Institutional review boards are a constant subject of complaints from scholars about delays or limits placed on research -- even if they understand that a given IRB may only be following the rules. Sarah Babb, a professor of sociology at Boston College, tells the story of IRBs in a new book, Regulating Human Research: IRBs From Peer Review to Compliance Bureaucracy (Stanford University Press). She responded to questions via email about her new book.
Q: How are IRBs different today from when they were first created?
A: When the first proto-IRBs emerged within National Institutes of Health in the 1950s, they looked a lot like peer-review committees: NIH researchers would sit on panels to consider the ethics of their colleagues’ research proposals. Even after the 1974 regulations and two rounds of revisions (in 1981 and 1991), most IRBs were mostly run pretty much as faculty committees, but with “nonscientist” members and some federal paperwork. The faculty volunteers weren’t very good with the paperwork, but since the feds hardly ever checked, it wasn’t a big issue.
Nowadays, most research is being reviewed by IRBs that look a lot more like “bureaucracies” in the Weberian sense. They have hierarchies, job ladders and an elaborate division of labor, kind of like an assembly line. And they are run by paid administrators with specialized regulatory expertise. The most bureaucratic of all are the for-profit “independent” IRBs, which compete for clients based on their ability to provide rigorous regulatory compliance with maximum speed. Independent IRBs don’t have faculty “volunteers” -- academics are paid per review to serve on particular boards. Everything else is handled by administrators. Today, most biomedical research is reviewed by these for-profit boards.
Q: You talk about how academic research is increasingly commercialized. Have IRBs responded to that change? How should they respond?
A: I would rather say that IRBs have been shaped and transformed by commercialization. Most biomedical research today is sponsored by private firms -- mostly (but not entirely) outside of academic institutions. Commercial studies mostly involve multiple research sites. Imagine 20 different clinics, all recruiting participants for the same study. But federal IRB regulations are set up around a local review model. This has always posed a problem for commercial sponsors, because clinics don’t have IRBs, and even if they did, you wouldn’t want 20 IRBs coming up with 20 different decisions.
Recognizing this problem, back in the 1980s the Food and Drug Administration put a loophole in its rules saying you could outsource IRB reviews instead of relying on a local board. In response, independent review boards were founded to serve commercial research sponsors. But more recently, even government-sponsored studies have started going over to the for-profit IRBs. This is because they are extremely efficient and very good at reviewing multisite research (which is now the dominant model throughout the biomedical field). Traditional IRBs have been losing staff positions. To get a competitive edge, many have emulated some of the practices of for-profit IRBs.
So in a way, the whole IRB world has become commercialized. In my book, I argue that this is an American pattern, where our inability to rationalize public systems leads to the emergence of markets and private “workarounds.” It’s interesting to contrast this with how they do it in the U.K., where they decided decades ago that the best way to deal with multisite studies was to have a single government-sponsored portal that everybody goes through. The British don’t have for-profit ethics review because it isn’t needed.
Q: Your book talks about how IRBs have become “compliance bureaucracies.” Was this inevitable?
A: The reason IRBs became compliance bureaucracies is that compliance is a lot of work. The regulations are complicated and confusing. There are actually two sets of IRB regulations, run out of two different federal offices, and [they are] not entirely consistent. Over time there have been a bunch of other ancillary regulations to keep track of -- HIPAA, conflict of interest and so on. The regulations are ambiguous, and federal offices don’t have the mandate or the resources to set precedents by ruling in particular cases. What regulators do (although less lately) is audit IRBs to make sure they have meticulously followed and documented mandated procedures -- for example, that they considered each of the criteria for waiver of documentation of informed consent. This means that IRBs have to generate and store mountains of procedural paperwork. Over all, it’s a big job -- one that calls for the attention of full-time specialists.
I found that there was a turning point at the end of the 1990s, when Senator [Edward] Kennedy and others were calling for a comprehensive rationalization and centralization of the federal system. Unfortunately, there was too much opposition from Congress and vested interest groups. But I suspect that if Kennedy’s plan had gone forward, you wouldn’t have seen the same growth in compliance bureaucracy: it would have centralized a lot of the interpretive and administrative work, similarly to the way you see in the U.K. today.
Q: Some argue that IRBs cannot simultaneously review the biological sciences and the social sciences. Should universities have separate IRBs?
A: This is a really great question. My impression is that it is really common to have separate IRBs for social and biomedical research, but that this doesn’t necessarily address the problems social researchers encounter with the IRB process.
Our problem is twofold: first, the regulations were designed around the routines of experimental research; and second, academic administrators tend to respond to risk and uncertainty with “hypercompliance” -- going beyond what the regulations require or intend. Imagine an anthropologist doing ethnography somewhere in an Indian village, trying to go through a three-page informed consent form (including all kinds of legalistic and regulatory language) with each person he interacts with. And then imagine that every time he slightly changes his research design, he needs to wait for his IRB at home to reapprove the changes. It sounds absurd, but this was a common experience at the beginning of the 2000s -- and I suspect that it still happens today.
What really has helped social researchers like me is something called the “flexibility movement,” a recent initiative among some IRB administrators to move away from hypercompliance. If you’ve noticed your IRB getting a lot more social research-friendly recently, you probably should be thanking the flexibility movement.
Q: What can universities do to improve the functioning of IRBs?
A: There’s a whole industry of IRB consultants and accreditors that are better qualified than I am to answer that question. But as a sociologist, I have the luxury of thinking about the big picture. What we have in the United States today is a highly decentralized and commercialized ethics-review system; recent revisions to the IRB regulations reinforce this. Because it generates high administrative costs, our system has created a sort of marketplace in ethics review, in which a small number of for-profit boards compete for clients based on their efficiency. This is a very weird way of handling ethics review, and very different from the way they do it in other wealthy democracies.
The advantage of our American system is that it can’t be easily dismantled by anti-regulatory politicians. But one disadvantage is that it creates potential conflicts of interests: you have researchers and sponsors shopping in a marketplace of IRB services and paying the board they select to regulate them. We haven’t had an outbreak of research scandals since the 1990s. I really hope that we don’t look back on this time as the calm before the storm.