Does autonomy require informed and specific refusal of life-sustaining medical treatment.

• Author: Bradley, Gerard V.

Does Autonomy Require Informed and Specific Refusal of Life-Sustaining Medical Treatment?

Any social order that prizes individual autonomy is committed to collectively informing its autonomous members. A good working definition of "autonomy" is "self-determining free choice." What distinguishes "choice" from mere appetite or desire is deliberation upon known, available options and their probable consequences, and a practical equilibrium of external pressures for and against those options.

Even the most autonomu-driven society will conclude that some of its members - e.g., infants and persons with mental disabilities - are unable to choose for themselves. Autonomy for persons who are incompetent is inapposite because they are incapable of exercising free choice. Thus constitutional rights derived from our commitment to autonomy - which some courts say include the right to forego life sustaining medical treatment - are exercisable only through the "knowing and voluntary" choices of competent individuals.(1) Distinguishing exactly who is competent, and whether particular proposals of competent persons are informed free choices, are collective decisions. "Collective" means society, acting through its common agent, the state and its laws. The law, and not those individuals, decides whether an effective exercise is made. At first glance then, autonomy would seem to require demonstrated assurances to society that any refusal of life-sustaining treatment is with the informed consent of the competent person affected.

Deciding what those assurances entails prompts a look at other aspects of our common life in relation to autonomy. We possess an abundance of examples. The marketplace is our dominant metaphor for collective existence, and not primarily for goods and services. Our most abiding market commitment is actually to the marketplace of ideas that a system of free expression implies. This prevailing first amendment metaphor aptly captures our constitutional law not only of free speech but also of church and state. The Supreme Court's Religion Clause jurisprudence has been little more than elaboration of religious autonomy, in the specific sense that what individuals believe about religion, if they believe anything at all, is necessarily a matter of individual choice. This choice may not be burdened by even faint traces of government endorsement of one faith, many faiths, or faith generally over disbelief.(2) The government here, as in other areas in which individual autonomy prevails, must be neutral about the outcomes of individual choices.

But neutrality does not mean passivity. Ours is a regime with a robust (if imperfect) commitment to individual autonomy, yet that government is busy indeed. The business of our government is not only to make trains run on time and to fill potholes. Even the most libertarian of economic theorists, where the implications of autonomy have been most fully explored, recognize that we live in a world of inadequate information. To the extent that is true, ours cannot be a world of autonomous free choice. Many of them recognize a potentially very large government role in empowering individuals, through information provision, to make autonomous choices. This is so despite these theorists' general skepticism of government which is such that, if Uncle Sam said at midnight that it was dark outside, they would go to the window to check. (The lawyer of course would ask Sam about the latitude and time of day and year.)

Government's role here is hardly limited to consumer advisories. The massive redistributive policies of the modern welfare state are most often justified as providing the material preconditions to individual pursuit of the individually chosen good life. Our massive system of compulsory public education is thought compatible with autonomy because it provides the nonmaterial preconditions to a life of intelligent, voluntary, and knowing choices. Interestingly, the libertarian critique here is most often confined to the question of whether education need be state administered as well as state subsidized. The argument for education vouchers, for instance, is an argument from autonomy which does not argue against compulsory schooling as such.

The critical implication is that a steadfast commitment to autonomy is not only compatible with but actually requires massive collective or state involvement. This conclusion is buttressed by empirical research in heuristics and in the sociology of knowledge, research which reveals even sophisticated adult human beings as beset by cognitive limitations and enmeshed in a host of dimly perceived and certainly not chosen pre-analytical commitments.(3) Add to these observations the limitations imposed upon individuals by their parentage, native abilities, compulsions, and chance misfortunes, and one comes up with a real world in which autonomy is something that a government must work to secure for its citizens.

Government does not secure autonomy by leaving its citizens alone as some courts would have it, as if autonomy were the residue of state inaction. It is neither theoretically nor practically true that where the state is held at bay, citizen behavior is therefore a matter of individual autonomy. The assumption that a "night watchman" state guarantees individual autonomy simply does not describe the contemporary regime. Characteristically that regime both eschews imposition of an overriding conception of the good or correct choice and recognizes that state action may be necessary to empower the individual against a plurality of nongovernmental, extra-individual forces.

If autonomy is in principle compatible with state maintenance of certain conditions that make it possible, what form should that maintenance take where a choice to forego life-sustaining medical treatment is proposed? Against a multitude of variables affecting the answer to that question, there is one invariable principle to which autonomy in different spheres attests: the more serious the consequences to the individual, the greater demands autonomy makes upon the state to ensure that a truly free and knowing choice is made. This demand is not a limitation upon possible choices due to their "incorrectness" or their consequences for others. It is not a state "trump" of autonomous choice because the choice "harms" another, a trump justified by the regime's obligation, born of equal concern and respect for all citizens, to ensure that an individual's maximum autonomy remains consistent with like autonomy for others. It is rather an autonomy limiting regulation that remains indifferent to the proposed choice's outcome.

As a practical matter the question of how "informed" and how "voluntary" a choice to forego life-sustaining treatment should be, is comprised of sub-inquiries such as these: (1) What kinds and how much evidence of choice are sufficient to show that it is indeed the choice of this self-governing individual? (2) What other options and probable consequences, and in how much detail, need be described to (or demonstrably within the ken of) the chooser? (3) What systemic and cognitive distortions are likely at work and how should they be combatted? (4) Does our response to (3) oblige us to discount the choice? and (5) In which direction - for or against implementation - should doubts about any or all of these sub-inquiries cut? None of these questions can be answered autonomously; that is, through a system of unregulated individual choices. The individual is not generally permitted to choose how "knowing" and "voluntary," in all these inseparable connections, his choice should be, and autonomy does not require it. Were he so authorized, there would not be a social order devoted to autonomy, but no social order at all. In any event, such a nonorder would be unintelligible. The relevant question in it would be: Did the agent knowingly and voluntarily choose that a life or death decision was trivial, one to be made whimsically? In other words, there can be no effective postponement of collectively imposed preconditions to effective choice.

No party to this debate has proposed that whimsical decisions need be implemented out of respect for autonomy. So that point will not be pursued

further. It is agreed that there are circumstances in which a decision to forego life-sustaining treatment should be respected as an expression of autonomous choice. The author's opinion, in brief, is that the appropriate informed choice criteria are those governing other serious medical procedures. Cases requiring less evidence of informed consent for decisions to forego treatment than to consent to it are illogical. These decisions are really flipsides of the same subject. The decision for treatment presents the same calculus of extent of invasion, probable effects, chances for recovery, and, most important, the consequences, including death, of not consenting. The suggestion here is that decisions to consent/not consent are equivalent to decisions to forego/not forego treatment.

What should be done with the more typical case of an incompetent patient? The first thing we must do is resist the temptation to make autonomy do work - resolve hard questions - it cannot do. We must recognize that for many of these patients, autonomy is inapposite. Saikewicz(4) is just an extreme example of how judges have used autonomy as a legal fiction to shield from view the true bases of their opinions. Such persons must be cared for according to collectively decided principles of respect for human life and death. Substituted judgment also substitutes fiction for sound legal judgment. Courts have sometimes mastered the alchemist's craft by transforming a decision by B to end the life of A (who is incompetent) into an exercise of A's "autonomy." The supposed logic here, is that autonomy does not cease with the end of the agent's capacity to choose. Well, it does. This whole situation is actually...