Notice Number: NOT-OD-16-094
Release Date: June 21, 2016
Effective Date: New Date - September 25, 2017 as per issuance of NOT-OD-17-027 (old Date: May 25, 2017 )
National Institutes of Health (NIH)
The National Institutes of Health (NIH) is issuing this policy on the use of a single Institutional Review Board (IRB) for multi-site research to establish the expectation that a single IRB (sIRB) of record will be used in the ethical review of non-exempt human subjects research protocols funded by the NIH that are carried out at more than one site in the United States. The goal of this policy is to enhance and streamline the IRB review process in the context of multi-site research so that research can proceed as effectively and expeditiously as possible. Eliminating duplicative IRB review is expected to reduce unnecessary administrative burdens and systemic inefficiencies without diminishing human subjects protections. The shift in workload away from conducting redundant reviews is also expected to allow IRBs to concentrate more time and attention on the review of single site protocols, thereby enhancing research oversight.
Overview of the Public Comments
In general, most of the comments that were submitted on the draft policy were supportive of NIH’s goal of enhancing and streamlining IRB review in multi-site research. Commenters, especially individual researchers, scientific and professional societies, and patient advocacy organizations, generally agreed that the use of a single IRB for multi-site studies involving the same protocol would help streamline IRB review and would not undermine and might even enhance protections for research participants. Most of the comments also favored the approach the NIH proposed to promote the use of single IRBs by making reliance on an sIRB an expectation for all non-exempt multi-site studies carried out at U.S. sites. At the same time, a number of commenters, mainly academic institutions and organizations representing them, did not agree with the scope of the proposed policy or that it should become a term and condition of funding, and suggested the NIH incentivize, not mandate, reliance on an sIRB.
Comments from researchers that supported the draft policy described unnecessary delays and additional costs caused by duplicative IRB reviews. They noted that IRB submission requirements at each site differ and take time to navigate and manage. They also indicated that review of the same protocol by multiple IRBs can sometimes lead to protocol and consent document changes that can introduce inconsistencies in the execution of the protocol across sites, lead to enrollment imbalances, and skew the analysis of the aggregated data. More often, however, multiple IRB reviews result in changes to consent documents that are merely stylistic and not substantive, or changes that focus on institutional interests (e.g., liability management) rather than human research protections. Commenters raised the concern that the current practice of requiring multiple IRB reviews may actually contribute to some researchers’ reluctance to participate in rigorous, multi-site research and may incentivize smaller and simpler study designs.
Scientific and professional societies generally favored the proposed policy. These stakeholders stated that the policy would decrease administrative burdens on clinical research staff, speed up participant recruitment, and streamline the research process and that these changes would result in enhancements to the efficiency of research and acceleration of research progress. They also suggested that the benefits of such a policy include enhanced adverse event monitoring and improvements to the quality and consistency of IRB reviews.
Most of the comments from patient advocacy groups and participant representatives were supportive of the proposed policy. These stakeholders pointed out that greater use of single IRBs will lead to enhanced protections through increased accountability and improved efficiency.
In general, comments from academic institutions, IRBs, and organizations that represent them cited concerns about the proposed policy, even though many also expressed support for its goal and agreed it could have a positive impact in reducing research review and initiation time to the study. These stakeholders suggested that the scope of the proposed policy is too broad and that the NIH should not make the policy a term and condition of award. They said that decisions about whether to use a single IRB should be voluntary and that the NIH should offer incentives to promote change. For example, they suggested that the NIH encourage investigators and institutions to use single IRBs in grant applications by providing additional funding to those grants that agree to use a single IRB. Some suggested that before issuing a broad policy, the NIH should pilot and evaluate a narrower use of single IRBs and provide appropriate resources to support the participating awardees. Others suggested that the NIH should fund research on existing central IRB models to evaluate potential benefits and costs before mandating single IRB review. A few commenters raised concerns about the timing of the policy in relation to the revisions of the Common Rule, stating their preference that the NIH not adopt a single IRB policy until Common Rule revisions have been finalized. However, other commenters praised the NIH for addressing the single IRB issue in the absence of an updated Common Rule. Finally, a few commenters discussed how the policy could be harmonized with other federal policies. One commenter recommended that the Office for Human Research Protections (OHRP) in the Department of Health and Human Services (HHS) provide guidance to support the policy’s stance on duplicative IRB review.
