Laura Stark’s 2012 book, Behind Closed Doors: IRBs and the Making of Ethical Research, devotes a chapter to what Stark calls “local precedents,” her term for “the past decisions that guide board members’ evaluations of subsequent research.” “By drawing on local precedent,” Stark claims, “board members can read new protocols as permutations of studies that they have previously debated and settled based on members’ warrants. The result is that IRBs tend to make decisions that are locally consistent over time.” (47)
But I keep getting stories about IRBs that are locally inconsistent.
Stark made her generalization after observing a handful of cases in which IRBs claimed to be either basing decisions on ones they had previously made, or making new decisions that would guide future rulings. But as far as I can tell, she did not systematically audit IRB files to test the degree to which IRB decisions are in fact locally consistent.
I have long had my doubts about this claim of consistency, since so many stories from frustrated researchers concern identical studies presented to the same IRB with differing results. “The process that was approved in the first application was denied in the second because it was deemed coercive,” one complaint notes. “In every case in which we submitted the approved IRB application from the previous year, the IRB required additional changes,” another laments. A candid IRB chair admits that “Investigators may get quite different and inconsistent advice from the committee depending on what it feels like that day.”
Two comments posted in response to Patricia Aufderheide’s essay, “Does This Have to Go Through the IRB?,” cast further doubt on claims of local consistency. Here they are, with the screen names of the authors:
At a university I don’t work at, I had a series of go-rounds regarding a caveat I placed in a proposal that boiled down to explaining that the participant pool I was working with is very small and there’s a chance some of the participants will talk to one another about this research. I had used identical language in two previous proposals at the same university. To appease the reviewer (who also strayed from reviewing to critique other areas of the proposal) I eventually removed the statement, even though I felt it was ethically appropriate to point this out and an unavoidable outcome of the type of research I am doing. The IRB personnel I worked with in that office helped as a go-between, but what a waste of time for her and for me, and I felt a tad compromised in the process.
At one of my previous institutions, the IRB disliked the size of the font on the recruitment flyer I wanted to put up around campus, saying that the size in which I indicated the amount of compensation was too big relative to the size of the words saying “research study,” and therefore that constituted undue influence to participate. Silly, but I guess I can run with it, except that I used the exact poster template that my lab had been using for years under a different protocol. Depending on who or when the protocol was reviewed, absurd ticky-tack font-size objections would require corrections and resubmissions of the protocols, which resulted in an additional three or four weeks of waiting for approval.
Such stories do not reveal the extent of IRB inconsistency, but neither do Stark’s observations. Without further investigation, I don’t think she can support the claim that “IRBs tend to make decisions that are locally consistent over time.”
Also, note how in the first story, the IRB interference had the effect of denying useful information to prospective participants.