Abstract
Keywords
Background
Institutional Review Boards (IRBs) were established by regulation to protect the rights and welfare of human participants in research. While exploitation and mistreatment of research participants in the name of science had been documented before the 1960s (Gordon and Shriver, 2022), such problems were felt to be rare or not relevant to the U.S. academic enterprise before the article in the New England Journal of Medicine by Beecher (1966) and the revelation of the Tuskegee Syphilis Study (Heller, 1972). These instances and others demonstrated that blindness to the rights and perspectives of research participants was not a matter of individual investigator misconduct but was tolerated by a system that simply did not provide for consideration of their interests. IRBs were meant to build such consideration into research review, to remind the scientific community that their work was in the public interest, and to make research with human participants consistent with the ethical principles articulated in the Belmont Report (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). Understandably, the regulations were responsive to the most egregious harms, and they focus on elimination of unnecessary risks, the requirement that the scientific question be important enough to justify the risks of harm that inevitably remain, and avoiding exploitation by making research participants partners in science through solicitation of their informed consent to participate.
Despite this focus on clearly recognizable (i.e. physical) harms, contemporaneous examples of psychological harm from research such as Milgram’s (1963) study of Obedience to Authority and Zimbardo’s Stanford Prison Experiment (Haney et al., 1973) led to the application of the same regulatory framework to social and behavioral research.
While there are no measures of aggregate research harms and thus no data, the IRB system has been effective at making another Tuskegee or Stanford Prison Experiment unlikely. From the beginning, however, some have seen IRBs as presenting an unjustified burden on research, particularly in the social and behavioral domain, and as highly public examples of exploitation or harm have become uncommon, the balance of concern between participant protections and research burdens has shifted (Kotsis and Chung, 2014). The preamble to the most recent update to 45 CFR 46, the Common Rule, made this rebalancing explicit (OHRP, 2017). At the same time, the research enterprise has changed structurally, with most biomedical research now funded by industry sponsors (Gilliland et al., 2022; Plottel et al., 2020), and technical developments have led to benefits and potential harms that could not have been anticipated when IRBs were established.
Long standing concerns about burdens and the emergence of these new harms led us to ask whether the established structures for the ethics review and oversight of research with human participants could be improved in ways that lead both to better science and greater public trust.
Methods
On May 5th, 2022, 16 individuals from across the United States with diverse roles in the research ethics and compliance community gathered at the Harraseeket Inn, in Freeport, Maine, for a full day brainstorming session to identify problems facing the current system of ethics oversight of research and to discuss possible solutions. Participants included IRB chairs, Human Research Protection Program (HRPP) directors, academic leaders from institutions that conduct biomedical and/or social behavioral research, representatives from professional societies and accrediting organizations, and leaders in the research ethics academic community.
The morning was dedicated to identifying and describing problems with the current system and what meeting participants would want to change. The afternoon was dedicated to discussion of possible solutions. All participants were asked to bring their personal perspectives; they were asked to participate as professionals experienced in this domain, not as representatives of the institutions for which they work. The intent of this charge was to avoid any reluctance to raise issues or propose solutions that would be hard to solve or difficult to implement in any particular institutional context. Discussions also explicitly followed the Chatham House Rule (Chatham House, 2022) and participants were promised that no perspective or opinion would be attributed to an individual. The gathering was supported by a grant from the Robert Wood Johnson Foundation.
The meeting began with a plenary session to review the agenda, set the context, and begin to discuss the current system for ethics oversight and review. Participants continued their discussions in small groups. After lunch, participants were asked to self-select to participate in one of two groups, one described as “starting fresh” and the other as “change within the system.”
All sessions were recorded and transcribed. Transcripts were reviewed and themes identified; themes were confirmed or refined by a second researcher. The following sections present the edited comments from each session and the themes discussed.
