Robert E. Goodin Australian National University My concern here is with the vulnerabilities of people who are "objects of medical interventions." I mean for that term to extend to "patients" in the case of therapeutic interventions as well as to "subjects" in the case of experimental interventions. The latter are the official focus of my discussion, but reflecting on cognate cases of vulnerable patients in therapeutic settings sheds useful light on the case of vulnerable research subjects It is the vulnerability of agents to other agents that most concerns me here (although I shall also, of course, be tangentially concerned with the various conditions that make them vulnerable to one another1). In general, one agent is "vulnerable" to another insofar as the interests of the former are sensitive to the actions and choices of the latter.2 In a therapeutic setting, the patient’s vulnerability is largely (if not exclusively) to the attending physician. In an experimental setting, the subject’s vulnerability is (again, largely if not exclusively) to the researcher. I. Vulnerability’s Two Faces Considerations of vulnerability might enter into bioethical reflections in two quite different ways. First is a broadly "consequentialistic" way, as in Goodin’s 3 Protecting the Vulnerable. By definition, if Sam is particularly vulnerable to Dr. Sue, then Sam’s interests are highly sensitive to Dr. Sue’s actions and choices. It therefore automatically follows that for any ethical theor y which attaches moral importance to promoting people’s interests, there is a moral reason for Dr. Sue to be particularly attentive to how her actions and choices will impact Sam’s interests. (Of course that reason is defeasible: there might be other stronger reasons for Dr. Sue to do something else.) That moral reason is stronger the more strongly—and, we might add, the more uniquely—Dr. Sue’s actions and choices are capable of affecting Sam’s interests. On this broadly "consequentialistic" account, Sam’s vulnerability (the potential impact of Dr. Sue’s actions and choices on Sam’s interests) is a reason for Dr. Sue to do something. Perhaps we might characterize that as a "responsibility" (or even perhaps a "duty"), derived from those consequentialistic considerations concerning Sam’s vulnerability to her. But for now, we can just leave it, more non-committally, as a "moral reason." A second way in which "vulnerability" might figure is as an exception or "disqualifying condition" within a broadly "deontological" conception of bioethics. On this model, we must above all respect the human dignity and moral autonomy of those with whom we deal. For medical practitioners and researchers, this is done by (among other things) securing "informed consent" from patients or subjects for procedures performed on them.4 Of course, there are many ways of telling a deontological tale. But because my concern here is with how vulnerability enters the picture, I shall focus on what I will call a "hard-line" version of the deontological ethic that makes informed consent a necessary if not sufficient condition of permissible medical interventions.5 "Vulnerability" enters into that sort of a consent-based deontological model as a disqualifying condition. "Vulnerabilities" of various sorts render agents incapable of giving meaningful consent to having those medical or experimental procedures performed on them. In the limiting case, vulnerable agents might not be able to consent at all— their vulnerability might be such as to deprive them of agency altogether. Here, however, I am going to focus on less extreme cases. These are cases in which vulnerable agents are still capable of giving what seems to be consent, but in which we are unsure of the moral value of that consent. The worry is that their vulnerability renders their putative consent "tainted," rather like a coerced confession in a criminal court or an agreement obtained by force or fraud in the law of torts. Where agents are sufficiently vulnerable, their putative consent might count for naught.6
II. If Not Consent, What? What follows when someone is not capable of meaningfully consenting to, or meaningfully withholding consent from, some intervention? Well, within a hard-line deontological ethic in which consent is the only right-maker, the upshot would be clear. If we do not have the permission of the person (or his or her authorized agent) to intervene in ways impinging his or her moral prerogatives, then we have no right to so intervene. This is as true of someone whose capacity to consent is compromised as it is of someone who is capable of consenting but who willfully withholds consent: in neither case do we have any right to so intervene, if consent is the only right-maker. Why a moral agent does not consent does not matter; the sheer absence of consent is all that matters in morally blocking our action. I hasten to add that this is a much harder-line deontological ethic than is embodied in any actual code of medical ethics. It seems so extreme as to preclude surrogate decisions for incapacitated patients, by anyone except perhaps their legally authorized representatives.7 In that and many other less extreme ways, this hard-line deontological ethic is very unlike any actual code of medical ethics operative anywhere in the world.8 Still, when excavating the true moral groundings for the ethical codes operative among us, it pays to begin by first exploring the limiting cases like this. This hard-line