In a recent special supplement to the Hastings Center Report entitled “Improving End-of-Life Care—Why Has It Been So Difficult?” Robert Burt wrote the following in an essay ominously entitled “The End of Autonomy”:
(1) No one should be socially authorized to engage in conduct that directly, purposefully, and unambiguously inflicts death, whether on another person or on oneself.
(2) 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 family member acting as health care proxy.1
Burt RA. The end of autonomy. Hastings Center Report 2005;35(Special Supplement):S9–S13.
Taken together and applied consistently and with conviction, there can be no reasonable doubt that these principles would, indeed, end autonomy in end-of-life care as we have come to know it. In the second point, Burt revives a thesis he initially propounded nearly 30 years ago in his book Taking Care of Strangers.2
Burt RA. Taking Care of Strangers—The Rule of Law in Doctor–Patient Relations. New York: The Free Press; 1979.
The first point, the one on which I focus in this article, resonates with a thesis Burt propounded in his much more recent book, Death Is That Man Taking Names: Intersections of American Medicine, Law, and Culture,3
Burt RA. Death Is That Man Taking Names: Intersections of American Medicine, Law, and Culture. Berkeley: University of California Press; 2002.
Nothwithstanding the rational implausibility of believing that death is a morally neutral biological fact or even a morally preferable outcome in some circumstances, we cannot readily erase a persistent contrapuntal conviction that death as such is inherently wrong—a logical and moral error.4
See note 3, Burt 2002:4–5.
Such a perspective—death as a cosmic shame—to borrow Ronald Dworkin's marvelous term, is quite consistent with the reigning curative (in contradistinction to the palliative) model of medicine, which views death as the ultimate enemy and the essence of medical failure. I leave for another discussion several not insignificant semantic quibbles, such as how Burt would have us understand “socially authorized,” or who might be the “many people” who must reach consensus for any decision to be made that might indirectly lead to death. With regard to the latter, we can only hope that he is not referring to the “many people” who deemed it within their purview to insinuate themselves into the decision to withdraw Terri Schiavo's feeding tube.
Embedded in Burt's concise mandate are two persistently problematic elements with profound legal and moral significance—causation and intent. Because commentators in the crucible of the climax and denouement of the Schiavo saga played unconscionably fast and loose with both of these, there is a need to take a fresh look at how we think about them in the context of caring for patients with grave or terminal illness. I argue that it is this underlying predisposition to view death as “a logical and moral error” that has motivated our efforts to manipulate the application of causation and intent so as to distance healthcare professionals from the dying process in a manner and to a degree that does not comport with reality or the needs of patients.
I first consider the concept of causation and its traditional role in ascertaining an individual's culpability for bringing about a particular consequence or state of affairs. Of particular concern is the role of causal analysis in clinical cases involving a patient's death. Second, I compare and contrast two approaches to the question of intent—the traditional one taken by the criminal law and that adopted by the Doctrine of Double Effect. Ultimately, I argue that neither causation nor intent can be ascertained with sufficient certainty so as to provide an adequate foundation for the critical moral distinctions involving dying and death in the clinical setting that have often been based upon them. Finally, I suggest that because of the inherent complexities, vagaries, and ambiguities of causation and intent in human behavior, particularly in the care of persons confronting grave or terminal medical conditions, they should play less of a role than we have formerly ascribed to them in establishing and maintaining the moral parameters of such care. Instead, we should establish those parameters based on a more realistic recognition of human mortality and the many ways in which patients are at risk of falling victim to, and hence may reasonably seek a physician's assistance in avoiding, a medical fate worse than death.
Actual and Legal Causation in Medicine
One of the problems in contemporary considerations of causation is the conflation of two separate and distinct domains—those of fact and value. Standing alone, the term “cause” projects an aura of fact. Determining the cause of a person's death, for example, would seem to be an inquiry into matters of fact, not value. After all, the proper goal of such an inquiry, or so we have been inclined to think, is not to discover what should be considered the cause of death, but rather what actually was the cause. However, the answer to such ostensibly simple and straightforward questions of causation turns out to be neither. The basis for and context of the inquiry into the cause in fact of an event is of considerable consequence. To quote one prominent legal commentator:
Much misunderstanding between lawyers and physicians could be obviated if members of both professions would realize that “simple” causation is not merely an abstract issue of fact and that the resolution of the cause problem depends largely upon the purpose for which cause is to be used. What is a cause for the judge need not be a cause for the physician. It is through the process of selecting what is to be regarded as a cause for purpose of resolving a legal dispute that considerations of policy exert their influence in deciding an issue of cause-in-fact.5
Malone, WS. Ruminations on Cause in Fact. Stanford Law Review 60 (1956).