Results
Themes for each session are presented as theme graphs with additional notes for explanation of the major emergent themes. The graphs depict each theme mentioned in the sessions as a dark blue circle; the diameter of the circle is proportional to the number of times that theme was reflected in a comment. Lines connecting the circles indicate themes that were mentioned together in a single comment; the thickness and darkness of these lines indicate how often each pair of themes appeared together. Connections between themes allow the themes to be grouped, as discussed in sections below.
Morning Plenary
Morning Plenary Themes.
Themes
The most common theme in this session was the “Role of the IRB,” which was strongly associated with “Limitations of the Current Framework,” “Research Quality,” “COI,” “Scientific Review,” and “Ethics from the Beginning.” Another prominent cluster included “Participant Voice,” “Community Voice,” and “Empathy”
Several comments under these major themes also discussed informed consent and the primacy of the principle of autonomy in research ethics. Participants noted that research was conducted in the commercial space (i.e. not under regulatory or IRB oversight) which exposed participants to the same risks as regulated research without consent and questioned whether individual consent was still an adequate model. They noted that consent did not mean comprehension, and that requiring individual consent could also reinforce exclusions and biases.
Morning Small Group 1
Morning Small Group 1 Themes.
Themes
The small group discussion introduced some new themes and refocused some of the themes raised in the earlier plenary session. Participants raised the issue of tension between compliance and ethics, and reflected further on the incentives for institutions, researchers, and IRBs.
Morning Small Group 2
Morning Small Group 2 Themes.
Themes
There was no interaction between the morning small groups, although each group was informed by the shared discussion at the preceding plenary session. The second small group raised several new issues, some reflecting the struggle of IRBs that are trying to adapt to evolving research methods.
Change within the system
Participants chose to participate in one of two parallel sessions: “Change within the System” or “Start Fresh.” It was telling of the general perspective of the group that only about a quarter of the participants chose to focus on incremental change within the existing research oversight framework.
Change within the System Themes.
Themes
The group’s discussions fell in three major areas that were consistent with discussions in the morning. These included extending the influence of the IRB and the consideration of research ethics to the earliest stages of protocol development; adopting a “learning IRB” model that included transparency of decision-making so that IRBs could learn from one another; and finding ways to alter the incentive structures for both researchers and IRBs to address ethics as well as compliance. Not surprisingly, given the self-selection of the group, several participants came from organizations that felt that they had solved many of these problems.
Start Fresh
Start Fresh Themes.
Themes
As noted in the “Change within the System” session, almost three quarters of participants chose to join the “Start Fresh” group to brainstorm solutions. Instead of considering ways to tweak incentives, extend the work of the IRB to participant and researcher education, or allow ongoing learning, this group discussed better ways to structure ethics review. Their discussion was notable for its focus and consensus – there were two major set of themes, but many smaller themes were addressed in one or more comments around these central ideals.
Suggestions for what to include on such a list of issues included:
• Have you talked to the right stakeholders?
• Have you considered unintended consequences?
• Have you considered to whom the benefits would accrue?
• Have you considered who will face the risks?
• Have you thought about vulnerable groups?
• Have you considered what kind of research is important to stakeholders on a broader level?
• Was a “landscape analysis” performed to assess what work has been done before and why the proposed study has social value?
The research team should have, or have access to, capabilities including:
• Community engagement
• Ethics
• Regulatory sciences
• Independent review
An ethical perspective should be embedded in the research team and study design should reflect both the perspective of research professionals and affected stakeholders, including communities and potential participants. Ensuring those perspectives are considered could be the role of an “ethics facilitator.”
The group noted that this level of community engagement and teamwork might come at significant “cost” in time taken to educate communities and to come to agreement. One way to reduce this cost would be to have community involvement at a programmatic or institutional, not an individual IRB or protocol, level. Participants noted that institutions typically have strategic planning processes; building in community voice in this context could save time, as there would be less debate about whether or not to undertake individual research projects.
Others noted that this kind of community engagement would be difficult to sustain for small institutions that might not have the resources, leading to the idea that community voice might best be captured at the
The complexity of the proposed research apparatus would be inconsistent with traditional ideas of investigator-driven research, and might result in a smaller number of studies, but studies of higher quality and greater social value. To get the research community to adopt this approach would require changing the current incentive structure, which rewards research quantity as highly as research quality or impact. Such a change would have to be embraced by funders as well as institutions, which would have to take ethical design into account in decisions about career advancement and tenure.