deontological ethic that treats consent as the only right-maker is especially interesting to explore in the context of the ethics of human experimentation, because at first brush, that looks like the ethic dominating thinking. The first "Directive for Human Experimentation" embodied in the Code laid down by the Nuremberg Tribunal specifies that: The voluntar y consent of the human subject is absolutely essential. This means that the person involved should have legal capacity to give consent; should be so situated as to be able to exercise free power of choice, without the intervention of any element of force, fraud, deceit, duress, over-reaching, or other ulterior form of constraint or coercion; and should have sufficient knowledge and comprehension of the elements of the subject matter involved as to enable him to make an understanding and enlightened decision.9 In short, at least at first brush: No consent, no experiment—on pain of something akin to practicing Nazi medicine. And insofar as the notion of vulnerability points to conditions invalidating consent, it thus provides what Ken Kipnis calls "a checklist of circumstances that...can invalidate the permissibility of research."10 But of course experimental settings are hardly the only ones in which doctors find themselves confronted with people who are vulnerable. Let us shift our focus to a therapeutic setting—a comatose patient who will die without urgent surgery, let’s say. There, "no consent, no intervention" would not be a sensible policy; certainly it is not the practice generally adopted in Emergency Rooms. There, after quickly checking the patient’s wallet and trying to contact the next of kin for instructions, the attending physician properly proceeds with whatever treatment "best medical practice" dictates in the circumstances— "as if " s/he had the patient’s informed consent, even though s/he does not. So too, upon reflection, should the experimenter. Indeed, the provisions of the US Code of Federal Regulations governing "Informed Consent of Human Subjects" now makes explicit provision for granting an "exception from informed consent requirements for emergency research."11 In certain tightly circumscribed conditions, experimental procedures that "hold out the prospect of direct benefit to the subjects" may be permitted on people from whom consent could not possibly be obtained ahead of time. One example was an experiment locating defibrilators in airports, to be used by laypersons on people having heart attacks before medics arrived, to see if that helped save lives: it was accepted by those authorizing that experiment that it was simply infeasible to obtain reliably informed consent from people in the midst of a heart attack; and since there was good reason to think that heart attack victims could benefit from those interventions, the experiment was authorized despite the fact that informed consent could not be obtained.12 Our concern here is with a medical researcher confronted with vulnerable subjects, and our question is, "What, then, follows from the fact that some subjects are "vulnerable" in ways that undermine their capacity to give or withhold meaningful consent for experiments to be practiced upon them?" The proper conclusion is not that we ought necessarily exclude them from our experiments—any more than in the "waived consent" cases we should exclude those who could benefit substantially from participation in the experiments, just because they cannot give informed consent. Rather, what follows is that we should apply "special scrutiny" to the conditions of their participation. Note, for example, the phrasing of the World Medical Association’s Helsinki Declaration: "When obtaining informed consent for the research project the physician should be particularly cautious if the subject is in a dependent relationship to him or her or may consent under dures."13 Just as Emergency Room physicians performing emergency procedures on comatose patients are supposed to think in terms of what those patients would "have good reason" to consent to, so too ought experimenters confronted with subjects whose seeming "consent" is compromised by their vulnerability pay special attention to whether those subjects would "have good reason" to consent, were it not for those vulnerabilities.14 This is sometimes expressed in terms of "hypothetical consent." That can be an evocative way of calling to mind the reasons people might have for consenting. But what it provides us with is a window onto the agent’s reasons, not a warrant rooted in the agent’s will. Hypothetical consent is no kind of consent at all, as generations of students have been rightly taught to scoff when confronted with "just so" stories of a "social contract."15 If the hard-line deontological model is correct, and consent is the only possible right-maker for interventions of this sort, then it is actual consent that is morally required to do the trick. If the prime moral directive is to respect the other’s moral agency and moral autonomy, then actual rather than merely hypothetical consent is the only moral warrant that there can be for intervening into the sphere of someone else’s proper moral prerogatives. For us to acquire a right to operate on (or experiment on) people whose actual consent is morally questionable, we have to move away from that hard-line deontological model toward a more consequentialistic one, which takes due account of vulnerabilities in its own peculiar way.
