Central IRBs: What are we in for? Can it work?

by Susan Trinidad, MA, Research Scientist in the Department of Bioethics & Humanities at the Center for Genomics & Healthcare Equality at the University of Washington

Yesterday’s pre-conference program on centralized IRB review raised and clarified a number of issues I hadn’t considered before.

For starters, the fact that most IRBs are responsible for so much more than human subjects review – HIPAA, FDA compliance, etc. – makes this a much more complex undertaking than it may appear at first glance. And not everything can be delegated. For example, a centralized IRB may have authority for review, but investigation of noncompliance, education and oversight of investigators, and many other tasks remain within the bailiwick of the relying institution (i.e., the institution that cedes review authority to a central IRB).

Some people have suggested that federal research funding may become contingent on institutions’ agreement to accept centralized review – similar to NIH data sharing policies that have been implemented in recent years. If that happens, centralized review will be a high-stakes issue sooner than we think.

Pearl O’Rourke, director of human research affairs at Partners HealthCare Systems,
started us off with introductory remarks and talked about the possibility of centralized review being “mandated” – i.e., required – depending on the outcome of the 2011 Advanced Notice of Proposed Rulemaking to the Common Rule (the ANPRM). She proposed a taxonomy ranging from status quo, in which the local IRB holds the bag by itself, to a “non-share model,” in which a central IRB conducts all reviews. For reasons that became clearer as the workshop went on, it seems likely that most implementations will fall somewhere in between.

Julie Kaneshiro of OHRP spoke to the question one audience member phrased half-jokingly as, “What was OHRP thinking?!” The majority of public comment has been supportive of proposed change, but as Emily Chi Fogler, an attorney at Partners, explained, teasing apart the responsibilities and authority of the central IRB and the relying institution is complicated and needs to be negotiated and defined in detail.

Reality checks are always a good thing: we were fortunate to hear from Elizabeth Hohmann, an IRB chair at Partners, about their experience with the NeuroNext Central IRB; Lynn Cates, director of the VA’s Program for Research Integrity Development and Evaluation; and Jacquelyn Goldberg, head of the National Cancer Institute’s CIRB Initiative. They generously shared lessons learned, potential pitfalls, and insights based on their experience.

The move toward centralized review is strongly supported by NIH and by many investigators; not coincidentally, the vast majority of arguments in favor of centralized review have to do with increasing efficiency and reducing costs. It’s tough to argue against efficiency, especially given the budget crunch all of us face. But one of the issues that we circled around and around yesterday was the inherent difficulty (and in some cases, impossibility) of delegating responsibility for considering the local context.

Can a central IRB really know what matters to the community, especially in participant populations that are culturally unique, vulnerable in unfamiliar ways, and/or are far removed from the coordinating center? If so, how? And if it can’t, could we end up sacrificing the protection of the most vulnerable participants – in the name of saving money and reducing the hassle factor for investigators and funders?