Secretary’s Advisory Committee on Human Research Protections
July 21-22, 2009 - Arlington, VA
TUESDAY, JULY 21
Barbara Bierer, M.D., Chair
Dr. Bierer thanked Dr. Genel and Dr. Botkin, who are concluding their terms of service as SACHRP members, for their many contributions. She also expressed appreciation to OHRP staff members Julia Gorey and Cecilia Chirinos for their work in preparing this meeting.
The Chair reported that SACHRP’s 10th letter containing recommendations has been submitted to the Secretary of Health and Human Services (HHS) for review.
Minutes for the previous meeting (March, 2009) were approved unanimously.
Dr. Bierer welcomed Dr. Koh, who has been confirmed as the Assistant Secretary for Health (ASH). She thanked him for attending the meeting and offering his remarks.
Howard K. Koh, M.D., M.P.H.
Dr. Koh thanked SACHRP members for their service and stressed SACHRP’s importance as an advisory committee. He said the committee has produced many important recommendations, and HHS values its members’ expertise. Calling the meeting agenda “impressive,” he found the topics to be addressed “of great interest.”
The Assistant Secretary of Health shared his own passion for improving research protections for human subjects, noting that the complex challenge “captures the essence of public health” as a “critical intersection” of multiple dimensions, including law, medicine, and ethics. He pledged to lend his own efforts to support SACHRP’s work.
Dr. Koh praised and expressed appreciation for the public service and many contributions of departing members Dr. Botkin and Dr. Genel.
Issues and Remarks from the OHRPDirector
Jerry Menikoff, M.D., J.D., Director, OHRP
Dr. Menikoff reported that much of OHRP’s current work on policy and guidance development is based on SACHRP’s recommendations. He stressed that OHRP wants to see SACHRP’s work “come to fruition.” Work in progress includes developing new guidance on continuing review, which is currently being prepared for internal government review; drafting guidance on research involving children; preparing to draft guidance on expedited review; and determining next steps in regard to the training and education of those involved in human subjects protection.
The Director added that in connection with Congressional hearings on for-profit Institutional Review Boards (IRBs), OHRP has clarified through its testimony what is involved in a Federalwide Assurance (FWA) and what it means to have a well-functioning IRB that is fulfilling its mandate. Dr. Menikoff’s testimony of March 26, 2009 may be accessed at: http://www.hhs.gov/asl/testify/2009/03/t20090326e.html.
Dr. Menikoff expressed OHRP’s appreciation for the “amazing work” of both departing members, Dr. Botkin and Dr. Genel. He called both “people of character.”
Dr. Strauss said that as a member of the Subpart A Subcommitteee (SAS) work group on continuing review, he saw the area as a major priority and was glad to know that OHRP agreed.
Julie Kaneshiro, Team Leader, Division of Policy and Assurances, OHRP; Darrell Wilson, M.D., Professorand Chief of Pediatric Endocrinology and Diabetes, Stanford University and the Lucile Packard Children's Hospital at Stanford; Susan Loess-Perez, M.S., Director, Office of Research Protections, DePaul University; Felix Gyi, PharmD, M.B.A., Chief Executive Officer, Chesapeake Research Review, Inc.; David Wynes, Ph.D., Vice President for Research Administration, Emory University
Ms. Kaneshiro reported on OHRP’s analysis of responses to its Advanced Notice of Proposed Rulemaking (ANPRM) on Holding External Institutional Review Boards Directly Accountable. The ANPRM was published on March 5, 2009 and may be found at: http://edocket.access.gpo.gov/2009/E9-4628.htm. Comments were due on June 3, 2009. OHRP received 30 comments, of which 13 were “joint” comments submitted on behalf of multiple entities. Ten comments came from individuals; eleven from nongovernmental organizations and private-sector entities; and nine from universities, investigators, or IRBs.
The speaker noted that OHRP cosponsored two meetings on alternative models of review for IRBs, in November 2005 and November 2006, in response to a SACHRP recommendation. A key conclusion of both meetings was that some institutions are reluctant to designate external IRBs and rely on cooperative review arrangements because, should there be any issue of noncompliance on the part of an external IRB listed on the institution’s FWA, OHRP currently holds the institution that is actually “engaged in human subjects research” accountable for noncompliance. Reports from the 2005 and 2006 meetings on alternative IRB models may be found at these sites:
Ms. Kaneshiro reported responses to the key question posed by the ANPRM – whether a regulatory change is needed to address this problem – as shown in the table below. In the presentations that follow, Dr. Wilson and Ms. Loess-Perez present the viewpoint of the two organizations that opposed a regulatory change.
Responses for and Against a Regulatory Change Regarding Accountability
Should HHS pursue regulatory change to hold IRBs/IORGs directly accountable for compliance with certain requirements of 45 CFR part 46?
Non-Governmental Organizations, Private Sector
Responses to additional specific questions were as shown in the table below. Note that not all respondents addressed every question. For typical or notable comments, see the PowerPoint for the presentation posted on this site.
Responses to Other Specific Questions
Question 1. Is there sufficient need for HHS to pursue a regulatory change to 45 CFR part 46 to hold IRBs/IORGs directly accountable?
Question 2.Would the proposed regulatory change reduce concerns about regulatory liability and contribute to an increase in collaborative IRB review arrangements?
3 no; 4 mixed
Question 3. Are there other approaches and strategies that would decrease concern about regulatory liability and increase collaborative IRB review arrangements?
Question 4. Would such a regulatory change have the unintended effect of making IRB organizations less willing to have their IRB designated on other institutions’ FWAs or have other unintended effects? If so, would there still be sufficient benefit to pursue the proposed regulatory change? Other possible negative consequences?
Question 6. How should regulatory requirements be categorized? Regulatory requirements seem to fall into three categories of responsibilities:
Are these categories appropriate? Is there a category of responsibilities that are inherently shared by the IRBs/IORGs and engaged institutions? Are the regulatory provisions identified under each category appropriate? How have institutions engaged in research and external IRBs/IORGs divided or shared regulatory responsibilities in the past?
Question 7a. If this regulatory change is pursued, should OHRP use the IRB Authorization Agreement or other forms of agreement to inform its compliance oversight evaluations?
Question 7b. Should there be new provisions that require specific content for agreements between external IRBs and FWA-holding institutions?
Question 7c. Should the regulation describe which regulatory requirements would need to be met by external IRBs versus institutions engaged in research?
Question 5, which reads as follows, was not phrased to elicit a yes-no response: If such a regulatory change were pursued, what kinds of administrative actions would be appropriate for OHRP to take against IRBs found to be out of compliance? Responses included the assertion that OHRP enforcement should be the same as FDA’s and that current OHRP sanctions are insufficient. The respondent making the latter comment suggested that OHRP seek the authority to issue fines for noncompliance if it does not have this authority at present.
In summary, the OHRP staff member said the majority of respondents favored pursuing a modification of the regulations. Most felt this would lessen concern about regulatory liability, thereby removing a barrier to use of cooperative arrangements and external IRBs. Respondents differed, however, in the extent to which they felt OHRP should specify which of the institution’s responsibilities should and should not be delegated; many felt this should be left to the discretion of the institution based on the nature of the study and other variables.
Dr. Wilson noted that while use of a Central IRB (CIRB) might theoretically be able to improve the otherwise “cumbersome and inefficient” approach to evaluating and approving multisite trials, his experience as the Chair of one of Stanford University’s IRB panels is that there are still many issues with CIRBs. Examples of such issues include:
- Each CIRB agreement is currently idiosyncratic in its structure, so that a single institution may have to interact with a variety of central IRBs with different roles and responsibilities;
- The current CIRB system is more difficult to manage locally than the overly redundant method of review by individual sites;
- It is not clear what should happen when one out of a hundred institutions disagrees with a CIRB
- Central IRBs have been shown to make errors, which, from the point of view of the local institution, defeats the purpose;
- Clear, dependable allocation of so-called “liability” is important.