Stakeholders from academic institutions were concerned that the membership of any given sIRB would not be able to achieve the level of local support for a particular research study or its acceptability in terms of all the participating sites’ institutional commitments and regulations, applicable laws, and standards of professional conduct and practice. Some commenters contended that only a local IRB is able to understand the specific protections required for a vulnerable population that comprises their research participant base. Some suggested that site-specific practices for recruitment and retention, especially for vulnerable populations, would pose challenges for an sIRB. A number of commenters stated that their institutional IRBs are in the best position to know and understand competencies of and potential conflicts of interest of specific investigators. Others stressed the importance of the relationship between an investigator and the local IRB and noted that IRB members can serve as mentors to investigators whose protocols they oversee.
Some commenters asserted that the proposed policy does not recognize the time and effort needed to identify and establish a single IRB of record, including negotiating and executing authorization agreements and standard operating procedures, conducting study initiation meetings, creating account activities, and modifying information technology (IT) systems. They suggested that the policy would result in the formation of hundreds of different “single IRBs of record” with which institutions and investigators will need to interact. Some questioned whether an sIRB would be equipped to ensure local compliance at a relying institution and expressed the concern that a compliance problem for an sIRB would lead to compliance actions against the sites relying on that sIRB. Several commenters who supported the use of sIRBs recommended that rather than having participating sites identify a single IRB for each protocol, the NIH should establish a central IRB to review all multi-site research studies, akin to the National Cancer Institute’s Central Institutional Review Board (CIRB). They suggested that this approach would create an even “playing field” for every institution, big or small, regardless of whether their own IRB has the resources to act as a single IRB of record.
Many commenters, regardless of whether or not they supported the proposed policy, noted that over the past several decades, the IRB’s role has been expanded to include functions that go beyond ethical review of proposed research. For example, IRBs are often responsible for reviewing compliance with institutional policies, such as conflict of interest and investigator training. Commenters in favor of the proposed policy thought that greater use of sIRBs would help to return sIRB review to its primary mission of ensuring appropriate protections for human subjects rather than protecting the institution from legal liability or damage to its reputation. They also suggested that when institutions rely on a single IRB of record for multi-site research studies, IRB responsibilities are clearer, which helps institutions to develop policies and to provide resources beyond IRB review (e.g., human research protections experts) to facilitate compliance with the institutional human research protections program. Some commenters opposed to the proposed Policy suggested that the ancillary responsibilities of IRBs are so intertwined with the research oversight responsibilities that using a sIRB would disrupt the existing system of “checks and balances” at institutions. They also argued that the opportunity for the IRB to recommend protocol changes for reasons unrelated to ethical review (e.g., scientific improvements, changes to study design) would be lost.
Many commenters, regardless of whether they supported or opposed the proposed policy, made a number of specific practical suggestions about implementation. These are summarized below.
Most commenters supported a broad application of the policy to all studies involving the same protocol carried out at multiple sites in the U.S. These stakeholders stated that use of a single IRB of record for all types of studies and populations and study arrangements would encourage standardization of clinical research protocols and more effective implementation of protocols and protocol amendments. In contrast, a number of commenters suggested that the NIH should narrow the application of the policy or phase it in over time. Ideas about how the applicability of the policy should be narrowed were wide-ranging. Some stakeholders suggested that the level of risk should be a consideration in whether the policy should apply, with some pointing to minimal risk research and others to research involving more than minimal risk as being more appropriate for single IRB review. Others suggested that the policy should apply only to multi-site studies that involve a large number of sites (e.g., greater than 10); only to research involving clinical trials; only to studies carried out within established cooperative groups; or only to lengthy studies requiring an extended period of IRB oversight, e.g., three years or more. Some commenters suggested that the applicability of the policy remain broad, but that it be phased in over time.