III. Consent in Consequential Context I take it that neither of the polar forms of ethics, deontological or consequentialistic, is altogether tenable in a health-care context. A hard-line deontological ethic would, as I have just been arguing, deprive us of the ability to treat emergency room patients clearly in need of medical attention but incapable of consenting to the procedures. Conversely, a hard-line consequentialistic ethic might risk turning us into Nazi physicians, practicing procedures on people who clearly (but, we have good medical reason to believe, wrongly) refuse to consent. In searching for a hybrid, my own inclination is to build on consequentialist foundations, grafting consent-based considerations onto that. Here, in brief, is the strategy I would propose. Recall the consequentialist story I told above, in which Sam’s vulnerability to the actions and choices of Dr. Sue gave moral reasons for Dr. Sue to be particularly solicitous of Sam’s interests in her actions and choices. All we need to get an element of consent into the picture is to recall that among those interests of which Dr. Sue is supposed to be solicitous are Sam’s "autonomy interests." Sam, like all moral agents, has an interest in being and being seen to be a self-governing agent capable of embracing and acting on reasons of his own. The reason Dr. Sue should, where possible, seek Sam’s consent to any procedure—even where she is sure it is one that best promotes all of Sam’s other interests—is that securing Sam’s consent is the only way also to protect his autonomy interests. This way of building autonomy interests in, alongside Sam’s other interests, has several advantages. It explains, in a way the hard-line deontological model cannot, why it might be all right to perform procedures on people incapable of consenting or whose consent is tainted by vulnerabilities of various sorts (their autonomy interests are not actually being overriden, insofar as they are not actually in play). It explains, perhaps better than any plausible variation on the hard-line deontological model can explain, why it is wrong for physicians to help kill anyone who genuinely wants to die, without further enquiries.16 (Presumably even those of us who would approve of physician-assisted suicide would ordinarily want the physician to ensure, for example, that the patient had a terminal illness rather than was merely "tired of living" or just was having a bad day.) There are many details left to be worked out with that hybrid model, to be sure.17 I do not want to belabour them here, however. Instead, I want merely to point out that something like that sort of hybrid is—contrary to what seems to be the common supposition—what actually lies at the heart of contemporary strictures surrounding the ethics of human experimentation. What the bioethics profession generally seems to remember as the rule of "the Nuremberg Code" is the first Directive, quoted earlier—the one that says "the voluntary consent of the human subject is absolutely essential." That is why we hear, so often, that "the Nuremberg Code’s foundational concern" is with "the concept of consent."18 That consent-based principle was undoubtedly of signal important to the Nuremberg Tribunal. After all, they listed that as their first Directive, and they elaborated on it at far greater length than any other item on their list. But what seems often to be forgotten is that the Nuremberg Tribunal did go on to list nine further Directives, only one of which has anything to do with the consent of the research subject.19 The other forgotten eight Directives of the Nuremberg Tribunal that are more consequentialistic in form are these20:
The experiment should be such as to yield fruitful results for the good of society, unprocurable by other methods or means of study... The experiment should be so designed and based on the results of animal experimentation and a knowledge of the natural history of the disease or other problem under study that the anticipated results will justify the performance of the experiment. The experiment should be so conducted as to avoid all unnecessary physical and mental suffering and injury. No experiment should be conducted where there is a priori reason to believe that death or disabling injury will occur... The degree of risk to be taken should never exceed that determined by the humanitarian importance of the problem to be solved by the experiment. Proper preparations should be made and adequate facilities provided to protect the experimental subject against even remote possibilities of injury, disability, or death. The experiment should be conducted only by scientifically qualified persons. The highest degree of skill and care should be required through all stages of the experiment of those who conduct or engage in the experiment.
