In December 2014, the National Institutes of Health (NIH) released draft guidance detailing its support and expectations for the use of a single IRB for multisite NIH-funded studies. But for many in the IRB community, the guidance raised more questions than answers.
“I don’t think the case has been made that the use of a single IRB is absolutely necessary,” says Erica Heath, CIP, director of regulatory affairs at Ethical & Independent (E&I) Review Services in Corte Madera, CA. “I think there are good cases for it, but I think other ways to create and achieve consistency and efficiency have not been explored sufficiently.”
In its guidance, the NIH stated that the use of a single IRB for its large, multisite studies is more efficient than multiple reviews and could eliminate redundant reviews and slow turnaround times. “In fact, the use of single IRBs may lead to enhanced protections for research participants by eliminating the problem of distributed accountability, minimizing institutional conflicts of interest, and refocusing IRB time and resources toward review of other studies,” the document states.1
“There’s no evidence, even with all the money and effort going into it, that the current IRB structure is actually protecting subjects,” says Mark Schreiner, MD, vice-chair of the Committees for the Protection of Human Subjects at the Children’s Hospital of Philadelphia (CHOP), and physician reviewer of IRB Advisor. “There’s a lot of duplication of efforts, and IRB budgets and staff are tight.” Multiple IRB review should be done if the review adds value, he says.
One issue with the guidance, Heath says, is that it doesn’t clearly define terms or what, exactly, the NIH is looking for. “When they say ‘a single IRB’ in the proposal, they don’t say what they want, and it’s really quite open,” she says. “I think a lot of people are a little bit at sea over what it is they [NIH] really want. It may be a good direction, but a bit hasty.”
In a recently released position paper, Public Responsibility in Medicine & Research (PRIM&R) stated that while central IRB usage would be beneficial for some NIH-funded studies, it may not be appropriate for all.
“PRIM&R believes that, before such a policy is implemented, further reliable empirical evidence is needed on the various ways in which a single IRB can be used to provide ethical review of multi?site research, and on whether such review is better, from the perspectives of subject protections, administrative costs, efficiency, and quality of review, than relying on local IRBs,” according to the position paper. “In the absence of sufficient evidence, we believe that a policy requiring the use of single IRBs for all domestic sites of multi-site NIH?funded studies is premature and ill advised,” the paper states.2
But this requirement from the NIH is nothing new, Schreiner says. “We’ve seen many studies where we’ve had to agree to have a central IRB as part of the [NIH] grant application. It’s not just hypothetical; it’s reality.”
Many IRBs do not have experience in deferring to a single IRB and may not know where to start, NIH officials acknowledged in the guidance. “Despite enthusiasm for central IRBs, there is confusion about the optimal structure for central IRBs as well as how best to meet regulatory requirements,” according to an NIH statement. “There are questions about the loci of responsibilities and whether the IRB or institutions will bear the blame if adverse events occur,” the guidance states.1
NIH should offer more ways to answer these questions and clear up confusion, says Cynthia Hahn, vice president of clinical research and regulatory affairs at the Feinstein Institute for Medical Research at North Shore-LIJ Health System in Manhasset, NY. “There really needs to be more guidance,” she says. Hahn is also on the steering committee of the Clinical Trials Transformation Initiative (CTTI) and co-authored research and recommendations for using central IRBs that was cited in the NIH draft guidance. “One thing the NIH needs to do is provide feedback to sites with things they should consider.”
Topics for consideration when using a single IRB, according to Hahn, include the following:
• What are each participating IRB’s bylaws? Do they conflict with the bylaws of other institutions?
• What are the bylaws for the medical staff at each institution? Are they in conflict with other institutions? Do the bylaws state that the medical faculty must use the institutional IRB?
• Contracts, bylaws, policies, and procedures may need to be updated.
• Who does institutional approval? If you use another IRB, who does approval for your institution?
“At the same time, you don’t want to duplicate work,” Hahn says. “You don’t want to replicate what the IRB already did. I think the NIH would be helpful to provide that kind of guidance.”
Drafting authorization agreements could also be a stumbling block for IRBs that don’t have experience in that area, Hahn says. “Lots of places don’t know how to do this and don’t know where to start. Some are comfortable with the one-page OHRP template, but it’s not sufficient,” she says.
“I feel sorry for IRB administrators because they will be faced with so many different kinds of reliance agreements,” Heath adds. “Agreements range from the one-page OHRP template to 20-page contracts. Administrators will have to go through and see the minimum their institution would need — if they need more conditions or fewer. Unless there is some move to get some consistency in their agreements, administrators will go nuts.”
Schreiner agrees that there is no guidance as to the qualifications the reviewing IRBs must have. “It’s clear to us that not all IRBs have the experience or manpower to play that role [of reviewing IRB],” he says. “The initial guidance should be fleshed out and include criteria for qualifications to be a reviewing IRB.”
