No Dice: The Resistance to Randomness

I

When James Childress included his 1970 Soundings article, “Who Shall Live When Not All Can Live?” in Practical Reasoning in Bioethics—a collection of his essays published in 1997—he followed it with a postscript, somewhat longer than the original article, in which he brought the deliberations of the intervening years to bear on his earlier arguments (Childress 1997c). The article plus postscript were then succeeded in that volume by reprints of two later publications, one from 1983 on the applicability of triage reasoning and practice to neonatal intensive care (also with a postscript), the other a 1991 article that considers fairness in medical resource allocation and focuses that discussion on the distribution of transplantable organs (Childress 1997b, 1997a). The three essays, taken together, contain Childress’s developed understanding of “medical utility,” especially as distinct from “social utility,” as a criterion integral to fairness in allocation of scarce medical resources—an original and influential contribution to biomedical ethics and to discussions of distributive justice.

In the 1970 article—his “first serious foray into biomedical ethics” (Childress 1997c, 180)—Childress describes and recommends a two-part selection process for considering candidates for renal dialysis or, in later application, kidney transplant.¹ Medical “acceptability” governs the first part of the process (Childress 1997c, 172–73), determining which patients need and would benefit from dialysis. As far as possible, the criteria for medical acceptability are to be cleansed of psychosocial considerations—factors most likely to reflect, even indirectly, discriminatory judgments of social worth—unless they can be shown in the case at hand to significantly affect medical suitability (as, for example, in relation to a particular patient’s ability to adhere to the medical regimen that must accompany dialysis in order to secure the benefits and safety of the procedure). In the second stage, allocation of time on dialysis machines among those deemed medically acceptable would be determined by a form of random selection, either in a lottery or, perhaps more appropriately in this case, on a “first come, first served” basis. The shift from utilitarian criteria in the first phase to random selection in the second honors personal dignity—the equal value of each person, important for all Childress’s arguments about allocation—and acknowledges inescapable limitations on human calculations of future outcomes and aggregate good. Childress’s overt aim in this proposal is to diminish the opportunity for unwarranted discrimination, an issue of unquestionable social and political importance in the United States at the time of this article’s composition.²

The postscript written for the 1997 republication of the article finds Childress adopting “medical utility, as distinguished from social utility” (Childress 1997c, 184), as a consideration both valuable and necessary in “final selection,” thus apparently allowing certain utilitarian considerations to modify the otherwise simple randomness of the second phase of the selection process, in the interest of good stewardship of scarce donated organs (too few dialysis machines is no longer the problem). By this argument, it would be ethically appropriate, for example, for a physician to decide to deflect a randomly allotted organ from a prospective recipient who is now nearing death for reasons the new organ will not resolve, in order that the donated organ go instead to someone who can benefit from it to a foreseeably and significantly greater degree. The definitions of medical and social utility are refined and clarified in the 1983 article “Triage in Neonatal Intensive Care” and in 1991’s “Fairness in the Allocation and Delivery of Health Care,” the terms are brought to bear on both stages of selection for organ transplantation. Or, perhaps better, on all three stages, for with the addition of attention to what Childress termed “final selection,” it is clear that the allocation process is now seen as comprising three steps: (1) selection by the local transplant team of individual patients as reasonable candidates for organ transplantation; (2) random allocation of donated organs among the candidates at a regional or national level; and (3) final determination, again by the local transplant team, of whether the allocated organ should be given to the designated...