... 8. During the course of the experiment the scientist in charge must be prepared to terminate the experiment at any stage, if he has probable cause to believe... that a continuation of the experiment is likely to result in injury, disability, or death to the experimental subject.21 Thus, while consent is of signal importance (as per Directive 1), consent is supposed to come only after a whole raft of consequential requirements have already been met.22 Essential though people’s consent may be to the ethical legitimacy of experimenting on them, those other background conditions are also equally essential. Consent is capable of transforming illegitimate experiments into legitimate ones, only when it is given against the background of all those other consequential conditions having also been satisfied. Upon reflection, this is surely unsurprising. After all, we would not let a surgeon perform an operation—even with the patient’s consent—unless there was some reason to think that the operation would do some good. (Presumably even with purely cosmetic surgery, we have to have good grounds for thinking that it will succeed, at least on its own terms, for it to be justifiable.) By the same token, we would not let an experimenter perform an experiment, even with the subject’s consent, if there were not evidence from animal experiments and so forth to suggest that the experiment might succeed. A subject’s fully-informed consent to experimental procedures that are unnecessary, incompetent, or gratuitously dangerous morally counts for naught. That—as much as the better-remembered principle that "the voluntar y consent of the human subject is absolutely essential"—is the law of Nuremberg. Of course, we would not want to proceed without first (if possible) obtaining the consent of the patient or subject, as well. We do not want to operate or experiment on unwilling subjects. But consequentialistic-style considerations constitute a prior hurdle. We do not want to perform operations or experiments on people without good reason to think that they will work, either. Why then, we might ask, is the issue about "consent" so much better remembered among medical ethicists? Well, recall for whom medical ethicists are principally writing: people at the medical coalface, which is to say, practitioners at the bedside. This is evident in, for example, the AMA’s "Principles of Medical Ethics": well over half of the nine guiding principles clearly pertain to relations between the doctor and his or her own specific patients. It is all the more evident in the Physician’s Oath prescribed by the World Medical Association’s "Declaration of Geneva," which has the physician avowing that "the health of my patient will be my first consideration."23 From the point of view of medical practitioners at the bedside—face-to-face with patient in therapeutic settings or subjects in experimental ones— the belief that "voluntary consent is absolutely essential" is indeed the thought that they should hold most firmly in mind. That is not because those other Nuremberg Directives (and their therapeutic equivalents) do not matter. It is merely because that is not where they matter. The bedside is not the place to meet those other requirements. They should have been (indeed, they have to have been) taken care of elsewhere, in planning the experiment or treatment regime. The bit of the Nuremberg Code that is uniquely the responsibility of a physician who is face-to-face with a patient or experimental subject is the bit about obtaining "voluntary consent." And given that it is practitioners at the bedside to whom codes of medical ethics are principally addressed, it is only right that that principle should be accorded heavy emphasis—heavier emphasis than it would be, if we were addressing instead (or even equally) planners of health care systems or designers of medical experiments. IV. Vulnerabilitries against Vulnerabilities When cashing out our morals in the currency of vulnerability, it is important to remember that those research subjects who we nervously deem "vulnerable" in ways that might compromise their consent to the experiment are often "vulnerable" in other ways, too. Often, they are vulnerable in the sense of suffering from medical conditions that make them suitable subjects for the experiment. Such people are vulnerable, too, in the sense that existing medical treatments of their condition are less than completely satisfactory (we would not be doing the experiment, certainly not on them anyway, otherwise); and the new trial treatment might be an improvement. When experimental treatment is expected to have positive therapeutic effects for the experimental subject as well, we are in the happy position of being able to justify the procedure on therapeutic grounds alone, and to treat any experimental payoffs as wholly unintended by-products. Often we are not in this happy state, though: realistically, we know that the benefits will be wholly or principally for other patients who come later, suffering from the same condition, rather than the patient upon whom the experiment is being performed. Still, even in this less happy scenario, we have a case of vulnerability-versus-vulnerability—the vulnerability of the experimental subject, versus the vulnerability of those who stand to benefit