The speaker proposed that OHRP develop one or more robust central IRB templates that clearly allocate the various responsibilities between the central and local IRB. The speaker believed that most responsibilities should generally be allocated to the central IRB. These agreement templates would provide a “safe harbor” for those local IRBs who use the CIRB process.
Discussion: Dr. Botkin noted that in addition to concerns related to regulatory liability, there is concern about the institution’s fiduciary responsibility to people who trust it to provide responsible oversight. He added, however, that from his perspective, an idiosyncratic response to risk characterization on the part of one out of a hundred institutions in a multi-site study is not necessarily a disadvantage. Finally, he noted that conflict of interest, radiological health, biosafety, and other issues must also be addressed.
The speaker emphasized that IRB review is only one part of human subject protection responsibilities. He believed that radiological hazards, for example, could be adjudicated centrally, but some responsibilities would remain with the institution. He repeated, however, that he was concerned that an IRB that has agreed to participate in a study might reach a different finding but not have a chance to present its reasoning. Dr. Strauss said he also found it troubling that a local IRB would be called upon to “rubber stamp” a decision.
Dr. Strauss asked whether the creation of CIRBs might evolve into a form of “IRB shopping” in which institutions that tend not to concur with the CIRB are less likely to be included in studies. The speaker noted that institutions with efficient and effective processes may be favored today, and this is understandable. He added that a respected institution with a well-constructed IRB might be asked to function as a CIRB for a study and would be more easily accepted than an unknown entity.
Ms. Loess-Perez was not sure that a regulatory change is needed in this area; rather, she felt that detailed and meaningful guidance with practical examples would be appropriate. She suggested that before implementing a regulatory change, OHRP should:
- Clarify what is meant by holding an IRB accountable. Would IRB members/chairs be held accountable?
- Clarify to what kinds of agreements the change would apply. How, for example, would it affect agreements between two FWA-holding institutions?
The speaker argued that shifting liability concerns to another institution will not necessarily increase the use of collaborative agreements for research, since other disincentives would remain. She noted that there is a “lot of work involved” in having a collaborative agreement, and small institutions, in particular, may find this is not worth the effort. For example, additional effort is required to create the agreement, obtain and track protocol documents, and coordinate a review process that is different from the established internal process. Like Dr. Wilson, she called for sample templates for IRB authorization agreements that are appropriate for differing types of relationships and divisions of responsibilities.
Dr. Gyi, the founder and CEO of Chesapeake Research Review, Inc., suggested the use of the term “professional IRB” rather than “commercial” IRB. He observed that many institutions find a “huge competitive advantage” in outsourcing IRB oversight. For example, they may find efficiencies and economies of scale, as well as being able to increase the size of their research portfolios without increasing the overhead burden assumed by the institution. Like the previous speakers, Dr. Gyi underlined the importance of clarifying roles and responsibilities, such as who will research allegations of noncompliance and misconduct. He also highlighted the need for a clear communication plan that addresses routine communications, “for cause” issues, and communications on urgent matters that require immediate attention. He added that the institution always retains some responsibilities for protecting subjects’ rights and welfare.
Dr. Gyi’s view was that as a regulatory issue, liability is a “red herring”: the U.S. Food and Drug Administration (FDA) already holds each entity to its defined responsibilities. Therefore, he suggested that OHRP develop similar guidance that establishes OHRP’s intention to hold each entity that commits to an agreement related to protocol review responsible for fulfilling its roles and responsibilities. Dr. Gyi observed that in his experience, no one model works for everyone, so the guidance should give institutions flexibility to establish appropriate models and agreements.
As an Institutional Official (IO) at Emory University, a large research university, Dr. Wynes strongly endorsed the principles described in the ANPRM. He noted what while multiple factors impact an IO’s decision to rely on an external IRB or an IRB Organization (IORG), this ANPRM addresses a key factor in those decisions. He suggested that some responsibilities should remain with the institutional IRB, including:
- Written procedures for initial and continuing review,
- Review more often than annually or when independent verification is required,
- Prompt reporting of changes to the IRB,
- Informing the investigator when research is suspended or terminated by the IRB, and
- Fulfilling IRB recordkeeping requirements.
Many other responsibilities, however, could be addressed on a case-by-case basis. Examples include addressing serious or continuing noncompliance and reporting IRB determinations of unanticipated problems involving risks to subjects or others.
Dr. Wynes stressed the importance of addressing the issue surrounding liability. He reported that OHRP had advised him that when using an external IRB, his institution would still have to self-report noncompliance if an issue arose in that regard.
Responsibilities for human subject protection. Dr. Genel stressed the importance of separating responsibilities for protocol review from those for oversight and compliance. He held that it was the responsibility for protocol review that would typically be delegated. Dr. Wilson agreed, but added that the approval of the consent form is embedded within the protocol review process; revisions to the consent form are handled differently depending on the review model.
Conflict of interest and other determinations. Dr. Botkin asked how Chesapeake addresses the question of conflict of interest. Dr. Gyi responded that the professional IRB has an internal conflict of interest panel that addresses any issues that might arise among its own reviewers. He feels that the institution is generally best equipped to look at its own internal conflicts and usually does so before the protocol comes to Chesapeake. Conclusions are reviewed by Chesapeake, which also considers any conflict of interest issues that might compromise the investigator.
A SACHRP member asked how considerations related to issues such as conflict of interest would affect the time available for protocol review. Dr. Gyi responded that such issues can be accommodated without increasing the normal time required for review. However, another panel member observed that preparing a detailed memorandum of understanding (MOU) for protocol review can take longer than the review itself.
Best practices in local context. A SACHRP member asked whether there are emerging best practices and mechanisms to address sensitivity to local context. Dr. Gyi responded that in his experience, a professional IRB can help institutions address related regulatory requirements. He remarked that many institutions do not understand the nuances of the regulations in this regard.
Guidance vs. legislation. A panel member asked whether OHRP had considered developing guidance instead of legislation to address the liability issue. Ms. Kaneshira responded that OHRP and legal counsel continue to consider the issue of whether the regulations would allow the issue of roles and responsibilities to be addressed through guidance.
Personal liability. In regard to personal liability issues, Dr. Gyi noted that a number of insurance products are available to ensure that individual IRB members cannot be held personally liable for IRB decisions. OHRP staff emphasized that the possible regulatory changes contemplated in the ANPRM were not intended to have any impact on the liability of individual IRB members.
Variety of models. Dr. Strauss underscored that the panelists had referred to several different models of review. He suggested that large institutions might be forced to grapple with a number of different standards of practice. Dr. Wynes responded that he would not envision one investigator working with a dozen different models, though there might be different models and procedures for various types of research. Dr. Bierer added that the new regulations or guidance would need to allow for flexibility of approach. She also noted that a standing research committee could make many such decisions on behalf of the institution so the burden does not fall on the IRB.
Dr. Wilson suggested that a small number of OHRP-approved templates might address most types of arrangements. Ms. Kaneshiro added that it may be possible to separate roles and responsibilities into three “buckets,” the third of which would be those that could be handled by either party.
Steven Hirschfeld, M.D., Ph.D., Captain, U.S. Public Health Services and Associate Director for Clinical Research, National Institute of Child Health and Human Development
Dr. Hirschfeld reported the experience of Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) with IRB review for multi-site pediatric clinical studies. (See http://www.nichd.nih.gov.) Approximately 60 networks and consortia are fully or partially supported by NICHD.
The National Center for Research Resources oversees a program of Clinical and Translational Science Awards (CTSA), which have allowed 40 institutions to date to engage in a program to develop a national infrastructure for clinical and translational research. See: https://www.ctsacentral.org/.
In the experience of NICHD, review by multiple IRBs results in multiple inconsistencies in review, as well as delays in activating studies and reaching accrual targets. As a step in addressing this concern, in 2007, a CTSA-NICHD workshop presented IRB members from different institutions with hypothetical case studies to determine whether they could reach consensus. The results were encouraging. The workshop was recorded and archived: https://webmeeting.nih.gov/p37176252/.