Participants observed that, if ethics is embedded in the study from the beginning, then the role of the IRB is redundant. Instead of revisiting questions that should have been explicitly asked and answered by the research team, independent review would only make sure that the
The group noted again that small institutions might not have the resources to sustain this complex research infrastructure and raised the possibility that independent review could be conducted at a regional level.
The group ended with two minor but important observations. They noted that all decisions and rationales should be explicitly documented, not simply to document compliance and to facilitate review, but because many of the decisions would be values-based. Values change with communities and with time, and it would be important to understand rationales so that research designs could change with time and context. Lastly, participants observed that the process above only addressed initial review, but review should be ongoing throughout the research study. In particular, and in contrast to current study monitoring conventions, the experience of research participants should be actively used to monitor the ongoing impact of the study on the participants and their communities.
Discussion
Problem identification
The morning was dedicated to problem identification and was notable for its exploration of underlying structural issues. The group did not discuss the usual complaints about IRB systems, such as its effect in delaying or burdening research or the perception that its demands were arbitrary but tried to identify the circumstances that led to these perceptions. Four large issues emerged, all of which were interrelated, including:
• an obsolete regulatory framework,
• an inappropriate incentive structure in research,
• the role of the IRB in research design,
• constraints on the IRB.
Regulations
Participants used the evolution of research technology (e.g. research with Big Data) as an example of how the current regulatory framework was no longer adequate. Obvious gaps in regulatory protection of research participants and their communities, and the limited authority of IRBs to address these issues, leads to the perception (and often the reality) that IRBs serve a compliance, rather than an ethics oversight, role. In turn, compliance separated from its ethical rationale is seen as a “necessary evil” and a burden on the research enterprise; that is, something that does not contribute to the quality or impact of the research.
Incentives
The second major problem discussed was inappropriate incentive structures in the research enterprise. In academia, recognition and career advancement are tied to volume of publication, not necessarily to the quality or impact of discoveries. Similarly, in sponsored research, the focus is on generating revenue and lowering costs. Both sets of incentives started out as surrogates for scientific quality and impact but have become ends-in-themselves. Neither academic nor commercial incentives recognize a value for ethics, which may be viewed as a source of burden or delay rather than an essential part of high-quality research. It is also impossible to insulate IRBs, sitting within academic institutions or acting as companies that must be responsive to the desires of their clients, from these incentive structures, which may lead IRBs to limit their review to issues of demonstrable compliance. This incentive-driven focus on compliance further undermines the ethical authority of the IRB.
Research design
The third issue raised was the limited role played by the IRB in research design. The IRB is often the last step after studies have been funded and designed. Consequently, the IRB’s requests for modifications force researchers to defend or redo work they have already done. This approach undermines the reality that ethics is a necessary part of high-quality research and may serve to further an adversarial relationship between the IRB and the research team. The group felt that the IRB and its concerns should be involved in research design from the beginning, a theme that would be expanded in the afternoon.
Constraints
Lastly, all these issues led to constraints on the resources and flexibility of the IRB. These constraints were clear in the inability of IRBs to respond to concerns about community voice or community harms consistently. All participants felt that the IRB system, and the research enterprise in general, was not well-aligned with current societal awareness and concerns around exploitation, equitable access to research opportunities, and fair distribution of the benefits of research. A focus on compliance, lack of incentives in this area, and timing of ethics considerations all contribute to such unresponsiveness, leading to frustration with the inability of the current ethics review system to address obvious ethical issues.