The draft policy proposed exceptions only if the designated single IRB of record is unable to meet the needs of specific populations or where local IRB review is required by federal, tribal, or state laws or regulations. Most commenters agreed that there was a need to allow for exceptions to the ues of a single IRB. There were a number of comments calling for additional exceptions to those proposed in the policy. Commenters who generally supported the proposed policy stated that exceptions should be very limited. Some were concerned that a determination that the sIRB would be unable to meet the needs of specific populations was an overly subjective criterion or that institutions would routinely request exceptions asserting that the needs of specific populations could only be met by local IRBs. Tribal Nation commenters pointed to the importance of firsthand knowledge of local tribal customs, cultural values, and tribal sensitivities and supported exceptions to address those needs and also as a way of respecting tribal sovereignty. Other commenters said that the policy should allow for situational exceptions, depending on the types and complexity of studies and study teams, types and numbers of involved institutions, resources available for the sIRB (including IT resources), available resources for investigators, accreditation status of the human research protection program, or when study sites have concerns regarding the constitution of the designated reviewing IRB, that IRBs’ experience reviewing a particular type of research was inadequate, or if relying on the single IRB would affect the institutional IRB’s accreditation status.
Assuring Consideration of Local Context
Commenters were divided about the extent to which individual sites’ local contexts would present a challenge for an sIRB. Some commenters suggested that in today’s highly interconnected world, local contexts would not be unique or different enough to affect the review of research protocols. Others suggested that local context does vary, not only from state to state and community to community, but even among institutions serving the same community.
Commenters identified a number of capabilities that the sIRB would need to have in order to be effective, and one comment identified four such capabilities:
Selection of the IRB of Record
A number of commenters called on NIH to establish criteria or a minimum set of requirements to assist in the selection of the sIRB, as well as a need for criteria for an sIRB to use in its evaluation of participating sites. One commenter suggested that NIH’s Policy should require the applicant, offeror, or intramural investigator to justify their proposed sIRB. Since the NIH funding Institute or Center (IC) must approve the sIRB, one commenter suggested that NIH describe the criteria to be used in making a determination that the proposed sIRB is acceptable.
Some commenters offered specific suggestions for sIRB evaluation criteria. Suggestions for evaluation criteria included the following:
Defining IRB and Institutional Responsibilities
Many commenters pointed out the importance of defining the sIRB’s role and scope of responsibility in relation to the responsibilities of the participating research sites. These commenters noted that responsibilities of IRBs defined by the 45 CFR 46 often constitute only one part of institutions’ overall human research protections program. Commenters called on the NIH to establish a common approach to the division of responsibilities by providing model authorization agreements or even a uniform agreement that should be used in all cases. In addition to helping ensure a well-functioning review process, clear roles and responsibilities would, some suggested, also help mitigate concerns about added liability that an sIRB might assume.
A range of views were expressed relating to responsibilities that would be assumed by the sIRB and those that would remain with participating sites. Some commenters suggested that in addition to fulfilling the requirements set out in 45 CFR 46, i.e., conducting initial and continuing reviews of protocols, amendments, unanticipated problems, protocol deviations, and required regulatory IRB reporting, sIRBs should adopt some of the responsibilities that are frequently delegated to local IRBs, in particular, acting as a privacy board for all sites. One commenter noted that systems would be required to ensure that duplicative reviews are not conducted by the sIRB and local IRBs, and several commenters expressed concerns about the difficulty of coordinating required sIRB reviews with additional reviews that are not required by regulation, such as reviews for conflict of interest, investigator qualifications, and scientific merit. Some of these commenters questioned how sIRB reviews required by the HHS regulations should be coordinated with other required reviews that may have been delegated to the local IRB. These commenters noted that many institutions have established systems and standard operating procedures for coordinating local IRB review with other required reviews, such as institutional biosafety reviews, radiation safety reviews, pharmacy reviews, reviews required by state or local laws, post-approval monitoring and for-cause auditing purposes, and research billing. One commenter suggested that sIRBs should not be responsible for adverse event reporting. Another commenter suggested that sIRBs should be responsible for maintaining databases of relevant state laws. In addition, a small number of commenters indicated that the regulations of other Common Rule agencies, FDA in particular, may create contradictory requirements, and called for clarification and a more unified approach.