On April 23, 2009, a workshop/Webinar provided an opportunity to explore models for multi-site pediatric review. Speakers who presented models they were familiar with were asked to address the following:
- Resources required for the IRB to function properly;
- Availability of pediatric expertise on the IRB: permanent or ad hoc?
- Average time for protocol review;
- Range of time for protocol review;
- Proportion of protocols that proceed without major revisions;
- Proportion that have specific stipulations that require major revisions before proceeding with enrollment;
- Monetary cost for protocol review and who pays;
- Opportunities and policies to consult with other IRBs;
- Level of interaction with local site IRBs for multi-site studies;
- The process for addressing local site IRB amendments or stipulations; and
- The process for safety monitoring.
The conference was recorded and archived at the following address: http://videocast.nih.gov/Summary.asp?File=15053
NICHD is engaged in developing a model for IRB review that will be implemented and assessed at selected CTSA sites. Member IRBs will be asked to sign on to a compact that outlines principles, processes, and performance standards for review of multi-site pediatric studies. The underlying assumption is that a shared understanding of principles, process, and performance can build consistency and trust. Under the proposed Federation Model, a Protocol Operations Center will coordinate the review. An initial IRB will review the proposed protocol, and each subsequent site will receive the comments of the initial reviewing IRB and any additional comments from other sites that reviewed the protocol. Amendments based on these comments will be communicated to all IRBs by the Protocol Coordinator. The model will be tested experimentally at several volunteer CTSA sites to elicit feedback.
Scope of application for the Federation Model. Dr. Genel asked how many sites would be involved. Dr. Hirschfield explained that because over 40 IRBs could eventually participate in protocol review, an operations center was envisioned. For the pilot, 6-12 institutions will be involved to test the model’s feasibility and metrics.
The speaker explained that it is not yet clear whether the model will be used by all CTSA participants or for certain protocols. Consideration of primary review will work on consensus principles, and it is not clear at present how it will evolve and how current networks will be affected.
Comparison to National Children’s Study. Dr. Botkin asked how the Federation Model compared to the review model used for the National Children’s Study (NCS), which uses a similar process. The speaker responded that key differences include the addition of a specific agreement regarding the process, principles, and performance aspects of protocol review. Also, the NCS is focused on one dynamic protocol with evolving amendments, while the Federation Model will not be limited to a particular therapeutic area or protocol source.
Input from “downstream” sites. Dr. Botkin also wondered whether “downstream” IRBs would be expected to do a “thumbs up/thumbs down” protocol review rather than “tweaking” the proposal. Dr. Hirschfield stressed the need for a balance between quality control, which includes the recognition that a reviewing site may find important issues not raised previously, and the need for a focus on human subjects protection within an effective and efficient process. He felt the scientific review could take place in other settings.
Dr. Strauss was concerned that the speaker was implying that IRBs should not be concerned about science. He stressed that an IRB is not exercising its responsibility if it does not look at inclusion and exclusion criteria, which are key to efforts to minimize risks. He held that IRBs need to have the opportunity to provide meaningful input. Dr. Hirschfield said he had in mind IRB reviews that offer lengthy suggestions for what kind of assay should be used; he agreed that eligibility criteria are central to human subjects protection. He felt, however, that the IRB’s review should focus on what is being done with and to people.
Power-sharing process. Dr. Marshall asked about the observed power dynamics in the groups brought together to seek consensus. Dr. Hirschfield responded that if questions were framed to focus on the process of resolving issues and reaching consensus, this provided a neutral space where participants could reach consensus.
IRB summary review. A SACHRP member asked about how the findings of IRBs would be summarized. The speaker said the plan is to develop a cover sheet that captures topics in the discussion in the same sequence, making it easy to process and produce. Each IRB will receive the entire protocol, and each will be able to decide whether to pursue an expedited review, a full discussion, or a consultation, so long as whatever procedure followed uses the agreed-on principles.
Evaluation. Dr. Strauss asked the speaker to comment on the evaluation envisioned for the Federation Model. Dr. Hirschfield said the evaluation process is in development, but a multidimensional approach is envisioned to understand the details of the assessment process and the adequacy of the review.
Elyse Summers, J.D., Acting Director, Division of Education Development, OHRP
Ms. Summers reported on the input received from a July, 2008 Request for Information (RFI) that asked the following key questions:
- Whether HHS should develop a regulation requiring that institutions engaged in human subjects research conducted or supported by the Department of Health and Human Services (HHS) implement training and education programs for certain individuals involved in the conduct, review, or oversight of human subjects research; and
- Whether OHRP should issue additional guidance recommending the implementation of such training and education programs.
Of the almost 100 comments received, six could be considered “joint” responses that represented more than one entity. Others included 16 from nongovernmental organizations (NGOs); 48 from Universities, Health Systems, Hospitals, and IRBs; and 32 from others (e.g., individuals, private sector entities, and government). In regard to new regulations, she reported the following:
- 24 comments were explicitly against new regulations. Most comments from organizations were in this category.
- 36 comments were for new regulations.
- The remainder did not explicitly state support or opposition to new regulations.
In regard to proposed new OHRP guidance on the subject, comments were as follows:
- 7 comments were against new guidance, all of which came from specific institutions;
- 25 comments were in favor of new guidance; and
- 63 comments did not explicitly state support or opposition to new guidance (e.g., the narrative described general support for education).
In regard to whether or not OHRP should stipulate specific content for education, of the 26 respondents that specifically addressed the question, 20 advocated generally for clear guidance describing expected content in broad areas, with allowance for institutional flexibility; 6 held that institutions should be given full authority to determine content of education programs.
In general, the SACHRP Chair summarized following the presentation, “everyone is in favor of education, but no one is in favor of regulation.”
A SACHRP member expressed appreciation for the RFI, which was consistent with SACHRP recommendations, and looked forward to hearing OHRP’s next steps. Several questions were posed by SACHRP members following the presentation.
How does OHRP plan to respond to this input? Dr. Menikoff said OHRP is deliberating next steps.
Why were respondents opposed to regulations? While many of them applauded OHRP for recognizing the importance of education, they felt that thoughtful guidance would be the best option. They generally welcomed assistance in putting together training programs but felt regulation was unnecessary.
What fraction of the regulated community responded? OHRP staff could not guess, but said that over 10,000 FWAs have been issued.
If you look at compliance findings, how many are attributable at least in part to a lack of education? OHRP staff could not guess, but noted that OHRP often recommends education and training as part of required corrective actions.
Does FDA require investigator training? FDA’s requirements are broad. Investigators must be qualified to assume supervisory responsibilities. Medical expertise and knowledge of clinical practice are expected.
A member suggested that FDA and OHRP explore harmonization in the area of educational requirements. The OHRP Director said he planned to discuss any further action with FDA and was very aware of the need for harmonization.
Dr. Bierer observed that the lasting impact of training and certification in these areas has not been studied, and it is not clear how they affect compliance or improve performance and attention to detail.
The Chair invited members and ex officios to provide input on future planning for SACHRP. This included consideration of the focus for a possible new subcommittee (since the Subcommittee for the Inclusion of Individuals with Impaired Decision-making in Research [SIIIDR] concluded its work at the previous meeting) and subjects for future panels.
Dr. Bierer began her own suggestions by reviewing the committee’s charge and reviewing its accomplishments. She noted that there is a specific reference to investigator conflicts of interest in SACHRP’s charter, although the Common Rule is essentially silent on the issue. She felt that informed consent is an essential topic to continue to pursue, but it is fully within the scope of the existing Subcommittee on Subpart A. She identified the following questions were key:
- What could, upon review, present further protections for human subjects in research?
- What could, upon review, decrease the risk or burden of participation in biomedical, and behavior and social science, research?
- What of the burden, not on human participants, but on IRBs and the oversight system itself? What research is not getting initiated because of regulatory burden?
