Yale University.
Calendar. A-Z Index.
Yale Bioethics

IRB Cases: Comment

Recusal without “conflict of interest”

Stephen R. Latham, JD, PhD

In this case, an IRB member abstained from a vote on a protocol involving research into assisted reproduction. Her abstention was not based on any particular objection to the protocol under review, but rather on her general religious conviction that no such research should ever be conducted. Because the Common Rule requires approval by a majority of members present at the meeting, rather than of members voting, her abstention had the same effect as a “no” vote. Four of eight IRB members approved the protocol, with three disapproving and one abstaining; the protocol failed for lack of the required majority.   

The IRB administrators seek now to resolve similar future problems by deciding to classify their members’ convictions about the impermissibility of whole areas of research as “personal conflicts of interest.” Common Rule conflict-of-interest provisions demand recusal of conflicted members from consideration of protocols. Recused members commonly absent themselves from discussion and do not vote at all. Thus, in future cases like this one, the IRB member would be determined to have a personal conflict of interest, would recuse herself or be recused by the chair, and would cast no vote at all rather than an (effectively negative) abstention.

The recusal solution seems appropriate, but getting there via the avenue of “conflict of interest” is both incorrect and unnecessary. It’s incorrect because the abstaining IRB member in this case doesn’t have anything we can coherently call a “conflict of interest.” “Conflict of interest” regulations and guidelines—both in the Common Rule and elsewhere—are aimed at defusing bias rooted in self-interest. Here, the abstaining committee-member has no self-interest at stake; she stands to gain nothing by disapproving the research. She believes that, as a matter of religious principle, certain sorts of research ought not to be permitted. But a belief is not an interest. She is motivated by her religious convictions not to promote this research. But not every source of motivation is an interest.

Consider that three other members of the IRB did not merely abstain, but actually voted against the protocol. We may assume that their votes, like hers, were rooted in their ethical beliefs and motivations. Were their ethical beliefs and motivations, like hers, sources of “personal conflicts of interest”? If so, then every IRB member who ever disapproves of a protocol has a “personal conflict of interest,” and we’re in trouble. If not, then we’re owed an explanation as to why a “personal conflict of interest” flows from some ethical beliefs and motivations, but not from others. To distinguish her beliefs and motivations as religious is unsatisfactory, both because it discriminates impermissibly against religious bases for ethical views, and because we have no information as to whether the votes of the other IRB members were or were not based in their religious views.

We get closer to the nub of the matter when we concentrate on the fact that her objection to this protocol is based on a general disapproval of the entire line of research, rather than on any troubling characteristics of this particular protocol. As the IRB chair helpfully puts it, “her position on this protocol had been determined prior to hearing any discussion of the study.” The language of “conflict of interest” is unhelpful here, where the difficulty is rather a conflict of ethical convictions. She happens to hold ethical views that are incompatible with her fulfilling her duties as an IRB member with regard to this protocol. For this reason, rather than because of any “conflict of interest,” she ought to recuse herself.


Conflicts of personal ethics with role-requirements occur commonly enough, and recusal from role-duties is a common solution. Citizens who oppose the death penalty are prevented from serving on juries whose job it is to determine the propriety of applying the death penalty to a particular case. Judges who morally disapprove of abortion have been known to recuse themselves from hearing cases which might require them to strike down, as unconstitutional under current case-law, various legislative efforts to limit access to abortion.


The key question here is whether it is a legitimate function of an IRB to determine whether a given type of research (as opposed to a given instance of it) is socially desirable or morally permissible. If it is, then our abstainer’s moral convictions seem as well-grounded and pertinent as any other member’s, and she should vote “no” rather than abstain. If, on the other hand, the IRB’s duty is to engage only in review of particular characteristics of particular protocols, leaving broader questions of the legitimacy or propriety of lines of research to other decision-makers, then her blanket disapproval of the whole line of research prevents her from taking meaningful part in the IRB’s proper inquiry, and she should absent herself.

There is considerable evidence for the second view of the IRB’s function. Common Rule 46.111 sets out specific criteria for IRB approval of research. It requires IRB enquiry into minimization of risks to subjects, balance of risks and benefits, subject selection, informed consent, data monitoring, privacy and confidentiality. It mentions additional protections for various classes of vulnerable subjects. Its only gesture toward more “macro” ethical considerations is its prohibition on any IRB’s considering “possible long-range effects of applying knowledge gained in the research (for example, the possible effects of the research on public policy).” The Common Rule is also explicit on the point that other institutions (e.g., the university) are free to engage in additional research oversight.

Legislatures, regulatory agencies, universities, courts, and even individual researchers are free to decide on the morality of lines or avenues of research. But IRB members are not; their job is not to frame the research agenda but to protect research subjects. We can imagine a person who morally disapproves of all research on human subjects, perhaps because he believes that all research involves impermissible instrumentalization of humanity. It would be folly to place such a person on an IRB, where he would vote against every protocol. His ethical position unsuits him for consideration of moral problems internal to the project of human subjects research. In the present case, we have no such extreme: our abstainer is only ethically unsuited for participation in consideration of research protocols in the assisted-reproduction area. For this reason, she should decline to participate in discussion of protocols in that area, and take no part at all in voting. The silence of the Common Rule (and of Roberts Rules) on recusal surely cannot be taken to prevent a person from leaving a meeting, or from being asked to leave temporarily in cases where she is unable to cast her vote on the basis of her views on the subject-protection issues properly before the IRB.     

Return to case description

Return to IRB cases main page