Bias vs. bias

Bias vs. bias

Ruth Ison


This case is presented as an illustration of the challenges of handling potential biases among IRB members that are personal in nature and that might influence a member’s objectivity in evaluating a research proposal. In this case, manifested personal beliefs are posited as equivalent to a conflict of interest, which has material implications for an IRB’s operations and procedures for the approval of research. The question before the reviewer is to evaluate the IRB’s actions in the convened meeting and the post-hoc operational discussions and actions.

The discussion is framed in terms of voting actions within an IRB’s parliamentary procedure: abstention versus recusal. Because an IRB member disclosed, at a convened meeting, her religious beliefs that prevented her from approving a research study in Assisted Reproductive Technology involving Intra-cytoplasmic Sperm Injection, this member abstained from voting Yea or Nay on the question of approval.  The net result of her vote was disapproval of the protocol, since a majority of members present did not vote for approval (eight members present, four yeas, three nays, one abstention). Subsequently the IRB Chairman and administrators decided that this member should have recused herself from the vote on the grounds that her religious beliefs constituted a sufficient personal conflict of interest that she should have followed recusal procedures (for this IRB, presumably) and left the room during the discussion and vote.  This would have changed the net outcome of the vote, because with only seven members present, the four yea votes would have approved the study.

In this case’s detailed discussion of the impact of personal beliefs and hidden or assumed biases, particularly those associated with religion, it is salient that the discussion of biases is limited to those presumably associated with the roles represented by IRB members.  Medical centers served by IRBs, however, have an inherent interest in attracting and conducting research to advance the institution’s academic prestige, and also attracting professionals of the highest caliber, to keep research grant funds flowing. Thus an IRB, whose existence depends upon the organizational and financial stability of the institution, also has an interest in approving research, versus disapproving it. The Common Rule appears to recognize this presumptive interest and addresses it in terms of an IRB’s members, requiring diversity of sex, race, culture, profession, scientific orientation, and institutional affiliation. The IRB’s diversity is intended to be representational for the purpose of “promot[ing] respect for its advice and counsel in safeguarding the rights and welfare of human subjects.”1 In other words, respect for the authority of the IRB to make research decisions involving human subjects is grounded in the degree to which its members exemplify “sensitivity to such issues as community attitudes,”2 as well as its professional expertise.  Moreover, an IRB is also required by this section of the Common Rule to be competent to review “specific research activities…in terms of institutional commitments and regulations, applicable law, and standards of professional conduct and practice.  The IRB shall therefore include persons knowledgeable in these areas.”3

What does this mean for an IRB’s operations and practice?  It certainly does not mean that a person’s religious or cultural beliefs constitute an unacceptable conflict of interest sufficient to require a recusal from a discussion or vote on a research proposal, if the IRB’s Standard Operating Procedures provide for such an action. Indeed, this member’s open disclosure of something in her cultural background that necessitated taking a particular ethically-based stand conforms with the Common Rule’s intent to include divergent opinions of a non-scientific nature (as well as those based upon principles of science). IRBs are intended to be a forum in which discussions of ethical, cultural, and scientific principles can take place in the context of each specific research study—otherwise, how can an IRB fulfill its duties to protect research subjects, who will be taken from the population served by the medical center? Even if the abstaining member’s religious beliefs were a minority view within the medical center’s demographic, it would nonetheless still be within the IRB’s purview (and due diligence) to include her views at a discussion in a convened meeting. Clearly this member’s beliefs did not prevent her from deliberating and voting on other ART research proposals.  It appears, however, that the latent biases of the IRB Chairman and administration in favor of approval of this research warrant further examination. 