Dr. Bierer observed that the burden placed on IRBs and the oversight system is not only a consequence of an arcane and complicated regulatory framework. The burden is aggravated by the fact that approximately 25% of IRB approved studies fail to enroll a single participant and approximately 40% of IRB approved studies enroll fewer than 5 participants. It has been estimated that the cost of IRB approval for studies that enroll 5 or fewer participants is $550M a year, every year. A further symptom of the problem is the length of time from submission of a protocol to approval (which is not to say that a metric for IRB effectiveness is speed). This, she held, is wasteful effort that contributes to the burden upon IRBs, that frustrates investigators, and that inhibits critical translational research in its broadest interpretation.
In conclusion, the Chair stressed that efforts should now be directed not towards eliminating independent review, but rather towards understanding the essential elements of independent review that do protect human participants. The Chair held that the guideposts that IRBs utilize to consider and adjudicate risk and benefit should not differ based on funding source or specific agent studied. Efforts should be directed towards harmonization and simplification of the Common Rule and the FDA, and importantly, of the application of Health Insurance Portability and Accountability Act (HIPAA) regulations, which were initially directed towards clinical care in billing and insurance coverage, to research. Members of the proposed subcommittee would need to have content knowledge, practical experience, creativity, and flexibility. She reminded members that such an effort has been suggested repeatedly by investigators, by IRB members and administrators, by a variety of public and private organizations, and by patient advocacy groups representing the public. Success would require not only cooperation, but a collective commitment to streamlined interpretation. It should:
- Address barriers to multisite and multicenter research,
- Seek to eliminate expensive and often redundant reviews,
- Focus on the benefits of cooperation,
- Address the benefits and challenges of international research, and
- Promote appropriate and aligned oversight by a variety of regulatory bodies.
In the end, the Chair held, such an effort will promote timely behavioral, social science, and biomedical research that will improve understanding of human biology and action and of the pathogenesis and treatment of disease, and that will contribute to the health and well-being of the American public.
Ms. Pattee stressed the importance of supporting alternative models for IRB reviews, including harmonization on the issue of external IRBs. She also noted the importance of harmonization between FDA and OHRP, through legislation if necessary. Finally, she was interested in efforts to measure the effectiveness of IRBs and promote quality improvement.
Ms. Bankert pointed to a “huge bottleneck” that delays approval of protocols that are “no more than minimal risk,” sometimes leading to the inability to complete the study. She wondered, “what pushes the paperwork process?” She was also curious about the implications of trial funding, i.e., funding by industry or by the Federal government. She said the issue of multi-site review should be part of the dialogue on these issues.
Dr. Genel stressed five priorities (not in order of importance):
- Harmonization of the Common Rule, FDA regulations, and HIPAA;
- Implementation of SIIIDR recommendations on individuals with impaired decisionmaking;
- Enhancement/development of centralized protocol review mechanisms;
- Helping secure appropriate OHRP resources to enable it to fulfill its mission;
- Paying attention to IRB review and oversight issues, including those related to independent IRBs and to issues related to social and behavioral science;
- Monitoring the accreditation process.
Dr. Genel called attention to an article on the increased regulatory burden, “Grinding to a Halt.” It may be found at: http://www.natap.org/2009/HIV/071009_04.htm
Dr. Botkin felt that while mission creep is a reality and there are significant problems in the IRB review system, research participants have been well served by system as it has evolved. He expressed concern about the lengthy time required to change regulations governing research oversight, which have become “ossified,” and wondered whether it was possible to envision ways that needed changes might be streamlined. He highlighted the importance of building a collaborative relationship between participants and investigators that goes beyond “informed consent.” SACHRP might wish to explore the emerging research in this area.
Dr. Gibbons also highlighted the importance of collaborative relationships with participant communities, pointing to the connection between research questions and improvements in the health of these communities. He noted the challenge and opportunities posed by the “new age of information sharing.”
Dr. Marshall was also interested in researchers, individuals, and communities as collaborative partners. She noted increased attention in the last 10 years to social and ethical issues surrounding such collaborative relationships, including the complex issues that arise in the international context. She wondered what could be done to build capacity and share information with international communities participating in U.S.-sponsored research.
Dr. Sodeke also wondered what SACHRP could do to foster community-based participation in research, which can begin with identifying community needs and continue with opportunities for involvement in every phase of the study. The rights and welfare of communities are frequently not considered as part of the oversight process. He also noted the varied abilities of IRBs to do the ethical analysis required. He encouraged SACHRP to explore whether training for IRBs would be helpful and inviting organizations such as the Community-Campus Partnerships for Health to discuss the scope of problems and needs in the area. See http://www.ccph.info/
Mr. Forster called attention to three pressing issues:
- Consistency among Federal regulations, including those of FDA and OHRP. He suggested early attention to “low-hanging fruit” such as differences in the anniversary date for continuing review.
- The need to unify definitions of vulnerable subjects.
- Attention to the consent process (possibly through SAS, but perhaps with new and specialized expertise).
Dr. Leiden emphasized the importance of harmonization as a focus point for any new subcommittee. She also recommended panel presentations on international research. Finally, she stressed the importance of assisting OHRP as requested to move ahead with SIIIDR recommendations.
Dr. Strauss focused on the importance of education, especially to communicate ethical ideals to investigators. He noted that they can memorize the requirements for informed consent and still have no real understanding of how to do it well. Best practices and good examples should be disseminated. As a country, he said, we have not done a proper job of educating and engaging the community around research ethics.
Ex officios also were asked to identify priorities. These included:
Dr. Trachtenburg (Indian Health Service) was particularly interested in community-research relationships.
Dr. Cates (Department of Veterans Affairs) was pleased with SAS’s recent progress on informed consent for the use of biospecimens. Her primary recommendation to SACHRP is for it to continue to make harmonization among Federal human research protection requirements one of its highest priorities. In particular, she said, there needs to be more consistency between Common Rule requirements and FDA requirements, as well as consistency among OHRP, FDA, and the Office of Civil Rights requirements as regards the HIPAA Privacy Rule.
Dr. Lux (Environmental Protection Agency) supported the issues identified by SACHRP members, noting that for his agency, group benefits and harms are a major concern and must often be considered at a national, or even global, level.
Mr. Rodamar (Department of Education) noted that for his Department, the ability to recruit and retain research participants in schools in a timely way is paramount.
Ms. Decot (Department of Defense) had no priorities to add, but wondered how much more feedback OHRP could process with the resources at hand.
Dr. Less (Food and Drug Administration) liked the idea of promoting consistency between OHRP and FDA. She noted that some things could be done quickly and would make a big difference. She added that many of FDA’s current efforts are focused on reducing administrative burdens.
Dr. McNeilly (Agency for Healthcare Research and Quality) supported others who highlighted community involvement. He hoped that rapid increases in technology would help disseminate information in related areas.
Dr. Budashevitz (National Institutes of Health) appreciated the efforts to prioritize previous recommendations and encouraged OHRP to continue to invite SACHRP’s participation as a sounding board as it seeks to implement recommendations it views favorably. He felt it would be helpful if OHRP shared details of its assessment of these recommendations and the challenge of implementation, with the idea that SACHRP might be able to provide additional input on policy options the Office might wish to pursue.
Dr. Geolot (Health Resources and Services Administration) would like to see SACHRP convene workshops on complex issues such as data repositories, alternative models, waiving informed consent, and group harms.
Dr. Kirchner (Department of Energy) supported all priorities mentioned. Additionally, he highlighted the difficulty many IRBs have in distinguishing between exempt and nonexempt research.
A spokesperson for the Center for Disease and Prevention was particularly interested in applications of the Internet in research and their potential in human subjects protection.
Dr. Menikoff assured SACHRP that “harmonization is on our agenda in a major way” and that OHRP is “interested in partnering with FDA.”
Subjects highlighted by speakers may be summarized as shown in the table below.