For further insight into how a judge might seek to ascertain what constitutes a cause in a medico-legal case, we can look to the writings of William Prosser, the most esteemed American authority on the law of torts: “Causation is a fact. A cause is something regarded as a necessary antecedent, something without which the event would not have occurred. The term includes all things which have so far contributed to the result as to be essential to it.”6
Prosser, W. Proximate Cause in California. 38 California Law Review 369, 375 (1950).
Smith, J. Legal Cause in Actions of Tort. 25 Harvard Law Review 103, 104 (1911).
To more fully appreciate the exquisite complexity of causation in medical–legal cases, let us consider that of Shorter v. Drury.8
695 P. 2d 116 (Wash. 1985).
During the D&C, Dr. Drury severely lacerated Ms. Shorter's uterus, resulting in significant bleeding. She and her husband were advised that without a transfusion there was a likelihood that she would die. They were urged to relent and allow the transfusion, but they did not and she did, in fact, die. What was the actual cause of Doreen Shorter's death? Was it the laceration of her uterus by Dr. Drury, which caused the substantial bleeding that necessitated a transfusion? Or was it her informed refusal of a transfusion, which expert witnesses at the trial testified to a reasonable degree of medical certainty that “but for” she would have survived?
In the subsequent wrongful death action Mr. Shorter pursued against Dr. Drury, he alleged that his wife's death was proximately caused by Dr. Drury's negligence in performing the D&C. The jury found Dr. Drury to have been negligent, and that his negligence had proximately caused Doreen Shorter's death, for which they assessed damages of $412,000. The jury also found, however, that the Shorters had “knowingly and voluntarily” assumed the risk that she might bleed to death were a transfusion to be required during or after the procedure. The recognized risk of bleeding was, after all, independent of any negligence on the part of the surgeon. That is the essence of a “risk of the procedure.” The jury attributed 75% of the cause of Ms. Shorter's death to the refusal of the transfusion. Consequently, Mr. Shorter's recovery was reduced to $103,000.
Both parties appealed, and the state supreme court affirmed the trial court judgment. Ultimately, matters of public policy proved to be much more decisive than matters of causation in determining the outcome of the decision on appeal. The court rejected Dr. Drury's argument on appeal that the release executed by the Shorters constituted a waiver of their right to sue Dr. Drury for negligence and recover damages for any harm that was proximately caused by it. The appellate court also rejected Mr. Shorter's argument that the release was only for harms resulting if the transfusion were required despite the nonnegligent performance of the procedure. The court expressed the belief that releases of liability such as the Shorters signed must be given full force and effect so that physicians will be able to provide necessary and appropriate medical treatment to patients under precisely such circumstances.
Shorter v. Drury is a prime example of why determinations of causation often are mixed questions of fact and value (or policy), and why a narrow and rigid focus upon the cause-in-fact may lead to unsatisfactory outcomes in contested cases. Although there is no question that Dr. Drury's negligence was the proximate cause of the laceration of Ms. Shorter's uterus and hence of her bleeding, his negligence was neither a necessary nor a sufficient condition for her death because the expert testimony at trial made it clear that “but for” her refusal to allow a transfusion, she would have survived. Both the jury and the state supreme court firmly believed that justice as between the parties would not have been done if the disposition of the case were determined entirely by the answer to the question of actual or proximate cause of Ms. Shorter's death. As in the Shorter case, in end-of-life care, values and social policy should be important factors, not necessarily in arriving at the cause in fact of a person's death, but in the determination of whether or not to ascribe moral and or legal responsibility to someone for the patient's death.
One reason for the complexity of causation questions in the medical domain is that, in both law and bioethics, we are concerned about not merely acts and their consequences, but omissions and their consequences. The distinction here is sometimes made by referring to acts as “simple causes” and omissions as “hypothetical causes.”9
Becht AC, Miller FW. The Test of Factual Causation in Negligence and Strict Liability Cases. St. Louis: Washington University Press; 1961:24.
The Nexus between Causation and Intent in End-of-Life Situations
The Schiavo case brought a renewed vigor (some might call it a frenzy) to the debate over the legitimacy of the active versus passive distinction, or killing versus allowing to die. This distinction has deep roots in the efforts of bioethicists to determine when moral responsibility for the death of or harm to a patient can be reasonably ascribed to another person, particularly a healthcare professional. Right-to-life advocates unequivocally asserted that removing Ms. Schiavo's feeding tube was an affirmative act that must be intended to bring about, and would invariably result in, her death (most likely from dehydration). Yet in the nomenclature of euthanasia, both withholding and withdrawing life-sustaining interventions are routinely characterized as “passive” euthanasia, in contradistinction to the administration of a lethal injection, which is characterized as “active” euthanasia. One ardent critic of such active versus passive distinctions is Supreme Court Justice Antonin Scalia, who formulates his assault on the distinction in the following terms: “Starving oneself to death is no different from putting a gun to one's temple…. [T]he cause of death in both cases is the suicide's conscious decision to put an end to his own existence.”10
Cruzan v. Director, Missouri Department of Health, 497 U.S. 261, 296–297 (1990) (Scalia, A., concurring).
