Erica Heath

	For a gloriously short policy statement, the 

 is guaranteed to create havoc. The idea is great: when an investigator submits a grant to any NIH institute, if the study will involve more that one performance site, the grant must include a plan to rely on one single IRB. However, implementation is likely to create chaos.

	The agency is looking “for ways to reduce procedural inefficiencies so that human subjects research can proceed efficiently without compromising ethical principles and protections.” This begs so many questions, including can anyone achieve maximum efficiency without compromise in other areas? Unlikely. 

	The first thing to notice is that the draft policy is from the NIH and applies only to NIH grants. The only studies involved will be from institutions that have signed a Federal Wide Assurance (FWA). The second thing is that the policy would be mandatory. Using a single IRB for all the sites would be a requirement for all NIH multi-site studies.

	Centralization of IRB review has been occurring naturally. For clinical sites without the federal grant requirements it has been a fact of life. Central IRB review using independent IRBs reviewing clinical sites started in the 1980’s. And the past decade has seen a momentous change with many academic institutions now outsourcing all of their industry-sponsored studies. But most have kept their federally funded studies in-house. The background section of the draft policy says that the regulations have always allowed reliance but that “too few institutions are taking advantage of the option”. There is no examination of why, but those reasons are likely what slowed the progression. Some of the obstacles have been removed and are as follows:

		Liability for review by the IRB being relied upon is clearer

		OHRP has altered its requirements to list every IRB on an institutional assurance

	There has been some development using a single IRB within the grants system. The National Cancer Institution Central IRB (NCI CIRB) initially conducted central protocol review and asked the local institution to conduct a “facilitated” review.  They now operate as a central IRB without involvement of any local IRBs. Several NIH institutes have been working on central IRB review for several years resulting in systems such as NeuroNEXT and StrokeNet. Both of these took intense effort and money and involved leaders in the grants and IRB fields. They are relatively new and are still working out the kinks. 

	There are efforts at consolidation: systems such as Kaiser have a single IRB per region and academic systems are cooperating to create system-wide reviews (e.g., UC Reliance Registry). Each of these innovations has reduced the total number of IRBs needed for any multi-site study.

	The NIH draft policy calls for an immediate jump from a natural progression to full immersion. The policy makes no distinction between large clinical multi-site studies and NIH grants for sociology studies involving three school districts. They are not dictating any particular structure or method. Having everyone jump into the fray at the same time may be a form of sink or swim that could cause many to sink.

	Central IRB review is certainly one way to achieve efficiency. But it might be inefficient to mandate it for everyone just yet. Perhaps an intermediate step should be to consider how to further encourage consolidation and cooperation among IRBs, to experiment with building strong networks and to demonstrate successful efforts.

, Presidentâ¨Ethical and Independent Reviewâ¨San Anselmo, CA and

Articles

The idea is great: when an investigator submits a grant to any NIH institute, if the study will involve more that one performance site, the grant must include a plan to rely on one single IRB.

Erica Heath

Articles

Centralization of IRB Review Fact of Life Outside of NIH