Possible Topics Highlighted by SACHRP Members and Ex Officios
|Topic||Number highlighting topic|
|Harmonization and simplification of regulations||5 members, 3 ex officios|
|Creating and exploring collaborative relationships between participants and investigators; partnering with community entities and applying new technologies to share information.||5 members, 2 ex officios|
|Facilitating review of “no more than minimal risk” studies; exploring ways to reduce regulatory burden||3 members|
|Measuring IRB effectiveness||1 member|
|Developing an operational definition of “vulnerability”||1 member|
|Exploring the consent process||1 member|
|International research||1 member|
|Importance of education, including dissemination of best practices||1 member|
|Group harms||1 ex officio|
|Recruiting and retaining large groups of research participants||1 ex officio|
|Complex issues such as data repositories||1 ex officio|
|Confidentiality||1 ex officio|
Holly Taylor, M.P.H., Ph.D., Assistant Professor, Johns Hopkins Bloomberg School of Public Health and Core Faculty, Johns Hopkins Berman Institute of Bioethics; Steve Joffe, M.D., M.P.H.
Assistant Professor, Pediatrics, Harvard Medical School and Pediatric hematologist/oncologist, Dana-Farber Cancer Institute (DFCI) and Children’s Hospital, Boston; Norman Fost, M.D., M.P.H., Professor, Pediatrics and Medical History and Bioethics, and Director, Program in Bioethics, University of Wisconsin
Dr. Taylor explored the question of how best to measure the achievement of ethical goals related to research oversight. Noting that many believe that IRBs today are paying increased attention to regulatory compliance at the expense of exploring ethical issues, she argued that the things that are easiest to count are not necessarily the most significant. While studies of ethical quality have focused on issues such as how the balance of harm and benefit is assessed, the quality of information provided to subjects, and the informed consent process, there is little assessment of whether IRB members really share the same understanding of these concepts.
The speaker noted that the IRB is one of many players with roles to play in improving the ethical quality of decisions, and that improving ethical quality means improving interaction among stakeholders. She proposed developing a working definition of ethical quality in conduct of human subject research, identifying the most targets for measurement, developing preliminary measures, and finally, after soliciting feedback from relevant stakeholders, testing their feasibility.
Dr. Joffee suggested that determining measures of quality and of IRB effectiveness begins with framing the objective of research oversight. Is it to seek optimal outcomes or to avoid bad ones? He argued that it is easier to agree on the latter. It is then necessary to show how structure and process influence outcomes. Examples of questions in three key areas include:
- Structure-Process: e.g., how do IRB resources or written policies & procedures impact quality or comprehensiveness of review?
- Structure-Outcome: e.g., how do members’ expertise & training impact likelihood that scientifically poor-quality protocols will be improved or disapproved?
- Process-Outcome (most important): e.g., how does attention to clarity of consent form impact likelihood that prospective participants will make informed decisions?
While structural aspects of oversight are easiest to quantify, consideration of process may lead to the most direct measures of IRB performance. First, however, it is essential to identify and validate the processes that are associated with the desired outcomes. The evidence base today offers relatively little to build on in this regard. Nevertheless, he argued that it is possible in the short term to begin to build a culture of quality improvement and evidence-based research oversight without waiting for final answers from the long-term effort to select outcomes, validate relationships of these outcomes to structure and process, and develop measures of key indicators.
Dr. Fost argued that IRBs today play a relatively small role in the effort to protect human subjects from harm through ethically responsible research, given the numerous other layers of protection that were less prevalent when the present system was developed. Instead, he said, they are often focused on protecting their institutions from harms that can result from shutdowns, lawsuits, and more commonly, poor evaluations by the Association for the Accreditation of Human Research Protection Programs and FDA. He agreed with former OHRP Director Gary Ellis who considered “catastrophic failures” as a result of inattention to major ethical concerns in protocol review unlikely. He likened much of the present approach, which emphasizes quantitative measurements of compliance, to requiring pilots to walk the aisles of their planes with clipboards to document that all tray tables are in a “full upright position.”
The speaker further argued that some highly publicized subject deaths may not have been avoidable and do not necessarily indicate systemic failures. He suggested that inconsistencies among IRBs were not all bad, noting that some inconsistencies might actually indicate more thoughtful strategies for balancing protection of subjects with ethically responsible research. More significant issues worthy of attention in his view included:
- Implementing proven methods of improving the consent process, such as interactive computer programs.
- Instead of the present focus on disclosure, increasing attention to documentation of meaningful informed consent, including consent monitoring, for high-risk studies.
- Analysis of “sentinel events” – problematic studies that lend themselves to failure analysis.
- Increased attention to highly publicized problems with industry-funded studies, such as biased design, concealment of adverse events, and the practice of ghostwriting articles that give a misleading picture of the science. (“Follow the money.”)
- Educating the public and Congress on the fallacy of “zero risk” and “zero tolerance for error.”
- Shift away from the emphasis on documenting compliance with requirements that have a low predictive value for human subjects protection.
- “Teasing out” situations that are really high risk and deserve more attention.
The extended discussion that followed included the following questions and key points:
Constructing paradigms. A SACHRP member asked how SACHRP would begin thinking about the issue of how best to measure IRB effectiveness. What paradigms would apply? What next steps would be appropriate? Responses varied:
Dr. Taylor reinforced the need to “open up the conversation” and encourage creativity as opposed to “maniacal box checking” driven by fear of noncompliance. Quality improvement should be seen as a long-term goal. Dr. Strauss noted, however, that regulatory frameworks in general are not “known for creativity.”
Dr. Fost proposed refocusing attention on issues where serious harm may occur.
Dr. Joffe urged learning as much as possible from instances in which something goes wrong (“sentinel events”). He also recommended identifying high-risk protocols and concentrating attention on feedback from these studies in particular.
Dr. Genel noted that SACHRP’s charge as an advisory committee is broad and can encompass a correspondingly broad view of human subjects protection. He agreed that an emphasis on the consent process is warranted.
Answering critics: Are IRBs really needed? Dr. Menikoff observed that there are many critics of the system who believe that IRBs are not needed. Can we measure what they are contributing? Dr. Fost responded that while there are still “crackerjack IRBs” that are doing excellent work, the trend toward increasing use of central IRBs is excellent. He noted, however, that some of the most problematic research originates in the institutions themselves; local high-risk intervention studies deserve attention.
Dr. Taylor added that SACHRP can help by focusing the Secretary’s attention on examples such as those highlighted by Dr. Fost, which involve protocols that warrant a higher degree of scrutiny.
Informed consent. Dr. Strauss held that new measures for quality improvement are not needed at present. There is a “vast literature” on consent that demonstrates that, as practiced, the consent process does not result in subjects understanding fundamental issues. The regulatory framework is restrictive, and IRBs do not have the flexibility to approve forms that do not contain all the required elements.
Dr. Joffe responded that measuring the outcomes of informed consent has not become a part of routine practice. However, quality improvement would provide a way to test whether changes are improving outcomes. Dr. Fost added that in his experience, investigators do a better job of ensuring informed consent when they know that comprehension will be tested. He recommended monitoring comprehension for high-risk protocols. Dr. Strauss said his IRB uses similar approaches. For example, it frequently conducts feedback surveys with subjects to determine how well the informed consent process prepared them for study participation
Accreditation. Dr. Fost described accreditation, as currently performed, as a “giant leap backwards.” He felt that many of the items being checked are not demonstrably linked to human subject protection and encourage a focus on regulatory compliance at the expense of real commitment to the goals of the Common Rule. Dr. Joffee added that the current principles and standards of the Association for the Accreditation of Human Research Protection Programs (AAHRPP) do not address quality improvement, and it would be helpful to add a standard that addresses this area.
Risk categorization. Dr. Taylor suggested retrospective studies that examine how protocols are categorized in terms of risk, then monitor adverse and unexpected events to generate a “feedback loop.”
Encouraging resource allocation. Dr. Botkin asked whether there were opportunities to encourage additional resources for developing and testing measures of IRB effectiveness. Dr. Taylor said the Clinical and Translational Science Awards (CTSA), mentioned earlier by Dr. Hischfield, allow institutions to carve out “research achievement programs” that look at quality-related issues systematically. The Johns Hopkins School of Medicine is currently pursuing a program that looks at issues such as good value, the culture surrounding research, and investment of research ethics. Dr. Joffee added that the issue fit well in the context of many NIH programs.
Detecting bias in industry-sponsored studies. Dr. Bierer asked whether it was possible to give IRBs a voice in addressing the issue of biased research designs that may cause harm to patients. Dr. Fost responded that in regard to studies sponsored by the pharmaceutical industry, where this arises most frequently, IRBs have no way of knowing all the biases that go into the design and will not see the documentation that would clue them in. Their role is further complicated if they have to report to an official who has, as a primary responsibility, bringing in money for the institution.
Barbara Bierer, M.D., Chair
Comments from the public were invited, but none were offered.
The Chair thanked speakers and suggested that SACHRP address recommendations on future planning and the work of today’s panels on Wednesday.
WEDNESDAY, JULY 22
Barbara Bierer, SACHRP Chair
The Chair invited discussion of possible recommendations based on yesterday’s presentations and discussions.
Dr. Strauss proposed that SACHRP express its support for OHRP’s continued consideration of possible regulations or guidance to increase use of external IRBs where appropriate by holding them accountable for their assigned roles. SACHRP discussed the wording of such a recommendation, and after passing an initial recommendation, revisited the subject to make corrections and finalize the recommendation as follows:
Recommendation Regarding ANPRM Follow-up. SACHRP endorses the concept that external IRBs and IRB organizations be held directly accountable for meeting certain regulatory requirements of the Department of HHHS regulations for the protection of human subjects, as presented by the ANPRM, whether that be through regulations or guidance.
Dr. Genel stressed that the recommendation does not mean that external IRBs and IRB organizations would be held responsible for roles and responsibilities they are not taking on as part of the protocol review process, such as a failure to carry out the consent process properly at the study site.
In response to a panelist’s remarks on the previous day, Dr. Pritchard clarified that the ANPRM is not an indication that OHRP is considering making IRB members liable for IRB actions.
The Chair noted that several panels were suggested, including presentations on:
- Community-based participatory research,
- International collaboration and research,
- Issues attendant on Internet research,
- Effect on communities and individuals as consequence of individual participation or community-based research, and
- Issues related to harmonization.
She asked members to consider the possibility of a subcommittee that would undertake a broad-based review focusing on harmonization of regulations and guidance from different offices, including OHRP, FDA, the Office of Civil Rights (OCR), and potentially the U.S. Department of Veterans Affairs (VA) and NIH. After a number of ex officios expressed interest in participating on such a subcommittee, SACHRP unanimously approved the following recommendation:
Recommendation Regarding New Subcommittee. SACHRP recommends establishment of a subcommittee to identify areas in which harmonization and simplification in the guidance and regulations of HHS agencies and signatories under the Common Rule would be of benefit to the research community and contribute to the protection of research subjects.
The Chair commented that the subcommittee’s charge is appropriately general and allows to subcommittee broad scope to approach questions that require engagement of other agencies and to focus its attention on the areas most pertinent to the protection of human subjects and the process by which that is achieved.
After discussion of the report by the Subpart A Subcommittee (SAS), which included a consideration of the complexity of addressing concerns related to Health Insurance Portability and Accountability Act (HIPAA) and the Common Rule in several scenarios, SACHRP also approved the following recommendation:
Recommendation Regarding OCR Representation. SACHRP respectfully suggests that a representative from the Office of Civil Rights (OCR) be invited to participate on SACHRP as an ex officio member.
Report of SubpartA Subcommittee (SAS)
Daniel Nelson, M.S., CIP and Elizabeth Bankert, M.A.
The Co-Chairs presented scenarios and suggested answers related to Informed Consent and Biospecimens in the form of Frequently Asked Questions (FAQ). So far SAS has developed 17 questions with prepared responses and another 11 with responses still in development. SACHRP also developed definitions for terms and concepts associated with these scenarios (see PowerPoints). The subcommittee has crafted each FAQ as a stand-alone question to increase clarity and focus.
FAQ 1.Tissue biopsies were obtained for clinical diagnostic purposes, which have now been satisfied. The hospital pathology department is willing to provide a portion of the remaining biopsy specimens to an investigator who will perform research assays. In order to allow matching with relevant clinical information, the specimens will be provided with identifiers such that the investigator can readily ascertain the identity of subjects.
Is consent of the patient from whom the biopsy was taken (or IRB waiver of consent) required for the secondary research use?
Response # 1.Yes. Under this scenario, informed consent of the subjects should either be obtained or waived under 45 CFR 46.116(d) because the samples are identifiable to the recipient investigator.
Discussion was limited to a question about why the research could not be exempt. Mr. Nelson explained that this was because of the presence of identifiers, so that specimens and information would remain readily identifiable to the investigator. The FAQ was approved.
FAQ # 2.Tissue biopsies were obtained for clinical diagnostic purposes, which have now been satisfied. The hospital pathology department is willing to provide a portion of the remaining biopsy specimens to an investigator who will perform research assays. The specimens will be coded such that the investigator will not be able to readily ascertain the identity of individuals (see definitions).
Is consent of the patient from whom the biopsy was taken (or waiver of consent) required for the secondary research use?
Response # 2.No. Under this scenario, neither consent nor waiver is required, because the activity is not considered to be research involving human subjects.
SACHRP approved this FAQ without discussion.
FAQ # 3.Blood samples were obtained for research purposes, with informed consent of the subjects, and the original study has been completed. The samples remain under the control of the original investigator. Another investigator wants to use a portion of the remaining samples to perform research unrelated to the original study.
If the original consent stated that “…your sample will only be used for research on diabetes,” but the secondary user is interested in studying mental illness, can the samples still be used if provided to the secondary user in a coded fashion (see definitions)?
Response # 3.The secondary use would not be considered human subjects research under 45 CFR 46 and, therefore, IRB review would not be required for the secondary use. However, institutions should establish mechanisms to determine whether secondary uses are compatible with the original informed consent. In the case at hand, it could be argued that coding before the secondary user conducts research does not negate the fact that the use appears to violate the original terms of consent. That is, there is an ethical obligation to honor the original terms, even if this project is no longer considered to be human subjects research.
Unresolved issue for discussion: Regardless of ethical acceptability, is this a regulatory violation?
There was extensive discussion of this scenario. Key points included the following:
- Consider each use of “should” to see if that is really what is meant.
- The implications of a “silent” consent form should be thoroughly considered.
- Depression and diabetes might be linked, so the example might be confusing.
- People are sometimes concerned about future use and do want to know what they are consenting to.
- The answer should be more positive in terms of who will make the decision about secondary use.
- It should be clear whether the IRB is the proper review mechanism or the institution should establish one.
- Some banks establish mechanisms for making such decisions. Is that ethically acceptable?
Speaking for OHRP, Dr. Pritchard said the Office would be willing to take a position on whether or not something is ethical, regardless of whether or not it violates a regulatory requirement. In regard to secondary uses of de-identified or coded samples that are not human subjects research but may not be compatible with the original terms of consent, Dr. Pritchard delineated three ways of thinking about such issues. He posed the question, “In what ways can you say that the requirements for human subjects research endure when the original research project is over?”
If the researcher has told the subjects, “I will keep your information confidential forever,” then publishes the names and data of all subjects the day after the research is officially concluded, this would still be, arguably, a violation of regulatory requirements. If a provision related to §116 imposed a requirement that is still in play after the research is over, that requirement is still in effect.
The second option would tie the requirement to anything that might be considered a procedure of the research with reference to §116 (a) (1). If the description of the procedures provides that the specimen will be destroyed, then even if the research is completed, the procedures described would still be expected.
The third way of thinking about the issue would tie the regulatory requirement to informed consent. The subject relies on whatever information is presented and promises are made, whether or not they are actual research procedures. The subject is consenting on that basis, and the promises made should be honored.