from the experiment. In any systematic application of our duty to "protect the vulnerable," those potential gains to vulnerable agents must be borne in mind, alongside all those other concerns about risks of vulnerable people being ill-used in research. For a particularly striking example, consider the continuing controversy over the exclusion of elderly people from clinical trials of drugs and medical procedures deployed commonly on the elderly. The issue is nicely set out in 1997 Editorial in the British Medical Journal: Practitioners face a difficult paradox in prescribing for the elderly. Those aged over 65 comprise only about 14% of the population in most industrialised countries, yet they consume nearly a third of all drugs. Ample evidence indicates that, even in healthy elderly people, aging impairs the way the body handles drugs. In ill elderly people these changes can be exaggerated considerably. In an ideal world data from premarketing and postmarketing surveillance studies would describe how a drug is likely to affect older patients differently from younger ones. Unfortunately, rather than being oversampled in clinical trials, to reflect their distribution in the drug consuming population, elderly people are inadequately represented.24 In some cases, the design of the trial literally excludes people over a certain age, for no scientifically justifiable reason.25 Other times the elderly are not formally excluded but are radically underrepresented in clinical trials. A recent RAND study analyzing patient and trial characteristics for 59,300 patients enrolled in 495 National Cancer Institute trials from 1997 through 2000 found, tellingly, that while 61% of cancer patients were elderly, only 32% of participants in clinical trails were elderly.26 Clearly, this is bad science. Insofar as the elderly are major users of those drugs and procedures, and insofar as the elderly do not react to them in the standard way, physicians clearly need those clinical trials to give them more information about the reactions of the elderly. And that is the way the issue is typically presented. The AMA’s Code of Ethics instruction on "subject selection for clinical trials" says, for example, that "Inclusion and exclusion criteria for a clinical study should be based on sound scientific principles."27 The reasons that the elderly are so often excluded or underrepresented in clinical trials of drugs, even those to be used predominantly by the elderly, are many and varied. The BMJ Editorial points to some: The "old old" are a messy lot physiologically. They are far likelier than the young to have coexisting medical problems, for which they are likely to be taking other potentially interacting drugs. They also have the distressing property of being more likely in the middle of a trial to suffer an infarct of the heart or brain or simply to drop dead. They are bad news for the drug development process. Yet another reason is that the elderly often count among those who would qualify as "vulnerable" in various respects, from whom we therefore cannot obtain meaningful consent. They are more likely to suffer cognitive impairments associated with the aging, they are more likely to be institutionalized, and so on. In addition to the pragmatic and scientific issues, then, there are also these ethical issues in obtaining meaningful informed consent associated with conducting clinical trials on the elderly. Ethical worries about their vulnerability, too, get in the way of including the elderly in experiments, in ways we scientifically should. Instead of construing this as an issue of "ethics" versus "good science," however, we can see it as an issue of "ethics versus ethics." Indeed, the ethical considerations on both sides can be seen to be broadly of a cloth, insofar as both involve an attempt to "protect the vulnerable." On the one side, we need to protect vulnerable research subjects from ill-usage in the experimental procedure. On the other side, we need to protect vulnerable patients from being prescribed drugs that have not been adequately tested on populations relevantly similar to their own. Saying that "vulnerabilities are involved on both sides" does not, of course, automatically tell us where exactly the balance should be struck. But the problem of weighing the competing considerations is nonetheless rendered far more tractable by getting them both on the same scale. Of course, there can be no thought of press-ganging subjects into experiments literally against their will. Kipnis is surely right to say that "the wrong committed by experimenting on an unwilling subject is of far greater seriousness than the wrong committed by unjustifiable exclusion."28 But that is not what is being contemplated, here. What is in view here is a consenting subject, albeit one who is vulnerable in ways that make us worry about the quality of that consent. In deciding on balance whether to go ahead and include the subject in the experiment, those worries notwithstanding, the benefit in prospect to other vulnerable agents ought I suggest be one further consideration weighing in favor of proceeding— particularly where, as perhaps with the elderly, the great majority of other potential experimental subject of the relevant sort would be consenting under similarly compromised conditions.