A new approach – embracing a heterogeneous research enterprise
In the afternoon, conference participants separated into two self-selected groups to consider how to address the problems identified in the morning. Not surprisingly, these groups reflected the experience of their members. The “change within the system” group was made up of individuals from smaller institutions with large portfolios of relatively low risk social behavioral research. Their experience reflected the greater flexibility available to Human Research Protection Programs (HRPPs) in this setting, perhaps related to lower institutional risk and less demanding incentive structures. Voices in the “start fresh” group, in contrast, reflected deep frustration trying to change practices in large, well-resourced institutions doing industry or federally funded biomedical research. Such research is resource-hungry and can present real institutional risk, leading to incentives that have become divorced from research quality or impact and enforced by entrenched research management and compliance bureaucracies.
This division in the nature of research is natural, and should be explicitly recognized and embraced in any evolution of the research oversight system. In some sense, the division is recognized in the current framework through mechanisms like regulatory exemptions and expedited review procedures. On the other hand, the current framework explicitly applies a single set of principles to
Recommendations from the “change within the system” group directly addressed the issues raised earlier in the day, proposing that IRBs or HRPPs be closely involved in training investigators, and be available as a resource for study design. They also recommended that IRB decisions be documented and available, moving IRBs from insular committees operated in ways that protect their institutions, to pieces of a system for learning and evolving research ethics. They made the novel suggestion that incentives could be changed by explicitly recognizing a focus on ethics in reduced fees for oversight and review services for researchers willing to explicitly embrace ethical principles as part of their research. All these suggestions are aligned with the primacy of the investigator role, and recognize the value of investigator-initiated, curiosity-driven research, a core justification for traditional ideas of academic freedom. The weakness of this approach is that it can lead to uninformative research, although this risk is mitigated by the early consideration of ethical issues. Uninformative research is not simply a waste of resources (i.e. an “opportunity cost”), but also wastes the contributions of research participants and communities, who implicitly or explicitly expect their contributions and, where applicable, their acceptance of risk, to be meaningful. The line between research that appropriately falls under this kind of oversight, and research that needs more robust and structured oversight, should be drawn at a point that balances these concerns against the value of investigator freedom.
Where that balance cannot be satisfied, participants and communities need further protections, and the recommendations of the “start fresh” group apply. This group described a new research oversight apparatus, driven by a process that forces explicit consideration of ethical concerns from the earliest stages of research design and enforced by independent process review. In many ways, this approach is a “reset” of the current approach, requiring institutional resources akin to those required for NIH’s Clinical and Translational Science Awards (CTSAs) and an oversight apparatus analogous to the current IRB system. The revised infrastructure would include an explicit focus on ethics and community involvement, and would be enforced using a robust checklist focused on process rather than specific design. Independent review would similarly focus on process, recognizing that many of the decisions to be made do not have a single right answer, but that all answers must be convincingly justified. This approach revisits traditional academic values, replacing individual investigator-driven research with team-based research focused on societal value and stewardship of common resources. It would be resource-hungry and would risk creating a new research management and compliance bureaucracy that might evolve to serve its own ends; this risk would have to be mitigated by constant vigilance and conscious design. In addition, it would almost certainly lead to a lower volume of research, both because of resource constraints and complexity, and its benefits in terms of societal value would need to be regularly revisited.
The recommendations of both groups fit well together to describe a future approach to ethical research that explicitly recognizes many existing but largely unspoken tensions. Separating research into two systems would allow researchers to be trained in ethics and community concerns from the earliest part of their careers, while preserving their intellectual freedom and ability to act out of curiosity. Where the social costs or risks made such freedom inappropriate, it would be knowingly surrendered to a team approach focused on bringing the benefits of scientific knowledge to society as-a-whole, while robustly protecting members of that society.
Supplemental Material
sj-docx-1-rea-10.1177_17470161231157053 – Supplemental material for The Harraseeket Conference – Revisiting systems for ethics oversight of research with human participants
Supplemental material, sj-docx-1-rea-10.1177_17470161231157053 for The Harraseeket Conference – Revisiting systems for ethics oversight of research with human participants by Stephen J Rosenfeld, George Shaler and Ross Hickey in Research Ethics
Footnotes
Funding
Ethical approval
Research ethics review
ORCID iD
Supplemental material
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
