interviews, or observing behavior in public places be exempt from the requirement of IRB review.

Note that this means excused from review, not just covered under an exemption that the IRB or its staff must grant!

It’s important to recognize that excusing research from IRB oversight would not imply that human subjects may be harmed. Rather, low-risk research should be conducted ethically on the part of the investigator, and investigators should be liable for unethical behavior and subject to sanctions when they commit it. But as in the regulation of speech, prior restraint should be reserved for those cases where the potential harm would be grave and irreversible.

Author: andrewperrin

University of North Carolina, Chapel Hill

5 thoughts on “irbs, mission creep, and prior restraint”

  1. I know Sarah Babb is doing some research on IRBs in the social sciences but I don’t whether if pertains to these questions. Has anyone read anything from her?


    1. There is considerable variation in how severe IRB mission creep is across institutions. One variable that seems to be highly associated with how reasonable IRBs are in evaluating research risks is the proportion of the board that is comprised of active researchers. In too many institutions, serving on the IRB is akin to serving on the Faculty Senate – nothing that a productive scholar would consider. The result is predictable when IRBs are populated with faculty who are not (have never been?) active researchers.

      Another factor is that an IRB industry has been created. IRB administrators and staff are encouraged to get their CIP certifications (Certified IRB Professional) and more problematic for researchers is the movement of many universities to seek accreditation through AAHRPP (Association for the Accreditation of Human Research Protection Programs). This accreditation is intended to give university administrators some confidence that the IRB is following federal regulations and thus avoid potential funding suspension for violations. The accreditation requires development of extensive documentation of policies and procedures and promotes extensive use of checklists at every level of the review process. From what I can see, the effect of this push for certification and accreditation tend to focus IRBs federal regulations and interpretations, not on facilitating research while insuring human subject protections.

      Some of us hold out the hope that the proposed changes in human subject protection at the federal level will move to most social science research being excluded from IRB oversight. The risk there may be the recommendation to use HIPPA-type privacy protections for all research – in which case we may be simply substituting one detour for another.


  2. I couldn’t agree more! The strangulation of secondary survey research analysis by IRB and privacy guidelines is especially harmful for researchers and students at less advantaged institutions like my own. Now, many of the best datasets for examining a variety of issues require special security clearance and servers, and they charge money for gaining access to data. While top tier schools may think that $1000 to access Add Health or the New Immigrant Survey is a trivial fee, for those of us in the third tier it means that some faculty or staff member will not get a computer upgrade. We’re basically dead in the water for our Add Health research team because we don’t have the necessary computer support staff to implement a secure server (we’ve been waiting for more than a year, so only one of us has had access to the secure machine during that time). I had to BEG to use the NIS, because their data use agreement required that you were an established research center with external funding! All of this nonsense about privacy and security is completely misplaced. Has there been a single instance of someone trying to find a survey respondent from a large scale national survey? NO (prove me wrong, if you can). Would anyone want to? NO! Frankly, much of this is simply an end-around data sharing requirements, privileging PI’s and heavily funded research centers, and preventing others from competing.


  3. Forty years down the road, it’s probably all a bit late for the whole ‘resistance’ thing. Still lots of social science work on IRBs. For a historical review see “Ethical imperialism: institutional review boards and the social sciences” by Zachary Schrag. Will van den Hoonaard’s “Seduction of Ethics: Transforming the Social Sciences” is mainly Canada focused for its data, but has some interesting stuff.

    Best of all is Laura Stark’s “Behind Closed Doors: IRBs and the Making of Ethical Research” which is about IRB review as a whole (rather than just focusing on the social sciences) and has a good balence of historical/ethiographic data.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: