Palliative Care: Ethics and the Law

Robert A. Burt

Michael K. Gottlieb

During the past 30 years, a dramatic change has occurred in the legal and ethical principles governing provision of care for terminally ill patients. Withholding or discontinuing life-prolonging treatment has now received widespread approval in judicial opinions, state statutes, and the intraprofessional norms of medical practice. The first significant judicial opinion was rendered in 1976 by the New Jersey Supreme Court in the Karen Quinlan case, which authorized the removal of a ventilator from a patient in a persistent vegetative state (1). Numerous state legislatures and courts subsequently endorsed the proposition that where patients or their surrogates requested this course, physicians could withhold or withdraw life-prolonging treatment with no fear of criminal or civil liability (2). In 1990, the United States Supreme Court bestowed ultimate legitimacy on this changed norm of medical practice by ruling, in effect, that mentally competent patients had a constitutional right to refuse life-prolonging treatment (3). Seven years later, the Justices unanimously declined to extend this ruling beyond a “right of refusal;” the court rejected any constitutional right to active acceleration of death by physician-assisted suicide even if requested by a mentally competent patient (4). The court did clearly indicate that state legislatures were free to authorize this practice; to date, only Oregon has done so (5).

From its very beginning in the Quinlan case, however, the “right to refuse treatment” has been afflicted by a disconnect between conception and reality. In authorizing ventilator removal, the New Jersey Supreme Court relied on the principle of respect for autonomous choice. The court gave no weight to the wishes of Karen’s parents or to those of her physicians; the only person with any legally recognizable claim was Karen herself. But Karen, of course, was in no position to make any choice about the continuation of the respirator. She was then incompetent, with no prospect of ever regaining competence; and before her cerebrovascular accident, when she had been competent; she had never expressed any opinion about her wishes in the event she might become ventilator-dependent. The court quickly bypassed the central problem in applying the autonomy ideal to her by positing that

If she had been competent, she would have had a right to choose withdrawal
She should not lose this right “merely” because she was now incompetent and
Her father could exercise this right for her, so long as he acted on the basis of what he believed to be her wishes rather than his own view of her best interests

From its modern origins in the Quinlan case, the autonomy framework for conceptualizing end-of-life decision making has therefore had a distinctly artificial cast of mind. It is only 30 years after Quinlan, however, that we can now clearly see what should have been evident from the beginning: the autonomy framework in the context of end-of-life decision making simply does not fit the facts.

This is not to deny that protecting patient autonomy in end-of-life care, as in all medical treatment and research, is an important principle. Nor does this deny that disregard for patient choice has been a longstanding and unjustifiable feature of medical treatment and research. But in reality, applying the autonomy framework in end-of-life decision making has had little practical effect and much fictitious posturing. Efforts to persuade people to execute advance directives to protect their autonomy if they should become incompetent have essentially failed. Where incompetent patients have previously completed advance directives and treating physicians disregard their clear import, however, courts have been willing to award civil damages (6). The fictive character of these directives is revealed with special clarity in the laws of some 33 states providing that where an incompetent person has not specified a health care proxy in advance, the state will make that choice itself on the premise that most people would want what the state wants for them—that is, spouse first, adult children second, and so forth (7). Most state surrogacy laws require that the surrogate follow a “substituted judgment” standard—that is, the surrogate is obliged somehow to discern what the incompetent patient would have wanted based on prior statements or general values (8). A few states provide that, where the patient’s prior wishes are unclear, the surrogate is permitted to act on the basis of the patient’s “best interests” (9).

The explanation for the failure of the advance directive movement emerged with considerable force in the empirical findings of the study to understand prognoses and preferences for outcomes and risks of treatments (SUPPORT) (10). Carried out in the 1990s, the study tested the most extensive, rigorous effort that had ever been tried to assist critically ill patients and their families in making informed choices about end-of-life care. Notwithstanding the magnitude of this effort toward promoting choice, it produced no effective results. The SUPPORT data instead revealed—in findings that have been subsequently confirmed in other settings—that most patients and their families simply did not want to make decisions about their end-of-life care. Although most patients in the study were persuaded to fill out advance directives, a substantial portion of these patients and their families ignored their prior directives as death drew near. They simply did not want to talk about
the imminently looming reality that they were facing death; and most medical professionals returned the favor with equal reluctance to speak about this inevitability.

There are two ways to respond to the consistently confirmed insufficiency of advance directives and end-of-life planning. One way—the dominant way during the past 30 years—has been to redouble efforts to promote patient and family choice making. The second way is to turn our focus of attention away from the autonomous choice framework in thinking about end-of-life decision making. This alternative is preferable—not to override autonomous choice but to remove this value from the center of attention and to recast our thinking about end-of-life care to promote different, although not necessarily inconsistent, goals.

Our preference for this alternative arises from more than the practical failure of efforts to promote individual end-of-life care choice making. More fundamentally, the pursuit in itself carries substantial social dangers that are likely to yield abuses as bad as, and even directly similar to, the abuses of physician authoritarianism that the autonomy framework was intended to correct. The crucial impetus for the modern embrace of the autonomy framework for terminally ill patients was mistrust of physicians, based on a belief that they regularly disregarded both the wishes and interests of their dying patients by pursuing aggressive, painful therapies with no realistic possibility of success, by withholding effective pain relief generally, and by abandoning their patients when death became patently unavoidable. The equivalent dangers in the autonomy framework arise from the practical reluctance of most people to exercise choice.

That reluctance, we believe, comes from cognitive difficulties that inescapably afflict everyone in contemplating the reality of his or her own death. Proponents of the autonomy framework during the past thirty years have not ignored these difficulties but instead have inveighed against them. From their perspective, the cognitive obstacles to rationalizing death are not insurmountable. According to their view, we should—individually and collectively—end our “denial of death” and reconceive death (as some say it once was conceived in western civilization, or in some other contemporary cultures) as a “natural part of life” to be accepted in the same way that we accept any inevitable biological givens. We suggest, however, that there is more at play in the avoidance of death’s biological inevitability than the simple fear of it. There is, more importantly, a cognitive drag on our ability to comprehend death. We may parrot the language of rational choice in comparing our fears about death with our fears about continued life. We may enact a convincing appearance of autonomous choice in contemplating death. But it is very difficult, at the core of our thinking, to convince ourselves that death is rationally comprehensible. Death is more than a future condition fraught with uncertainties about benefits and detriments. It is more than the absence of life. It is the absence, the intrinsic contradiction, of meaningfulness. The very concept of the choice-making self, the construct on which the autonomy principle depends for its coherence, is radically unsettled—even made incomprehensible—by the actual, imminent approach of death.

The more conventional view is that some people may be afflicted with this difficulty but some—perhaps many or even most—are not; and the task for application of the autonomy principle is to devise guidelines for distinguishing those who are and are not “competent” to exercise rational choice. But the difficulties of drawing this distinction are so profound and the consequences of our inevitable failures to make convincing distinctions are so grave that we should not put this differentiating enterprise at the center of our practices about end-of-life care.

We believe that the most convincing explanation for the medical abuses inflicted by physicians on dying patients is the physicians’ own sense of the “wrongness” of death. The incomprehensibility of death readily translates into a conviction that death is a kind of grammatical error, a misfit in a world that can be rationally comprehended. In the medical lexicon, death is therefore understood as an error to be corrected, opposed, and negated. Displacing physicians and blaming their commitment to rational mastery over death does not, however, cure the problem posed by the incomprehensibility of death. Its conceptual status as a grammatical error leads to medical triumphalism and the abuses of dying patients that follow from this relentless warfare against error. Its incomprehensibility translates with equal ease in lay terms to a conviction that death is “wrong” as a moral proposition. Even if one’s rational intellect can comprehend that there are worse things than death, and that it is nonsensical to imbue a biological inevitability with moral condemnation, nonetheless a persistent undertow pulls continuously in the opposite direction. This moral ambivalence toward death might be consciously denied. And some people may be more capable than others of rigidly maintaining this denial into the maw of death itself. But for most people, successful resistance to this moralized understanding is akin to success in refusing to think about elephants in response to a command that you must not—whatever you do, you must not—think about elephants.

The actual consequences of thinking forbidden thoughts about elephants are much less fraught with danger than thinking forbidden thoughts about the moral wrongness of death. If actual death is a moral wrong and you cannot avoid dying—indeed, if you actively embrace dying—then it follows that some wrong has been committed and someone must be punished for this wrongdoing. Physicians might, of course, punish their dying patients for this transgression. But when physicians are removed from their previously central choice-making role, the impulse to blame does not vanish; it simply moves to a new target. The ironic consequence of the autonomy principle—that decisions about death are the legitimate prerogative of no one but the dying person—is that blame too will be attached only to the dying person. As the choice belongs to the individual, the punishment will be individually self-inflicted. The precise content of this punishment would vary—perhaps patients’ insistence on aggressive and painful, although patently futile treatments, perhaps their refusal to request effective pain relief, perhaps their embrace of premature death. But in all such cases, the abuse previously inflicted by physicians on dying patients will reappear, for the same underlying reasons, as abuse inflicted by dying patients on themselves.

Abuse of dying patients—either self-inflicted or iatrogenic—is not, however, inevitable. It is only ambivalence about death—some lurking, ineradicable sense of its wrongfulness juxtaposed against all rational arguments for its inevitability and even preferability—that is inevitable. The impetus for turning this ambivalence toward abuse is denial—not “denial of death,” in the conventional sense of that cultural construct, but denial of the negative valences in the ambivalence toward death. Like trying not to think about elephants, the negative valences cannot be entirely repressed; and if they are banished from consciousness by a single-minded insistence that death is “good” or “dignified” or “accepted,” the unconsciously buried sense of wrongness and guilt accompanying death will push toward expression in action. This is the dynamics by which an unacknowledged sense of wrongdoing and guilt expresses itself by wrongful action that implicitly invites condemnation whereas, at the same time, the action is explicitly enshrined in protestations of righteous conduct.

The challenge for social regulation of end-of-life care is to identify the circumstances in which this malign dynamic is likely to take hold and to design countervailing schemes. Reliance on patient autonomy is not an effective countervailing
scheme, any more than the now-discredited reliance on physician autonomy for deciding whether and when death should occur. We believe, however, that the following three principles are alternatives responsive to the current overinvestment in patient autonomy:

No one should be socially authorized to engage in conduct that directly, purposefully, and unambiguously inflicts death, whether on another person or on oneself.
Decisions that indirectly lead to death should be acted upon only after a consensus is reached among many people; no single individual should be socially authorized to exercise exclusive control over decisions that might lead to death, whether that individual is the dying person, the attending physician, or a single family member acting as health care proxy.
As much as possible, end-of-life care should not depend on explicit decisions made at the bedside of a specific dying person but rather should be implicitly dictated by systems-wide decisions about available resources, personnel, and institutional settings—that is, by setting up default pathways that implicitly guide and even control caretaking decisions in individual cases.

Principle 1: No Infliction of Death

The rationale for the first principle is that the direct, purposeful, and unambiguous infliction of death leaves no psychological space for acknowledged ambivalence. Whether the infliction is carried out on oneself or on others, it demands an unambivalent claim of rightness and righteousness that is psychologically impossible and therefore invites self-contradictory expressive actions. (The clearest model for this demand is in the administration of the death penalty, where the possibility of executing an innocent person must be entirely excluded and yet the difficulties of obtaining this ironclad assurance lead to a lingering sense of guilt and abusive inflictions in the enterprise.)

