IRBs and Clean Secrets

There’s a comment piece I wrote that appears in today’s issue of the journal Nature that talks a bit about the role of open data and IRBs. But I worry that perhaps in the number of iterations it made before publication the main point got muddied a bit. So here it is:

Funding agencies and journals should require authors to submit and openly publish protocols as submitted to the IRB.

There are a bunch of reasons for this. First, IRB protocols should be public. Right now, it’s treated on the one hand a bit like the Napoleonic code: it doesn’t matter what others have decided, the board decides entirely on the basis of your submitted application. This has some real negative implications.

First, the same protocol may be accepted at one campus and rejected at another. Or research in the same stream (or seeking to replicate) may be rejected at a later date.

Second, the IRBs have to make an original determination each time wasting, in many respects, the efforts of other competent IRBs who have already made a determination. They don’t need to be bound by earlier determinations, but don’t you think it would be worthwhile to be at least aware of them?

Third, when applicants feel as though they are handled unfairly, a transparent system is better for all.

Fourth, the best way to learn to do ethical research is to be able to observe the process from the periphery, and listen to the queries of IRBs and the responses by researchers. Releasing the approved protocols may not get at that deeper conversation completely, but it at least provides a small window.

Fifth, the protocols are an excellent way of “indexing” open data. Open research data often is published with codebooks and other ancillary material, but IRB protocols in many ways are the ideal introduction to an open collection of data, explaining why it was collected, how it was collected, and how it might be used.

For these reasons, among others, there should be open repositories of IRB protocols. Now, we could just try to convince individual IRBs on campuses to open up their process and publish protocols they approve, and I hope that they might. But IRBs are by nature a conservative group, intended to protect, not to disrupt. In many cases, they are made even more conservative by the institutions that they are housed in, and concerns by that institution either that they might be sued by subjects or that they might be investigated by federal regulators. (Those regulators, naturally, have access to the protocols and the decision process of the IRB once an investigation or audit begins, but they might not want to provide any sort of “probable cause.”)

Individuals might be encouraged to submit their own protocols to a repository, and in fact, self-archiving has made an important impact on the way publishing happens. But there are enough open questions surrounding this that it’s a hard place to start the ball rolling.

Funding agencies, and more recently journals, either insist on or facilitate the sharing of research data. In many ways IRB protocols are an important part of those research data. If funders required that IRB protocols be shared just like any other research data, and if journals provided authors the resources to share these protocols, it would revitalize in some ways the role of scholarly publishers and it would make for ethical oversight that was more robust and transparent.

But what if you were not IRB approved? It may be you didn’t need pre-approval of research by the IRB, and as I argue in the article above, I think this should be the case for much of the research that is currently placed under some level of review by IRBs. But if you don’t have to have IRB approval, I think funders should still require you to talk about the ethical considerations of your research, and journals should require you to publish this online when you do not have an IRB-approved protocol to provide.

What this does is creates an environment in which ethical post-review is encouraged. Certainly, when it comes to drug trials–and even to invasive forms of research of vulnerable populations in the social sciences–there should be some sort of oversight before the research occurs. But even after it occurs, peer-reviewers and the reading public should be able to see how the researchers weighed the needs and rights of subjects against the importance of their research questions.

This entry was posted in Research and tagged , , , , , , , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Posted 12/9/2011 at 1:01 pm | Permalink

    Great article. As a long-time member of an IRB, I completely agree with the arguments you make here. Of course, as you will admit, sometimes IRB’s do important work, such as when they prevent (in one case I remember) the distribution of a non-anonymous survey to students that asked them to reveal potentially illegal conduct. In general, though, there is plenty of overregulation. I will never forget a case from several years ago in which we were asked to approve a study that asked veterans of the wars in Iraq and Afghanistan their opinion of the foreign policy of the Bush administration, and the study was denied because of fears that the question might trigger PTSD. Ridiculous. If that study triggers post-traumatic stress, then vets should never be allowed to take history or political science courses.

    • Posted 12/9/2011 at 2:41 pm | Permalink

      Thanks! I definitely agree, and suspect I’ll find myself on an IRB soon enough. I’m sorry that the emphasis–especially in the Nature piece–came out on the side of reducing IRB involvement. I do think it should be easier to do social science research without pre-review (and OHRP seems to agree!), but I think the openness issues is more important.

  2. Posted 12/16/2011 at 11:53 pm | Permalink

    IRBs and Clean Secrets

Post a Comment

Your email is never published nor shared. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>