[K]illing versus letting die are properly applied only to a circumstance in which one person intentionally causes the death of another human being…. One person can kill another by intentionally allowing the other to die, and killing can occur by omission just as well as by commission…. Law, medicine, ethics, and ordinary language all recognize that some forms of allowing to die constitute acts of killing (emphasis added).11
Beauchamp TL, Childress JF. Principles of Biomedical Ethics, 4th ed. New York: Oxford University Press; 1994:220.
Returning to the Schiavo case, or the case of Nancy Cruzan, which was the one under consideration when Justice Scalia offered his critique of the distinction, the argument as to why it is ethically acceptable to withdraw a feeding tube from either of these permanently vegetative patients but ethically unacceptable (for example) to administer a toxic substance by means of the feeding tube relates much more to the intent of the person who withdraws the tube or utilizes it as a route of administration of a lethal substance than it does to matters of causation. In a strict sense, there is nothing passive about removing a surgically implanted feeding tube or extubating a patient undergoing mechanical ventilation. Similarly, the consequence of any of these measures is the death of the patient. Removing a feeding or a breathing tube from a patient for whom it is an essential form of life support constitutes a minimally sufficient condition for bringing about the patient's death.
Defenders of the killing versus letting die distinction have sought to attach great significance to the patient's underlying medical condition that has necessitated the life-sustaining intervention such as artificial nutrition and hydration or mechanical ventilation. It is, so the argument runs, not the withdrawal of the intervention, but rather the underlying disease necessitating the intervention in the first place that causes the patient's death. Yet if we apply traditional legal causation analysis, but for the withdrawal of the intervention the patient would not have died when and as she did. Similarly, extubation of a ventilator-dependent patient or discontinuation of artificial nutrition and hydration constitute substantial factors in the patient's death at that time, arguably more substantial than the underlying disease, whose lethal potential was being successfully resisted by those very interventions.
Through the shift from causation to intent, those who support the right to refuse or “appropriately” discontinue treatment but who oppose physician-assisted suicide and voluntary active euthanasia seek to persuade us that in withdrawing life support the physician is merely respecting the patient's right to refuse treatment, whereas in physician-assisted suicide or voluntary active euthanasia the physician's intent must necessarily be to kill the patient. To assess the legitimacy of this move, let us approach intent as we did causation, by considering the traditional view of the law of intent.
Intent in Law and Bioethics
A reasonable place to begin this inquiry is with the analysis of responsibility in the criminal law and the dichotomy of the objective and subjective, or physical and mental, components of a criminal offence. The objective component is the actus reus, for example, homicide in the case of murder, breaking into the real property of another in burglary. The subjective component is mens rea, or criminal intent. Such intent relates to the act which is a violation of the criminal law, which does not require the specific intent to violate the law. Criminal liability generally requires both a voluntary act (as opposed to an omission) and the requisite intent to bring about the consequences that constitute a violation of the law. Intent, however, is a subjective state of mind known for certain only to the person whose intent is the object of the inquiry. Proving beyond a reasonable doubt (the burden of proof in a criminal case) that an individual had a particular state of mind (mens rea) at the time an action was undertaken would be an insurmountable task. Consequently, the law has traditionally invoked a presumption so as to enable the state to have a reasonable chance of meeting its burden of proving not only that the actus reus was perpetrated by the defendant, but at the relevant time it was accompanied by the requisite intent. The formulation of this presumption is usually in the following terms: “everyone is presumed to intend the natural and probable consequences of his act.”12
Perkins RM. Criminal Law, 2nd ed. Mineola, NY: Foundation Press; 1969:747.
There is a stark contrast between this legal presumption and the approach of the Doctrine of Double Effect (hereinafter double effect) that has been so frequently invoked in bioethical discussions, particularly those involving care of gravely ill or dying patients. Indeed, the almost reflexive manner in which the double effect finds its way into the ethics of end-of-life care belies its confusing origins and the persistent controversy over whether the doctrine can ever, and if so, under what circumstances, be consistently and coherently applied. A thorough discussion of the historical origins of double effect in the medieval natural law tradition exemplified by Thomas Aquinas, as well as the finer points of the enduring controversy, are beyond the scope of this article.13
Such a discussion can be found in Woodward PA, ed. The Doctrine of Double Effect—Philosophers Debate a Controversial Moral Principle. Notre Dame, Ind.: Notre Dame University Press; 2001.
- The act in question is morally good or morally neutral.