V. Conclusion The problem I have been wrestling with is why we should hesitate to experiment or operate on people who are vulnerable in ways compromising their capacity to consent— but why it might be all right to go ahead, despite those hesitations. My own preferred solution is to introduce "autonomy interests" into a consequentialistic model, which imposes a general duty on all doctors (and all others) to protect the interests of those who are especially vulnerable to their actions and choices. Other philosophers no doubt would prefer more deontological foundations. To them, I offer this closing observation. If you think the Nuremberg Tribunal got it broadly right, then for an experiment to be permissible it has to be both "right" and "good." That is to say, the researcher not only has to have the consent of the experimental subject; she also has to have good grounds for thinking that some good will come of the procedure. To those taught to see deontology and consequentialism as mutually exclusive alternatives, this requirement that the experiment be "both right and good" constitutes an interestingly different hybrid. It bears pondering how many more situations, other than medical experimentation, might manifest that same structure.
References American Medical Association (AMA) (2001), "Principles of Medical Ethics"; available at www.ama-assn.orgama/pub/ categor y/2512.html (accessed Jan 2, 2004). American Medical Association (AMA) (2003), "Code of ethics"; available at www.ama-assn.org/ama/pub/category/8423.html (accessed Nov 7, 2003). Arvon, J. (1997), "Editorial: Including Elderly People in Clinical Trials: Better Information Could Improve the Effectiveness and Safety of Drug use," British Medical Journal, 315, 1033; available at: bmj.bmjujournals.com/archive/7115/7115e5.htm (accessed Nov.) Bayer, A. and W. Tadd (2000), "Unjustified Exclusion of Elderly People from Studies Submitted to Research Ethics Committee for Approval: Descriptive Study," British Medical Journal, 321, 992. Bugeja, G., A. Kumar and A. K. Banerjee (1997), "Exclusion of Elderly People from Clinical Research: A Descriptive Study of Published Reports," British Medical Journal, 315, 1059. Dworkin, R. (1974), "The Original Position," in N. Daniels (ed.), Reading Rawls, (New York: Basic Books). Goodin, R. E. (1985 ), Protecting the Vulnerable (Chicago: University of Chicago Press). Gurwitz, J.H., N.F. Col and J. Avorn (1992), "The Exclusion of the Elderly and Women from Clinical Trials in Acute Myocardial Infarction," JAMA, 268, 1417. Hobbes, T. (1651) Leviathan (London: Crooke). Kipnis, K (2001) "Vulnerability in Research Subjects: A Bioethical Taxonomy," in Ethical & Policy Issues in Research Involving Human Participants (Bethesda, Md.: National Bioethics Advisory Commission, 2001), vol. 2, 174-84; available at wwww.georgetown.edu/research/nrcbl/nbac/human/overvol2.html (accessed Nov 17, 2003). Kipnis, K (2003) "Seven Vulnerabilities in the Pediatric Research Subject," Theoretical Medicine & Bioethics. Kipnis, K. (2004) "Ethical Pitfalls in Waived-Consent Trials," OHRPSRA Conference, Waikiki, Feb 10, 2004. Lewis, J. H. et al. (2003), "Participation of Patients 65 years of Age or Older in Cancer Clinical Trials," Journal of Clinical Oncology, 27 (#7), 1383-9. Mill, J. S. (1859), On Liberty (London: Parker & Son). Nuremberg Tribunal (1949), "Directives for human experimentation," Trials of War Criminals before the Nuremberg Military Tribunals under Control Council Law No. 10 (Washington, D.C.: Government Printing Office), vol. 2, 181; available at ohsr.od.nih.gov/nuremberg.php3 (accessed Dec 28, 2003). Schuch, P. H. (1994), "Rethinking Informed Consent," Yale Law Journal, 103, 899. US Code of Federal Regulations (1996), "Informed Consent of Human Subjects," CFR, 