Dr. Pritchard said that general counsel has informed OHRP that the first option is at least a defensible interpretation of the regulations. The “jury is still out” on the second and third options. Ms. Odwazny clarified that if OHRP chose Option 1, it would need to establish the new interpretation through formal guidance and give the public an opportunity to comment. In regard to Option 2, she noted that some courts have considered informed consent to be a unilateral contract. She felt that the standard for interpreting informed consent as authorization for future research should be what the subject might reasonably be expected to understand.
Dr. Botkin found the implications of this scenario to be “enormous.” He said there should be no issue when a change to a protocol actually reduces risk: “If the principle is strict fidelity to the original agreement,” he said, “we will significantly restrict the conduct of research.” Dr. Strauss differed, highlighting the relevance of privacy issues and the difficulty of anticipating what might be detectable through in rapidly emerging procedures, particularly in regard to genomes. Dr. Botkin agreed that there could be ethical issues in such future use, but he resisted the conclusion that subsequent use would be prohibited unless it was envisioned in the original agreement. Another SACHRP member emphasized the need to promote the construction of informed consent documents that do clearly describe future use and to specify in agreements with secondary researchers, when relevant, that they will not seek identifiers in the future and will follow the terms of the original consent.
Possible new wording was presented to address some concerns. However, the FAQ was not approved. Some members felt the answer had far-reaching repercussions that required further consideration.
FAQ # 4.It is increasingly common to collect and store specimens for future unspecified research.
How broad can this consent be without requiring investigators to obtain additional consent for specific uses? Alternatively, how specific must this consent be to allow for future use of biospecimens?
Response # 4.There is a tension between the desire to be as specific as possible when informing subjects of what will be done, and the reality that specifics are, by definition, not known at the time of consent.
Many institutions and IRBs have found it prudent to be “broad yet specific,” in ways that give subjects a reasonable idea of research that might be conducted in the future, but without placing unreasonable restrictions on what that might be. Thus, subjects can be informed that future studies may involve genetic research, drug development, or searching for links between genes and environmental factors like diet or lifestyle, or between genes and diseases. While examples might be given of specific diseases (e.g., cancer, diabetes, heart disease), being overly-specific or restrictive in this regard may result in problems later, when investigators propose other uses. Thus, IRBs and investigators should consider the downstream implications before promising subjects that “your specimens will only be used for research on XYZ.”
Future uses of identifiable specimens should be reviewed by the IRB, which should determine whether the research is compatible with original terms of consent, or whether additional consent may be required.
Alternatively, the creation of a repository with an oversight committee and “honest broker” mechanisms that distribute specimens to investigators in coded fashion can remove subsequent uses from IRB review, to the extent they no longer constitute human subjects research. In these cases, special attention should be given upfront to ensure that the repository (which is human subjects research and does require IRB approval) is established with policies and procedures to effectively manage subsequent uses in keeping with what the IRB approved.
Further complicating this situation are requirements that consent be specific in certain ways to satisfy the HIPAA Privacy Rule. [NEED OCR INPUT]
Mr. Forster pointed to the importance of the overlay of the Common Right with HIPAA. A representative of HHS’s Office of Civil Rights (OCR), which administers HIPAA, said that OCR’s current interpretation is that authorization must be study-specific. Ms. Bledsoe commented that for this scenario, it would be important to determine whether the data are de-identified according to the HIPAA Privacy Rule, whether they constitute a limited data set, or whether they are “identifiable” under the Rule. (“Disclosure” or authorization for each use is only required for data that are identifiable under the Rule. Also, a waiver of authorization could be obtained to permit disclosure of identifiable data.) SACHRP agreed that an OCR representative should be invited to join SAS. Members also agreed that the FAQ would not be useful unless HIPAA requirements were also addressed.
Mr. Nelson observed that it is not clear how specific informed consent need to be in regard to future uses. Some argue, for example, that a general statement to the effect that the sample could be used “for genetic research” is sufficiently specific.
FAQ 5. When can informed consent be waived for use of previously-collected human specimens and data (e.g., when does such research meet “minimal risk” criteria, what does “practicability” mean with regard to the informed consent waiver criteria)?
Response # 5.The criteria for waiver of consent under 45 CFR 46.116(d) include that the research involves no more than minimal risk; the waiver would not adversely affect the rights and welfare of subjects; the research could not practicably be carried out without the waiver; and whenever appropriate, the subjects will be provided with pertinent information after participation.
Points to consider in applying these criteria include the nature of the research; the protections in place to maintain privacy and confidentiality (e.g., coding, limited/controlled access, honest broker mechanisms); the change in level of risk, if any; the ability to locate or contact subjects; risk of introducing bias into the research; potential anxiety or confusion for subjects; the number of subjects; the length of time since specimens were first collected; and the likelihood that subjects would object to the proposed secondary use, based on the nature of original collection.
In regard to the “risk of introducing bias into the research,” Mr. Forster noted that “practicability” can be construed to mean “just not practical” or to an element of bias. He was concerned about the reference to bias as a consideration. He also wondered how much future use of samples would really be likely to cause “potential anxiety or confusion for subjects” and under what circumstances.
Dr. Strauss focused on the reference to the waiver not adversely affecting the rights and welfare of subjects (a regulatory criterion). He noted that people are often given the ability to retract or destroy samples. If the purpose of the study can be changed without informing the subjects, however, this power is of no real value.
Another concern expressed by a member was that if people with the worst outcomes are least likely to allow the use of their tissues, results would be biased. Ms. Bledsoe added that some anecdotal reports suggest that when people have had diseases and survived, they may not wish to be reminded of this period later, since it can cause anxiety.
A member of the public advised that the scenario explicitly state that the intent is to address situations in which a waiver would be appropriate.
In summary, SACHRP action on the scenarios discussed was as follows:
- FAQ 1 and the response were approved unanimously.
- FAQ 2 and the response were approved unanimously.
- FAQ 3 was not approved and was returned to SAS for further consideration.
- FAQ 4 and the response were considered not to be useful without overlaying considerations related to HIPAA, which provides the “ceiling” on what is allowable in this circumstance. SAS was asked to redraft the FAQ with these in mind.
- FAQ 5 and the response were approved by a majority of members.
Bernard Lo, M.D; Jack Harris, M.D. Vice President, U.S. Medical Division, Lilly USA, LLC; Guy Chisolm, Ph.D.,Vice Chair, Lerner Research Institute, Cleveland Clinic; Ross McKinney, Jr., M.D., Director, Trent Center for Bioethics, Humanities, and History of Medicine, Duke University School of Medicine
Dr. Lo reported on the findings of an Institute of Medicine (IO) committee on Conflict of Interest in Medical Research, Education, and Practice, which may be found at: www.nap.edu/catalog.php?record_id=12598
The committee’s dual charge was to
- Examine conflicts of interest in medical research, education, and practice and in the development of clinical practice guidelines, and
- Develop analyses and recommendations to inform policies to identify, limit, and manage conflicts of interest in these contexts without damaging constructive collaborations with industry.
Recommendations pertinent to SACHRP included the following:
- [Congress should] require pharmaceutical, medical device, and biotechnology companies to publicly report payments to physicians and researchers, health care institutions, professional societies, providers of continuing medical education, etc.
- [Disclosure should be standardized through] a broad-based consensus process to develop standard categories and formats for disclosure of financial interests to institutions.
- [Medical research institutions should, with limited exceptions] adopt a policy that investigators generally may not conduct research with human subjects if they have a significant financial interest in the outcome of the research.
- Nonclinical research and new models of collaboration between academia and industry should be identified and evaluated for potential COIs. Dr. Lo observed that COIs in basic research are “underdiscussed.”
- Institutions should create board-level policies and procedures to identify and respond to conflicts of interest at the institutional level, and NIH and the Public Health Service should develop rules for grantees on institutional conflicts of interest.
- [HHS should] develop a research agenda to assess positive and negative effects of conflict of interest policies.
He suggested that SACHRP consider the following areas:
- Institutions should cross-check investigators’ disclosure with publicly available information; there are many examples of researchers omitting large payments from disclosure statements.
- Consider providing guidance on investigators with COIs in human subjects research.