The post-hoc reformulation of the question before the IRB as an issue of recusal versus abstention, particularly in this case in which such a procedural change would have changed the outcome of the vote from disapproval to approval by manipulating the quorum of present members, suggests powerful institutional biases at work. These political biases undermine the IRB’s role of human subjects protection in far more subtle and dangerous ways.  As noted above, medical centers have a strong material interest in the type and scope of its research enterprise. IRBs within this context can be perceived by institutional executive leadership as a threat to the growth of a research enterprise, with the result that IRBs’ liberty and authority to make principled, scientifically sound decisions with regard to each piece of research can be severely constrained.  Such constraints are manifest in terms of actions to conform to leadership’s dictates to ensure members’ and staff’s job security, academic promotion or tenure, and/or enhancement of the reputation and prestige of the institution.  It is common, for example, for an IRB’s Chairman, affiliated members, or administrators to encounter internal political pressure to approve Dr. X’s protocols (or to require few modifications) if Dr. X holds a powerful position in the institution such as department chair, dean, or provost, or if Dr. X. is known for her ability to secure significant grant funding.  IRBs that make every attempt to evaluate research on its merits alone, even to the extent of blinding the name of the principal investigator among IRB reviewers, can easily be labeled as “obstructionist,” with the job security of its administration always at risk and the availability of senior scientific members’ time to serve on the IRB limited by competing professional and institutional priorities. 

Consider an assistant or associate professor coming up for academic promotion, who also serves on the medical center’s IRB.  A research proposal under consideration for approval has the professor’s department chairman as its principal investigator. Should the assistant professor/IRB member disclose a material conflict of interest in this case, which is clearly based upon the potential for personal gain or loss by the IRB member in question if his department chairman should discover that the IRB member up for a promotion voted to disapprove the chairman’s study?  Although any retaliatory action on the part of the department chair is illegal and unethical, what is to prevent it from occurring under another pretext? This situation presents a clearer case for recusal based on conflict of interest than the case of the abstaining member’s religious beliefs.  The abstaining member had nothing of personal value to gain or lose, whichever way the votes fell—all she could do was to take a stand based upon her principles to abstain from voting on the question because of her openly acknowledged and deliberated biases, which rightly took place in a convened meeting, and which should have been reflected in the minutes of the meeting. In the case of internal political pressure on IRB members, chairpersons, or administrators, the biases, potential for personal gains or losses, or constraints upon professional relationships within the organization are frequently hidden, but can exert a powerful restriction on discussion and decision-making, and thus a restriction upon protection of research subjects.

Thus it cannot be assumed that the IRB Chairman and other IRB members and administrators were without their own latent personal biases and beliefs, or even potential material conflicts of interest, in their consideration of the ICSI research study. The issue of religion is spurious: what ethical difference does it make if one’s biases are religious or political in nature? The religious member’s biases were at least conscious and willing to be presented in open deliberations. Clearly something significant was at stake, however, to warrant the post-hoc flurry of administrative activity after the study was disapproved, but this “something” was absent from the discussion in the case presentation.  Insufficient detail is given about the relative novelty of the research methodology, or other of factors influencing its prestige or desirability for institutions to participate.  Of note in this instance is the Common Rule’s provision found at 45 CFR 46.112, which permits institutional officials to provide additional review of research that has been approved by the IRB.  Institutional officials are precluded from approving research previously disapproved by the IRB.  Yet the focus on the matter of abstention versus recusal and whether one’s religious beliefs constitute a personal conflict of interest, or what are the limits of one’s representational roles, further obscures the interest of the institution (or at least of the IRB Chairman and administrators) to find a way for this research to have been approved, even with a significant majority voting Nay (their rationale is not provided). 

This IRB has considerable strengths within its administrative infrastructure, as evinced by its evaluation of its Standard Operating Procedures, detailed evaluation of parliamentary questions, and commitment to undertake continuing education on the topic of conflicts of interest.  It would do well also, however, to examine and reflect upon its governance policies regarding the IRB’s role as a check on unlimited institutional authority to approve research through its authority to revise and approve IRB procedures, the IRB’s rules of membership, roles and reporting structure of the staff, the IRB’s broad institutional reporting relationships, and accountability to its community.

1. Title 45 Public Welfare Part 46 Protection of Human Subjects: 45 CFR 46. 107

2. Ibid.

3. Ibid.

Case Studies homepage