21, 50.24; available at http://history.nih.gov/history/laws/21CFR50.24.html (accessed Feb 21, 2004). US Code of Federal Regulations (2001) "Protection of Human Subjects," CRF, 21, 50 ; available at http://history.nih.gov/history/ laws/21CFR50.24.html (accessed Feb 21, 2004). World Medical Association (1948), "Declaration of Geneva: physician’s oath," available at www.cirp.org/library/ethics/geneva (accessed Jan 2, 2004). World Health Association (1989), "Declaration of Helsinki" (1964, as amended 1975, 1983, 1989); available at ohsr.od.nih.gov/ helsinki.php3 (accessed Jan 2, 2004). Endnotes * An earlier version was presented to the American Philosophical Association, Pacific Division, Pasadena, March 2004. I am grateful for comments there, particularly from Ken Kipnis and Jeff Blustein.
These are the sorts of things catalogued most ably, for the experimental case, in Kipnis 2001; see further Kipnis 2003. Goodin 1985, ch. 5. Goodin 1985, 62-70, discussing doctor-patient relations alongside lawyer-client ones as instances of "professional responsibilities" more generally. Goodin (1985, 62-70) raises doubts about the parallel voluntaristic account of professional responsibilities of doctors toward their patients: there it is argued that the reason doctors have special responsibilities toward their patients is not because of any voluntarily self-assumed obligations but, rather, because patients’ vital interests are particularly vulnerable to their doctors’ actions and choices. Similar issues are raised as regards researchers and experimental subjects in H. Schuch1994. An alternative deontological ethic might for example make its prime directive "respect for persons," which is ordinarily manifested by securing their informed consent before undertaking any medical procedures on them, but can also be manifested in various other ways as well. But if we do not necessarily have to secure a person’s informed consent to a procedure, vulnerability compromising that person’s capacity to give meaningful consent is then not necessarily a problem. To see why vulnerability might be a problem, in deontological terms, we therefore need to focus on versions of deontological ethics that prioritize informed consent. Because, in Hobbes’s (1651, ch. 14) terms, the "sign" does not, in the case of such agents, "sufficiently argue their will." A hard-line deontological analysis would accommodate the latter case by literally equating the consent of a legally authorized representative with the consent of the person whose agent it is: when the agent consents, the person whose agent it is thereby consents. In the absence of that formal legal authorization, however, there seems to be no way on a hard-line deontological ethic for that transfer of agency to be affected. There is no plausible story that could be told about how, "when the surrogate decision-maker has consented, the person on whose behalf the decision is being made has thereby consented." Cf. World Health Association 1989, Principle 11. Nuremburg Tribunal 1949. Kipnis 2001, 177. US Code of Federal Regulations 1996. Kipnis 2004. World Health Association 1964, Principle 10 (emphasis added). That Principle goes on to say that, "In that case the informed consent should be obtained by a physician who is not engaged in the investigation and who is completely independent of this official relationship"; and the next
Principle goes on to say that, "In case of legal incompetence, informed consent should be obtained from the legal guardian in accordance with national legislation. Where physical or mental incapacity makes it impossible to obtain informed consent, or when the subject is a minor, permission from the responsible relative replaces that of the subject in accordance with national legislation."
