| Comment Record|
Dr. Robert Lowman ||
2002-06-04 16:26:27 |
University of North Carolina at Chapel Hill |
| Comments for FDA General |
1. General Comments
Thank you for the opportunity to comment on the issue of IRB shopping. I write as an individual, not representing the University of North Carolina at Chapel Hill, where I am Associate Vice Chancellor for Research.
I have chaired two different IRBs for a total of 13 years. For the last 11 years, the 8 IRBs at UNC-Chapel Hill have reported to me in my role as Institutional Official. The answers to your questions are numbered below.
1. At a university such as UNC-Chapel Hill, the problem of IRB shopping is virtually non-existent. We review every protocol for research conducted by our own faculty, staff and students and every protocol to be conducted on our premises or using our facilities. The only exception is in situations where we enter into a formal, written agreement with another institution to rely on the review of its IRB in lieu of our own. Because of the requirement to obtain approval from one of our own IRBs, there is little or no incentive for faculty, staff or students to take their protocols to another IRB, and hence little incentive for shopping.
There is a small amount of shopping that takes place on our own campus, as some individuals may believe the IRB that normally reviews protocols from their academic unit is not the best qualified to do so in special cases. This is seldom an instance of an individual going to a second IRB after being disapproved at the first. Instead, it is a selection of the right or best IRB for the initial review. The IRBs here communicate with each other whenever this occurs, and the IRBs decide where the review will take place.
The percentage of protocols ultimately rejected here is trivial. Investigators routinely and overwhelmingly make the modifications required by the IRB in order to obtain approval. Often those required changes are quite substantial.
Our sample of IRB shopping--if it exists at all--is too small to allow any generalizations regarding the circumstances under which it might occur.
2. If the FDA decides to require disclosure of previous reviews, whoever is responsible for submitting a protocol to an IRB should be responsible for the disclosure, whether sponsor or investigator. For multi-site studies, any failure to receive approval should be disclosed to all IRBs considering the same protocol, but the mere fact of multi-site review need not be reported, as the reviews generally take place more or less simultaneously.
In multi-site studies, the sponsor should be the clearinghouse. That is, investigator X would notify the sponsor if an IRB failed to approve his/her protocol. The sponsor would notify investigators Y and Z, who would, in turn, notify their respective IRBs.
3. For multi-site studies, IRBs should be notified of a lack of approval, even if it occurs after they have themselves already approved the same protocol. Again, the sponsor should be the clearinghouse. I would allow sponsors to delegate this responsibility to a data management center or a CRO, but the delegation should be in writing.
4. Only reviews that do not result in approval should be disclosed to other IRBs. That would include outright disapproval, but it would also include reviews resulting in stipulations never met by the investigator to gain approval.
5. Because the overwhelming majority of protocols is ultimately approved, failure to approve is the special case or exception. In the interests of minimizing the regulatory burden, only reviews that do not result in approval should be disclosed.
6. A short description of the grounds for the failure to approve should be conveyed to the second or subsequent IRB. It would be cumbersome, but the IRB that failed to approve may need to approve the language used to describe its action. One method of circumventing this need for language review by the first IRB might be to require that actual language from the first IRB minutes describing the disapproval or unmet stipulations be conveyed to the second or subsequent IRB.
7. Voluntary notification of prior review without approval is the only practical method of enforcement of this requirement. This is the method used with adverse event reporting or the reporting of serious deviations from protocol or violations of the regulations. The FDA has available to it the full range of sanctions that can be applied to institutions or individuals, as it deems appropriate for non-compliance.
Again, thank you for the opportunity to comment on this important topic.
Robert P. Lowman, Ph.D.
Associate Vice Chancellor for Research
The University of North Carolina at Chapel Hill