Kennedy Institute of Ethics Journal

The exclusion of potential subjects on the basis of increased risks is a common and, in most instances, noncontroversial practice in human subjects research. Investigators routinely exclude potential subjects with reduced kidney function, bleeding disorders, and allergies to the drugs under investigation. Surprisingly, this use of exclusion criteria takes place in the absence of any explicit guidelines. In the present paper, I attempt to fill this gap in the literature and regulations by offering a specific account of the conditions under which potential subjects should be excluded from research on the basis of increased risk. This account serves two purposes.

First, subjects who are excluded on the basis of increased risk are prevented from making their own enrollment decisions. Therefore, it is important to ensure, in spite of the broad range of potential subjects and the varied reasons for excluding them, that potential subjects are excluded in [End Page 307] a systematic and fair way. Second, certain instances of subject exclusion, most notably the exclusion of pregnant women and women of childbearing potential, raise more difficult issues. By providing a general framework for assessing the exclusion of potential subjects based on increased risks, the present analysis covers these cases as well and, thus, offers a solution to one of the thornier ethical dilemmas facing investigators and IRBs today.

Briefly, I argue that standard exclusion practices suggest that whenever two groups of potential research subjects offer the same opportunity for scientific and individual benefit, but different levels of risk, the group that would introduce greater risks should be excluded. With this framework in place, I consider potential subjects, such as men and women of child-parenting potential, whose research participation introduces increased risks to third parties. I argue that the present approach is justified, not because it reduces harms to specific individuals, but because it reduces aggregate harms. Therefore, it should be applied to all riskier potential subjects regardless of who faces the risks in question. After defending this approach, I consider several objections.

First, I argue that the present approach is not paternalistic; the exclusions it calls for are not justified by what is thought to be in the best interests of the individuals excluded. Rather, the argument is that these exclusions minimize the aggregate harm of human subjects research. In addition, the present approach does not violate the obligation of researchers to respect subject autonomy because it does not reduce the number of potential subjects who are given the opportunity to enroll in a particular protocol. Finally, the exclusions are consistent with reasonable side constraints on attaining the goals of research.