Several commenters stated that coordinating these additional reviews with sIRB reviews would limit the gains in efficiency realized from reliance on an sIRB. One commenter recommended that the NIH develop a template IRB authorization agreement and guidelines to define the institutional obligations that are distinct from the IRB review responsibilities. Another commenter recommended that the NIH publish guidance delineating the specific regulatory requirements for which the sIRB would be responsible, shared responsibilities, and responsibilities that an sIRB could negotiate with IRBs at participating sites.
Resources and Funding
Several commenters described the proposed policy as an unfunded mandate, or stated that it would result in a shifting of expenses from one institution to another. Many commenters expressed the concern that if costs associated with using a single IRB are taken from a participating institution’s indirect costs, there would be insufficient funds for the local Human Research Protection Program (HRPP) that still has institutional oversight responsibilities, even if the IRB of record is external. Most commenters with experience using a single IRB of record for multi-site research studies recommended that indirect costs remain unchanged for relying institutions in order to ensure that the human research protections infrastructure are available for institutional responsibilities, e.g., post-approval compliance monitoring, conflict of interest reviews. Many commenters noted funding sIRBs through indirect costs would divert funds required to conduct research and serve as a disincentive to conducting multisite research. The majority of commenters stated a preference for including the additional costs associated with a single IRB review in the study budget as direct cost, although one commenter stated a preference that sIRB review be included as an indirect cost in order to maximize the amount of funding available for research.
Several commenters stated that the costs and resources needed to establish sIRBs were not addressed by the proposed policy. Infrastructure needs noted by these commenters included additional staff and/or staff time to perform sIRB-related activities, costs to create or adapt electronic managements systems that are interoperable with outside institutions, and the time and cost of developing communication tools to link investigators to IRBs outside their institution. Other commenters familiar with the operations and use of sIRBs noted that while initial financial support from the NIH may be required to establish or expand the capacity of some IRBs to serve as the IRB of record, most sIRBs should be able to become self-supporting eventually.
Commenters had questions about whether plans for single IRB review would be required in grant applications and how plans would be reviewed.
Need for Implementation Guidance
A number of commenters pointed out how important it would be for the NIH to provide practical guidance to facilitate the implementation of the policy, with some commenters stating that, in the absence of such guidance, burden and costs would only shift between institutions rather than adding efficiency to the IRB process. A few commenters noted that this guidance could be developed using the experiences of IRBs that have already implemented centralized IRB review processes.