Thus, for example, under each of the dimensions of "vulnerability" he discusses, Kipnis (2001, 178-9) proposes "measures researchers might take to address" the limitations those vulnerabilities imply for capacity for meaningful consent. In some cases, those measures are designed to make vulnerable people’s consent more meaningful (in the case of cognitive vulnerabilities, e.g., with "plain language consent forms" or "supplementary educational measures") and other times involve measures to make sure that vulnerable agents’ interests are protected, albeit not through the agency of those agents themselves (in the case of cognitive vulnerability again, through "the proper use of surrogates and advocates"). My favorite version remains Ronald Dworkin’s (1974, p. 18): "Suppose you and I are playing poker and we find, in the middle of a hand, that the deck is one card short. You suggest that we throw the hand in, but I refuse because I know I am going to win and I want the money in the pot. You might say that I would certainly have agreed to that procedure had the possibility of the deck being short been raised in advance. But your point is not that I am somehow committed to throwing the hand in by an agreement I never made. Rather you use the device of a hypothetical agreement to make a point that might have been made without that device, which is that the solution recommended is so obviously fair and sensible that only someone with an immediate contrar y interest could disagree. Your main argument is that your solution is fair and sensible, and the fact that I would have chosen it myself adds nothing of substance to that argument."
1 6 "If I consent to your killing me, you would not thereby be permitted to do so. That some deed is okay with me does not always mean it is okay," as Kipnis (2001, 176) observes. The best stor y hard-line deontologists can tell here, presumably, is akin to Mill’s (1859, ch. 5) argument against slavery contracts: respect for autonomy does not oblige us to respect autonomous choices to extinguish autonomy, whether by selling oneself into bondage or by killing oneself either. That would oblige physicians to engage in the cruel prolongation of a terminal patient’s autonomous existence, however painful and ultimately pointless, contrary to the patient’s clear and rational preference for a more dignified end of his choosing. Many hard-line deontologists will reply "quite so" to that proposition, of course; but the reference to "dignity" in that case description might give pause to at least some deontologists who see respect for persons as being linked as much to "human dignity" as it is to "moral autonomy."
Especially, perhaps, the worr y that other interests can outweigh autonomy interests too easily and often, if autonomy is seen as just one interest among many. Unless we make autonomy interests lexicographically prior to all others, we will be unable to satisfy those who demand that no medical procedures could ever be performed without the consent of a person who is actually capable of granting or withholding consent. Kipnis 2001, 176. Nuremburg Tribunal (1949, Directive 9) requires that, "During the course of the experiment the human subject should be at liberty to bring the experiment to an end if he has reached the physical or mental state where continuation of the experiment seems to him to be impossible." "Forgotten" in the literature of medical ethics, if not the practice of IRBs, where it is standard practice to subject proposals first to a consequential-style risk-benefit calculus and then an informed consent test.
Nuremburg Tribunal 1949. See similarly World Health Association (1964), whose Basic Principles 1-8 correspond broadly to the Nuremberg Tribunal (1949) directives above, with Principles 9-12 pertaining to issues of informed consent. Similar conditions to those quoted above are written into the provisions of the US Code of Federal Regulations, both pertaining to protection of human subjects in general and to the waived-consent experiments discussed above (US Code of Federal Regulations 2001 and 1996 respectively). Note that I describe these considerations as "consequential" or "consequentialistic in form" —by which I just mean "outcome-oriented." Why those should matter is, of course, easily analyzed in terms of consequentialistic ethics. But it is of course possible to incorporate them into a deontological ethic as well (as, for example, duties of beneficences are within Kant’s ethics). The task of those preferring to construct a hybrid model from a deontological starting point would be to find some such way of these consequential considerations within that sort of ethic. AMA 2001. World Medical Association 1948. Arvon 1997. Gurwitz, Col and Avorn 1992. Bugeja, Kumar and Banerjee 1997. Bayer and Tadd 2000. Lewis et al. 2003. See similarly Gurwitz, Col and Avorn 1992. AMA 2003, sec. E-2.071. Kipnis 2003.