No abstract

Most recent thinking about the vulnerability of research subjects uses a "subpopulation" focus. So conceived, the problem is to work out special standards for prisoners, pregnant women, the mentally ill, children, and similar groups. In contrast, an "analytical" approach would identify characteristics that are criteria for vulnerability. Using these criteria, one could support a judgment that certain individuals are vulnerable and identify needed accommodations if they are to serve as research subjects. Seven such characteristics can be evident in children: they commonly lack the capacity to make mature decisions; they are subject to the authority of others; they (and their parents) may be deferential in ways that can mask underlying dissent; their rights and interests may be socially undervalued; they may have acute medical conditions requiring immediate decisions not consistent with informed consent; they may have serious medical conditions that cannot be effectively treated; and they (and their parents) may lack important socially distributed goods. Each of these vulnerabilities can call for special care in the design and implementation of research protocols.

It is broadly held that confidentiality may be breached when doing so can avert grave harm to a third party. This essay challenges the conventional wisdom. Neither legal duties, personal morality nor personal values are sufficient to ground professional obligations. A methodology is developed drawing on core professional values, the nature of professions, and the justification for distinct professional obligations. Though doctors have a professional obligation to prevent public peril, they do not honor it by breaching confidentiality. It is shown how the protective purpose to be furthered by reporting is defeated by the practice of reporting. Hence there is no conflict between confidentiality and the professional responsibility to protect endangered third parties.

There is a broadly held view that neonatologists are ethically obligated to act to override parental nontreatment decisions for imperiled premature newborns when there is a reasonable chance of a good outcome. It is argued here that three types of uncertainty undercut any such general obligation: (1) the vagueness of the boundary at which an infant's deficits become so intolerable that death could be reasonably preferred; (2) the uncertainty about whether aggressive treatment will result in the survival of a reasonably healthy child or, alternatively, the survival of a child with intolerable deficits; and (3) the inability to determine an acceptable ratio between the likelihoods of those two outcomes. It is argued that the broadly held view accords insufficient weight to the fact that newborn intensive care increases the likelihood of harm to the child by effecting survival with intolerable deficits. Though treatment may offer a reasonable chance of a good outcome, it is argued that there are situations in which neonatologists should nonetheless defer to parental nontreatment decisions.

At the time of this writing, a widely publicized, waived-consent trial is underway. Sponsored by Northfield Laboratories, Inc. (Evanston, IL) the trial is intended to evaluate the emergency use of PolyHeme®, an oxygen-carrying resuscitative fluid that might prevent deaths from uncontrolled bleeding. The protocol allows patients in hemorrhagic shock to be randomized between PolyHeme® and saline in the field and, still without consent, randomized between PolyHeme® and blood after arrival at an emergency department. The Federal regulations that govern the waiver of consent restrict its applicability to circumstances where proven, satisfactory treatments are unavailable. Blood-the standard treatment for hemorrhagic shock-is not available in ambulances but is available in hospitals. The authors argue that the in-hospital stage of the study fails to meet ethical and regulatory standards.