In addition to general requests for implementation guidance, a number of commenters made specific guidance suggestions. These suggestions included the need for guidance covering:
In addition, commenters requested:
Implementation of the Policy
In developing the final policy set out below, the NIH carefully considered the many thoughtful comments we received on the Draft NIH Policy on the Use of a Single Institutional Review Board (IRB) for Multi-Site Research (NOT-OD-15-026). While we found no compelling reason to narrow the essential scope of the final policy—it will cover all domestic sites of NIH-funded non-exempt multi-site studies as was proposed—we have clarified the policy intent and modified several provisions. The final policy is intended to apply only to studies where the same research protocol is being conducted at more than one site; it does not apply to studies that involve more than one site but the sites have different roles in carrying out the research. Applicants/offerors will be expected to submit a plan identifying the sIRB that will serve as the IRB of record for all study sites. It will be the responsibility of the applicant/offeror to assure that the sIRB is qualified to serve; the applicant’s plan will not be evaluated in peer review. The additional costs associated with sIRB review may be charged to grants or contracts as direct costs, provided that such costs are well-justified and consistently treated as either direct or indirect costs according to applicable cost principles in the NIH Grants Policy Statement and the FAR 31.202 (Direct Costs) and FAR 31.203 (Indirect Costs). Exceptions to the policy will be granted, as was proposed, if the use of an sIRB is prohibited by federal, state, or tribal laws or regulations. We will also grant exceptions where the federal, state, or tribal prohibition on the use of an sIRB is established by policy, and we will consider granting an exception if a request is made and a compelling justification provided for why an exception is needed. Such justifications could be for reasons other than that the sIRB is unable to meet the needs of a specific population, as was proposed in the draft policy. The final policy also clarifies that multi-site studies within ongoing, non-competing awards will not be expected to comply with the policy until a competing renewal application is submitted.
The NIH recognizes that the policy will begin a paradigm shift in IRB review. As such, the final policy will not take effect until May 25, 2017. In the interim, the NIH will issue guidance and provide resources to assist awardees in adapting to the shift.
Guidance on how costs associated with sIRBs may be charged as direct versus indirect costs can be found in Guide Notice NOT-OD-16-109 Other guidance materials will be issued before the policy’s effective date and posted along with the policy on the following site: http://osp.od.nih.gov/office-clinical-research-and-bioethics-policy/clinical-research-policy/models-irb-review. Among other topics, the guidance will address:
Finally, while the NIH anticipates that that there will be challenges associated with implementation, we expect these to be short-lived. Once the transition to the new way of operating is made, the benefits of widespread use of sIRBs will outweigh any costs and, ultimately, reduce burdens to the research process. At the same time, the NIH will also closely monitor the implementation of the policy, consider its impact on research such as improvements in time to initiation of research and reduction of unnecessary burden, and be vigilant about any diminution in the protection of human subjects.
National Institutes of Health Policy on the Use of a Single Institutional Review Board for Multi-Site Research
Final NIH Policy on the Use of a Single Institutional Review Board for Multi-Site Research
The National Institutes of Health (NIH) Policy on the Use of a Single Institutional Review Board of Record for Multi-Site Research establishes the expectation that all sites participating in multi-site studies involving non-exempt human subjects research funded by the National Institutes of Health (NIH) will use a single Institutional Review Board (sIRB) to conduct the ethical review required by the Department of Health and Human Services regulations for the Protection of Human Subjects at 45 CFR Part 46. This policy, which is consistent with 45 CFR Part 46.114, is intended to enhance and streamline the process of IRB review and reduce inefficiencies so that research can proceed as expeditiously as possible without compromising ethical principles and protections for human research participants.
Scope and Applicability
This policy applies to the domestic sites of NIH-funded multi-site studies where each site will conduct the same protocol involving non-exempt human subjects research, whether supported through grants, cooperative agreements, contracts, or the NIH Intramural Research Program. It does not apply to career development, research training or fellowship awards.
This policy applies to domestic awardees and participating domestic sites. Foreign sites participating in NIH-funded, multi-site studies will not be expected to follow this policy.
Consistent with the Roles and Responsibilities section, applicants/offerors will be expected to include a plan for the use of an sIRB in the applications/proposals they submit to the NIH. The NIH’s acceptance of the submitted plan will be incorporated as a term and condition in the Notice of Award or in the Contract Award. This policy also applies to the NIH Intramural Research Program.
The Authorization Agreement, which is also called a reliance agreement, is the agreement that documents respective authorities, roles, responsibilities, and communication between an institution/organization providing the ethical review and a participating site relying on the sIRB.
