Last December, the National Institutes of Health proposed a major change in conducting ethical oversight of medical research. In its “Draft NIH Policy on the Use of a Single Institutional Review Board for Multi-Site Research,” the NIH suggested that it would require one ethical committee, a central IRB, to take jurisdiction over multiple research sites instead of relying on many local IRBs. Although the policy’s comment period closed months ago, the proposal also received attention from the 21st Century Cures bill that recently passed the House of Representatives. The current language of that bill requires the NIH to finalize the policy on central IRBs within 12 months of the 21st Century Cures bill becoming law.
The debate over the merits of central versus local IRB review is not new, but I was curious to see where opinions in the research community currently lay. The NIH’s proposal generated a rich selection of opinions to consider. The 16-paragraph announcement inspired 348 pages of comments from 167 sources. The comments reflected some of the key issues in considering local versus central IRB review. After reading through the comments I created a scoresheet to summarize them.
Responses to the “Request for Comments on the NIH Draft Policy on the Use of a Single Institutional Review Board for Multi-Site Research”
|Comments by||For||Against||Not Sure||Total|
For a “For” or “Against” rating, I looked for a clear statement of support or opposition; otherwise the comment landed in the “Not Sure” column.
Regarding the categories of commenters, here are some brief explanations:
- Associations: This was a diverse collection. Any comment that came from a general group went here. This category of commenters included groups of physicians, research professionals, research ethics professionals, advisory committees, and tribal nations.
- Central IRBs: Here I collected comments from three independent IRBs and one federally-funded central IRB.
- Individuals: These were commenters who said they were expressing personal opinions. As the table shows, over half of the comments came from here.
- Institutions: This group was for official responses from hospitals, universities, or sections within them.
- Sponsor: One pharmaceutical company provided comments.
(Note: The draft policy and most of the comments mostly assumed “central IRB” meant that a review committee at one research site in the study would oversee research in the others. This sometimes is called a lead IRB model, and the NIH has utilized it before. An independent IRB such as Quorum Review represents another model, where an ethics board not affiliated directly with any sites in a study has oversight of the research, but the bulk of the NIH-related comments focused on the lead IRB model.)
Overview of the Responses
While a strong majority of comments supported the NIH’s proposal, a quick study of the table shows how institutions bucked that trend. Almost two-thirds of all comments clearly endorsed the idea, but less than half of universities or hospitals did. This is an especially interesting group to consider, because they are the ones which the new policy could affect the most. The institutions’ administrators and researchers could have the most adjustments to make in the face of a mandate of single IRB review for federally-funded studies.
With that in mind, we can delve a bit further into the specifics of what the opponents and the supporters had to say. When I went through the comments, I found quite a few common themes on either side. I’ll first summarize the most common arguments first, and then share some excerpts from the 167 comments themselves.
Opinions of those AGAINST
Here were the most frequent arguments against the NIH policy:
- Encourage, don’t mandate: Observers objected to the NIH making the change mandatory. Those observers requested that the NIH consider creating behavior-changing incentives instead.
- The policy’s scope is too broad: Many of the commenters argued that the range of NIH-sponsored research is too wide and varied for a universal rule of centralized review.
- The evidence isn’t there: The most outspoken critics said the NIH’s proposal failed to prove that central IRB reviews save time, save money, or provide the same ethical protections as local IRB review.
- The policy will make things worse: Many observers argued that relying on central IRBs for multiple studies would increase rather than decrease the administrative burden on sites. Instead of relying on the particulars of one IRB (that is, the local IRB), institutions could face new rules from another IRB with each NIH-funded study.
- A loss of collaboration: One advantage of local IRBs is local understanding. The proposal’s opponents worried that central IRB review would curtail awareness of local conditions, as well as communication between investigator and IRB reviewer.
And of those FOR
A solid majority of commenters endorsed the NIH’s idea. While the backgrounds of the commenters varied, their arguments were largely consistent:
- It will save duplication of effort: The associations and the individuals who endorsed the NIH’s proposal cited examples or personal experience of how a central IRB review simplified a multi-site study.
- Multiple IRB reviews do not improve participant protections: In contrast with the opponents, supporters of the policy agreed with the NIH that central IRB reviews do not reduce the protection of study participants.
- Increased clout of one IRB that oversees many: Commenters argued that an IRB overseeing multiple researchers could have more influence in a study than many separate IRBs.
Common requests from all sides
From the collection of comments, criticisms, and kudos rose two prominent recommendations. Considering the diversity of the comments, the consistent appearance of these two suggestions was telling:
- Ensure that the central IRB for any NIH study meets clear qualifications and capabilities of the central IRB in question. Accreditation was a common suggestion, as was a reliable electronic process.
- Include training and document templates as part of the policy. The NIH should include guidance and assistance with any implementation.
All told, I was surprised at the level of support that the proposal received. The comments from the individuals and the associations were the most enthusiastic. I was not surprised, however, at the level of thoughtfulness and insight that went into some the comments. The proposal raised important issues of resources, priorities, and ethics. Here are excerpts of some specific comments that appeared:
Association: Alzheimer’s trials already are long, and multiple IRB submissions can delay study results. Members of this Alzheimer’s-related Association “overwhelmingly support” central IRB review.
Association: This trade association “salutes this bold stroke.”
Association: The policy seems reasonable, and contains avenues for necessary exceptions. The final policy should require criteria for a central IRB.
Association: While favoring the policy, this association believes that selection criteria for the central IRB are “of paramount importance.”
Association: “We commend NIH for advancing a policy that encourages efficiency while still protecting those who participate in research.”
Association: Non-tribal IRBs cannot understand tribal customs and society. Tribal nations should have an exception and retain their own reviews.
Institution: “Ceding IRB review is contrary to developing an IRB culture of a shared responsibility.”
Institution: It will be counter-productive to prepare submissions for another IRB while the local IRB continues to review other research and support affiliated entities.
Institution: The evidence for benefits is not in. It is too early to mandate central IRB review for all multi-site studies.
Individual: “What you are likely to generate is chaos rather than efficiency.”
Individual: “(W)e are confident that the CIRB [Central IRB review] will have a significant positive impact on study planning, start-up, and conduct.”
Individual Multiple IRB reviews do not improve participant protections. If anything, they protect the institutions.
Individual: “(A) dream come true.”
Association: “The amount of effort, time, and money spent on IRBs alone is enormous and adds nothing to research.” The letter compares the experience of a study with seven IRBs favorably to another with fifty-one.
Sponsor: Works with about 175 IRBs in 250 studies each year in the United States. The complexities from that delay trial starts and create “unnecessary administrative variation.” This sponsor supported the proposal, but noted the difficulty of implementing quickly.
Association: This group believes it is too early to require single IRB review for all NIH-funded studies. Empirical data on advantages of central IRB are “sparse to nonexistent.”
Association: “Mandating single IRB review . . . is not the appropriate solution to improve turn-around time for human subject’s research.” If NIH implements the policy, it should provide a set of central IRB qualifications, including accreditation, knowledge of state laws, and an effective record-keeping system.
Institution: Signed by eight IRB Chairs, one university’s letter argues the data are insufficient to prove a central IRB model is better. The NIH should provide incentives, not requirements.
Institution: Suggests alternatives to the lead IRB model. The alternatives include designating some institutions’ IRBs as dedicated central IRBs for the NIH; utilizing independent IRBs in NIH research; or creating freestanding IRBs at the NIH through outsourcing.