- Provide case studies on institutional COIs to stimulate discussion.
- Address the issues of ghostwriting and the need to publish negative results. Should these be considered misconduct?
Remarks by Jack Harris: Preserving the Value of Industry Interactions with Health Care Professionals
Dr. Harris focused on the efforts of the biopharmaceutical industry to ensure human subject protection. Key documents that reflect this commitment include:
The Pharmaceutical Research and Manufacturers of America (PhRMA) Principles on Conduct of Clinical Trials and Communication of Trial Results: http://www.phrma.org/principles-and-guidelines-clinical-trials
PhRMA Code on Interactions with Healthcare Professionals
The Advanced Medical Technology Association (“AdvaMed”) Code of Ethics on Interactions with Healthcare Professionals: http://www.zimmer.com/web/enUS/pdf/AdvaMed_Code_of_Ethics.pdf
He expressed support for the IOM committee’s efforts and many of its findings. He said that industry should be viewed as a key part of the solution in the effort to ensure that interactions are legal, ethical, and appropriate. He said it should participate in any effort to establish appropriate guidelines, which are in the best interest of industry as well as healthcare professionals and researchers.
In response to discussion questions, he said that in general, other companies share Lilly USA’s interest in transparency; that Lilly USA stopped the practice of ghostwriting a number of years ago, and that Lilly USA does not use investigators for whom the value of an investment is dependent on trial results. He added, with respect to one of the challenges of transparency of clinical trial results, that “no one wants to publish negative results.”
Remarks by Guy Chisolm: Conflict of Interest in Human Subjects Research
Dr. Chisolm explained how the Cleveland Clinic publishes information on the relationships its researchers and physicians have with industry. (See www.clevelandclinic.org , select “Find a Doctor,” and select “industry relationships.”) He noted that the federal government, several state governments, companies, and academic medical centers are planning disclosure Web sites for industry payments to doctors and scientists. He said discrepancies among these sites are inevitable due to different reporting formats, and site visitors may perceive discrepancies that are actually due to these variations. He therefore recommended:
- Encourage broad public disclosure, but determine a template for uniformity and construct the disclosure web sites to be understandable by research subjects, patients, and the public.
He also suggested that HHS consider addressing the need to educate the public on the benefits of academic-industry partnerships in research, the potential for conflicts of interest, and “best practices” for effective management of conflicts of interest.
Remarks by Ross McKinney: Conflict of Interest in Human Subjects Research
Dr. McKinney reviewed a number of pertinent guidelines that address issues related to conflicts of interest in research. Pointing to a 2007 study of selected research abstracts that found that authors with a declared consulting conflict of interest reported positive results more frequently – 97.8% of the time as opposed to 89% for authors with no such conflict of interest – he said conflicts of interest are indeed a “real problem.”
The speaker also stressed the dangers to subjects from off-label use of pharmaceutical products. Emphasizing the importance of consistancy in addressing potential conflicts of interest, both to protect subjects and to facilitate multisite studies, he called for a joint rule on the part of Common Rule signatories that addresses issues related to conflicts of interest in clinical research.
Structuring institutional authorities. Dr. Botkin raised the issue of how best to structure institutional authorities to address conflicts of interest. He noted that while IRBs are independent, an institutional committee on conflict of interest may be subject to pressure from influential individuals. Dr. McKinney said that Duke University School of Medicine has chosen to use the IRB whenever Duke intellectual property is involved. Dr. Chisolm said that the Cleveland Clinic places its conflict of interest program under the guidance of the Board of Trustees.
Per capita payment. Dr. Menikoff asked panelists to explore the per capita payment issue. Dr. Menikoff suggested that if the doctor receives $500 to treat a patient and $5000 to enroll that patient in a study, this would be important for the patient to know. Dr. McKinney noticed that much of such payments goes to the study and to staff other than the investigator. Dr. Bierer agreed, noting that it makes a difference whether the funds go to the research center or the doctor. Dr. Harris said that money that went to the doctor directly would be reported as fee for service under the rules followed by Lilly USA; however, if funds simply constituted a percentage of the investigator’s salary it would not be considered reportable.
Disclosures and Bias. Various speakers agreed that “nothing about disclosure mitigates bias.” In fact, there is some evidence that subjects think better of the person who discloses conflicts. Dr. Chisolm said that patients and their families should know more about the investigator’s ties and what they mean. Others agreed that education is important for all those involved.
Uniform guidelines or standards. Dr. Bierer noted that most speakers called for uniform guidelines or standards in conflict of interest disclosures, both in the interest of making multi-site reviews easier and facilitating public disclosures. She asked whether all speakers concurred that this would be a useful step. All agreed. Dr. Harris responded, “the more consistent, the better.” He said discrepancies in reporting could lead to misunderstanding.
Additional comments included:
Dr. Strauss stressed that any recommendation to the Secretary of HHS should emphasize the importance of addressing this issue for the entire subject community, not just those regulated by HHS.
Ms. Bankert asked Dr. Lo whether the IOM committee had considered the issue of consistency. He responded that the uniform rules should not injure the public. Dr. Chisolm agreed, noting that all interested parties should be at the table. Both stressed the importance of the interaction between industry and academia.
Mr. Forster observed that administrative burdens might be reduced by uniformity. For example, his institution asks investigators to fill out two forms related to COI, one for the COI committee and one for the IRB.
Discussion of recommendation. Dr. Bierer introduced a draft recommendation related to “Standards for the Reporting, Disclosure, and Review of Financial Conflicts of Interest” for discussion purposes. The following points were made in the ensuing discussion:
Members discussed whether or not it was a good idea to open the door to regulation. Mr. Forster commented that guidance already exists but has not promoted consistency, so it should be a possibility.
It is important to encourage the participation of all affected parties or stakeholders.
Dr. Lux commented that the problem appeared to be of crisis proportions, and SACHRP should stress the seriousness with which it regards the issue and the danger of harm to patients.
A consistent approach to the problem at the Federal level is of critical importance. Individual institutions are then more likely to “step up to the plate.”
After discussion, SACHRP unanimously passed the following recommendation:
Standards for the Reporting, Disclosure, and Review of Financial Conflicts of Interest.SACHRP recommends that the Secretary of HHS work with affected stakeholders to consider the need for regulations or guidance that define consistent standards for the reporting, disclosure, and review of financial conflicts of interest for those responsible for the design, conduct, or reporting of human subject research. These regulations or guidance should be uniform within the components of HHS (OHRP, NIH, FDA, etc.) and should include (1) uniform standards of reporting to any applicable entity (2) uniform standards of public disclosure (3) uniform expectations for appropriate information to permit human subject consent to be informed, (4) thresholds for evaluating and potentially limiting participation in human subject research based on financial COI s and (5) potential management approaches to limit undue influence and bias. Taken together, these approaches should promote industry-academic-public partnerships, allow transparency, ensure fair reporting, promote human subjects protections, and enhance public trust.
Public comments were invited, but none were offered.
Barbara Bierer, M.D., Chair
Noting that in its 8th letter to the Secretary of HHS, SACHRP asks OHRP to “require” that institutions provide training for a variety of specified audiences, the Chair asked if SACHRP members were prepared to reaffirm the recommendations made on March 29, 2007. The following recommendation was passed unanimously:
Reaffirmation of Previous SACHRP Recommendations on Education. SACHRP reaffirms its recommendation of March 29, 2007, recommending that OHRP require institutions to ensure that initial and continuing training is provided for IRB members, IRB staff, the Institutional Signatory Official and the Human Protection Administrator (e.g., Human Subjects Administrator or Human Subjects Contact Person), investigators, and other members of the research team with responsibility for conducting human subjects research.
Recorder’s note: The minute as passed actually referred to recommendations passed in June, 2007, but the correct date is March 29. As the intent of SACHRP was very clear, this change has been made in the above text.
Secretary’s Advisory Committee on Human Research Protections
July 21-22, 2009
Certification of the Summary of Minutes
I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
Barbara Bierer, M.D., Chair Date