But there are positives to the use of a single IRB, Hahn says.“I think the whole thing [the NIH draft guidance] stems on people knowing that the days of single-site studies are basically over,” she says. “More and more studies are becoming multinational and multisite. Studies are increasingly complex, addressing rare diseases and conditions, and experts on those diseases might be across the country from you. I don’t think that’s a bad thing. [Single IRBs] could produce better quality reviews, not just more efficiencies. Most of the research has been driven by efficiencies and people need to focus more on a quality review.”
There are pluses and minuses to multiple IRB review and single review, Schreiner adds. For instance, multiple IRB reviews could catch something that a single IRB may not. “On the other hand, if there’s an issue with the trial, a single IRB has much more impact to forcing changes to a trial than 30 sites do,” he says. “If one of 30 sites raises objections, they are dropped as a site. There is a lot of leverage that comes with a single IRB review.”
The single IRB policy could have unintended consequences for smaller regional IRBs that operate in the shadow of a large academic center, says W. Parker Nolen, MBA, CCRC, CIP, network manager of the IRB at Community Health Network in Indianapolis.
“My initial reaction is that it’s nice that streamlining is occurring, but it’s not necessarily in the best interest of those who don’t have a lot of NIH funding,” he says.
It’s a question of access to study participants, Nolen says. If a researcher wants to use the patients at a particular hospital as part of a multisite study, “the inability of the IRB to at least look at the consent form is a bad idea,” he says. “It exposes the hospital system to a great degree of liability. They’re then faced with the decision to turn over the population to a researcher elsewhere, or to become a subcontractor for the grant.”
“Turning it over blindly is never a good idea,” he adds. “The IRB may not allow physicians and patients to participate because it’s too great a risk.”
It’s also a question of patient safety and personal liability, Nolen says. “It places hospitals at a large disadvantage if they have to say, ‘Wait, we need to get on this as a subcontractor,’” he says. “My gut says it’s really bad for the long run — in situations like that, research could start drying up at hospital systems.”
But having a single IRB review is not an abdication of the relying IRBs’ responsibilities, Schreiner says. “There are many things that remain with the relying IRB,” he says. “The relying IRB is still responsible for the review of consent forms and for adverse event reporting.” The relying IRB can also object to the central IRB’s review. “Just because there is central oversight doesn’t mean they [relying IRB] have to accept it as-is. Having an IRB of record does not absolve the reliant IRB of its responsibilites of research.”
Experts propose alternatives
In its position paper, PRIM&R encouraged the NIH to consider incentives for IRBs to voluntarily adopt the single IRB model, and suggested the following:
1. “Convene an expert panel to host open meetings to develop criteria regarding the types of research that lend themselves to the single IRB model and those that do not;
2. “Sponsor research on existing models of review by a single IRB for multi?site research to gather more evidence about both the quality and cost of review;
3. “Create incentives (e.g., preferential treatment in the award process) that encourage and reward the use of, and require the collection of data on the use of, single IRB review and/or elements of single IRB review processes; and
4. “Develop tools, guidance, and best practices to help facilitate the use of single IRB review mechanisms (e.g., model reliance agreements, standard operating procedures, etc.).”2
There are also many different models of collaborative IRB systems, Heath says.“If it’s mandatory, you forgo other forms and styles,” she says. “You get a lot of people working together who may not want to. There are a lot of good things about central review — I do it — but there are better ways of being successful.”
Other models include collaborative IRBs, IRB consortia (such as university system consortia), and other reliance models. There are also methods for shared communication, Heath says.
“The first several IRBs could look at something and all their questions and answers could be compiled and given to the next IRBs to forestall their questions,” she says.
In a perfect world, Nolen says, the draft guidance would be a good model. “But in reality it’s kind of symptomatic of academic myopia — they think that almost everything coming through NIH is through large academic medical centers, and it’s not,” he says. “I understand the philosophy of it; the NIH has lost a lot of funding from sequestration. They need to be efficient as well.”
“There are a lot of things to work on, but it doesn’t mean we shouldn’t go forward,” Schreiner adds. “The biggest issues are how to implement it [the model] and operationalize the IRB.”
- Request for Comments on the Draft NIH Policy on the Use of a Single Institutional Review Board for Multi-Site Research. National Institutes of Health. http://grants.nih.gov/grants/guide/notice-files/NOT-OD-15-026.html#sthash.I3F6X3W4.dpuf
- PRIM&R’s Response to NIH’s Draft Policy on the Use of a Single Institutional Review Board for Multi-site Research. http://primr.blogspot.com/2015/02/prim-response-to-nihs-draft-policy-on.html. February 3, 2015.