Our current regulations for end-of-life decision making do offer psychological space for acknowledged ambivalence in various ways. The rules that permit withholding or withdrawing life-sustaining care provide some comforting assurance that these actions do not in themselves inflict death because the underlying illness is the cause of death. At the same time, the logical tenuousness of this reasoning promotes conscious acknowledgment of ambivalence—that is, of the close proximity of these actions to wrongful conduct. (This protective dynamic is compellingly revealed in a case presentation by Miles Edwards and Susan Tolle, describing their actions and reactions in removing a competent, conscious postpolio patient from a ventilator, in response to his insistent request. Although rationally convinced of the moral correctness of this course, Drs. Edwards and Tolle report the sense of wrongdoing—of “purposeful killing” that nonetheless attended their actions and that powerfully troubled them (11).

The logical tenuousness of the distinction between relieving pain and hastening death in the high-dosage administration of opioids to dying patients—the so-called “double effect” principle—has the same psychologically protective function, serving simultaneously as permission and a warning sign about dealing with death. According to the double effect principle, consequences that would be wrong if caused intentionally become acceptable, even when foreseen or indeed expected, if the actions creating those consequences were intended for a morally permissible purpose (12). Typically, reliance on the principle of double effect requires the presence of the following four conditions:

The act creating the risk of adverse consequences must be beneficial, or at least morally neutral
The actor must intend the beneficial effect and not the harmful effect, although the harmful effect may be foreseen
The harmful effect must not be a means to the beneficial effect and
The beneficial effect must outweigh the harmful effect (13)

In the context of palliative care, under this principle, a physician may administer pain medication to a patient, knowing that medication may hasten the patient’s death, so long as the intent of the physician is solely to relieve the patient’s pain. In this scenario, the “beneficial” effect of pain relief is intended, whereas the “harmful” effect of death is merely foreseen.

The protective function of logically tenuous rules tends to erode over time, as their routine application dulls everyone’s sense of the close correspondence between permitted and forbidden conduct authorized by these rules. The clearest indication of this erosion is in the arguments put forward with increasing social acceptance by advocates for physician-assisted suicide and euthanasia. These advocates insist that withholding or withdrawing life-sustaining treatment or application of the “double effect” principle are logically identical to purposeful, unambiguous infliction of death, and that this logical identity means that all these steps are morally equivalent and morally correct. These contemporary advocates fail to see that, far from justifying this “next step” toward purposeful killing, the plausibility of their logical claims about existing practices should raise concerns that these practices have themselves lost their function as protective expressions of ambivalence toward death. Our guiding principle for social regulation should be that the more comfortable clinicians and patients are with actions implicating death, the more socially dangerous these actions become. Preserving these “illogical” lines between accepting and hastening death—between physician-assisted suicide and withholding or withdrawing treatment or administering high-dosage opioids—is in the service of promoting conscious awareness of moral discomfort. Eliminating this discomfort, as urged by advocates for physician-assisted suicide and euthanasia, is logical but terribly wrong—and socially dangerous because the unconsciously buried conviction of wrongdoing ultimately will express itself in eruptions of blameworthy conduct.

Principle 2: Consensus Decision Making for Death

Our second principle, that social regulations should not designate any single individual to exercise exclusive control over decisions that might lead to death, would require a more radical departure from existing arrangements. Forged on the anvil of autonomous individual choice, existing arrangements search relentlessly for a single designated decision-maker based on a clear-cut hierarchy of authority. The desperate intensity of this search is revealed by the state laws, noted in the preceding text, that denominate proxy decision makers even where an incompetent patient has made no prior selection. In particular, this intense search is apparent in the provision of those laws regarding multimember proxies, such as parents or children or siblings; the laws typically specify that for this class of proxies, majority vote shall prevail and, in the event of tie votes, the class is disqualified from decision-making authority. The implicit goal in these laws is not simply to find some single decision maker but to find an unambiguous choice about life-sustaining treatment.

Only gold members can continue reading. Log In or Register to continue