A multi-site study uses the same protocol to conduct non-exempt human subjects research at more than one site.
Participating site in a multi-site study is a domestic entity that will rely on the sIRB to carry out the site’s IRB review of human subjects research for the multi-site study.
sIRB is the selected IRB of record that conducts the ethical review for participating sites of the multi-site study.
Roles and Responsibilities
Applicant/Offeror. In the application/proposal for research funding, the applicant/offeror is expected to submit a plan describing the use of an sIRB that will be selected to serve as the IRB of record for all study sites. The plan should include a statement confirming that participating sites will adhere to the sIRB Policy and describe how communications between sites and sIRB will be handled. If, in delayed-onset research, an sIRB has not yet been identified, applications/proposals should include a statement that awardees will follow this Policy and communicate plans to use a registered IRB of record to the funding NIH Institute/Center prior to initiating a multi-site study. The applicant/offeror may request direct cost funding for the additional costs associated with the establishment and review of the multi-site study by the sIRB, with appropriate justification; all such costs must be reasonable and consistent with cost principles, as described in the NIH Grants Policy Statement and the Federal Acquisition Regulation (FAR) 31.302 (Direct Costs) and FAR 31.203 (Indirect Costs).
Awardees. Awardees are responsible for ensuring that authorization agreements are in place; copies of authorization agreements and other necessary documentation should be maintained in order to document compliance with this policy, as needed. As appropriate, awardees are responsible for ensuring that a mechanism for communication between the sIRB and participating sites is established. Awardees may delegate the tasks associated with these responsibilities.
Funding Institute or Center (IC). Funding ICs are responsible for management and oversight of the award, including communicating with the awardee regarding the implementation of its proposed plan to comply with the sIRB Policy. In the event that questions arise about the awardee’s plan, including the IRB that has been selected to serve as the sIRB, the funding IC will work with the awardee to resolve them.
sIRB. The sIRB is responsible for conducting the ethical review of NIH-funded multi-site studies for participating sites. The sIRB will be expected to carry out the regulatory requirements as specified under the HHS regulations at 45 CFR Part 46. In reviewing multi-site research protocols, the sIRB may serve as a Privacy Board, as applicable, to fulfill the requirements of the HIPAA Privacy Rule for use or disclosure of protected health information for research purposes. The sIRB will collaborate with the awardee to establish a mechanism for communication between the sIRB and the participating sites.
Participating Site. All sites participating in a multi-site study are expected to rely on an sIRB to carry out the functions that are required for institutional compliance with IRB review set forth in the HHS regulations at 45 CFR 46. Participating sites are responsible for meeting other regulatory obligations, such as obtaining informed consent, overseeing the implementation of the approved protocol, and reporting unanticipated problems and study progress to the sIRB. Participating sites must communicate relevant information necessary for the sIRB to consider local context issues and state/local regulatory requirements during its deliberations. Participating sites are expected to rely on the sIRB to satisfy the regulatory requirements relevant to the ethical review. Although IRB ethical review at a participating site would be counter to the intent and goal of this policy, the policy does not prohibit any participating site from duplicating the sIRB. However, if this approach is taken, NIH funds may not be used to pay for the cost of the duplicate review.
Exceptions to this policy will be made where review by the proposed sIRB would be prohibited by a federal, tribal, or state law, regulation, or policy. Requests for exceptions that are not based on a legal, regulatory, or policy requirement will be considered if there is a compelling justification for the exception. The NIH will determine whether to grant an exception following an assessment of the need.
This policy applies to all competing grant applications (new, renewal, revision, or resubmission) with receipt dates on or after May 25, 2017. Ongoing, non-competing awards will not be expected to comply with this policy until the grantee submits a competing renewal application. For contracts, the policy applies to all solicitations issued on or after May 25,, 2017. For the intramural program, the policy applies to intramural multi-site studies submitted for initial review after May 25, 2017.
Please direct all inquiries to:
NIH Office of Science Policy