- The good consequence is not brought about by means of the bad one.
- The actor does not intend to bring about, although she may reasonably foresee the likelihood that the bad consequence will naturally follow from her act.
- The good and bad consequences must be roughly commensurate.14
Solomon WD. Double effect. In: Becker LC, Becker CB, eds. Encyclopedia of Ethics. New York: Garland Publishing Co.; 1992:268–269.
What is most remarkable about contemporary invocations of double effect is not that they are common in, for example, Catholic bioethics, but rather that they are even to be found in such influential secular sources as the U.S. Supreme Court decisions in Glucksberg and Quill15
Washington v. Glucksberg, 521 U.S. 702 (1997); Vacco v. Quill, 521 U.S. 793 (1977).
The AMA EPEC project was a nationwide series of “train-the-trainer” regional conferences sponsored by a generous grant from the Robert Wood Johnson Foundation. Double effect appears in EPEC modules that suggest significant clinical, legal, and moral distinctions between aggressive pain and symptom management and withholding or withdrawing life-sustaining treatment on the one hand and physician-assisted suicide (providing a lethal prescription) and voluntary active euthanasia (providing a lethal injection) on the other hand.
Applying Causation and Intent Analysis to Terminal Sedation and Physician-Assisted Suicide
Terminal sedation involves the total sedation of the patient because of intractable pain or suffering. Once complete sedation is achieved, thereby eliminating an experience of pain or suffering, nutrition and hydration are withheld or withdrawn and, depending on how close to the point of death from the underlying disease the sedation is administered, the patient may die from either the underlying disease or, like Nancy Cruzan and Terri Schiavo, from dehydration. Regardless of whether the patient dies of the underlying terminal condition or from dehydration, double effect is invoked to morally justify the physician's conduct. The physician's intent in totally sedating the patient is presumed to be the relief of otherwise intractable suffering, and the intent in withholding or withdrawing artificial nutrition and hydration is presumed to be either complying with the patient or surrogate's request (thereby respecting the patient's autonomy) or declining to provide that which no longer provides a medical benefit, inasmuch as there is no expectation that the patient will ever regain consciousness prior to death. Although the patient's death, probably from dehydration in the terminal sedation case, is inevitable, the physician's intent is presumed to be respecting the patient's autonomy right to refuse medical interventions.17
A concise discussion of the psychological complexities of intent in the clinical setting can be found in Quill TE. The ambiguity of clinical intentions. New England Journal of Medicine 1993;329:1039–40.
In the case in which a physician honors a patient's request for a lethal prescription, however, the physician's intent, and probably the patient's as well, are presumed to be to cause death rather than to relieve or escape from suffering or to respect the autonomous choice of the patient to determine the time and manner of her own death. Furthermore, when a patient brings about her own death through the ingestion of a lethal prescription, a second element of double effect is violated as well, in that the good outcome of an escape from intractable suffering is a direct result of the purportedly bad outcome, causing the patient's death, because the direct and immediate cause of the patient's death is, of course, the ingestion of the lethal prescription.
In the case of physician-assisted suicide, sound argument can be offered that the causal field must bring to the foreground, not push to the background, the imminent death of the patient from the underlying terminal condition. The focus of both the dying patient and the attending physician, on the question of intent, can reasonably be characterized not as bringing about the patient's death but rather orchestrating the patient's dying and ultimate death in a manner that is consistent with the patient's values, goals, and perception of her own best interests.18
The original development of the concept of orchestrating the end of life appears in Loewy EH, Loewy RS. The Ethics of Terminal Care—Orchestrating the End of Life. New York: Kluwer Academic/Plenum Publishers; 2000.
Details on the series at www.pbs.org/wnet/onourownterms/.
In the case of both the lethal medication and the withdrawal of life support, the direct and immediate cause of death involves a medical procedure, either swallowing a prescription drug or discontinuing a life-sustaining intervention. But the rightness or wrongness of the outcome can and should be determined not by the simple and relatively straightforward fact that it causes death, but rather by the extent to which it accommodates the patient's authentic wishes and circumstances, and by a more expansive view of the role and responsibility of the physician in the care of such patients than that of merely prolonging life. Reintroducing the proper role for mixed questions of fact and value that we first encountered in our discussion of Shorter v. Drury, the causal connection between a physician's act or omission and a patient's death may play less of a role in our ultimate determination of moral and legal acceptability of the outcome than the degree to which the physician was seeking to respect the patient's values as she confronts her inevitable death from a terminal condition.
In the case of physician-assisted suicide, it is the ingestion of the lethal prescription by the patient, not the writing of the prescription by the physician, which sets in motion the death-inducing properties of the prescription drug. For that reason, it is particularly curious why the focus is so obsessively upon the intent and act of the physician in writing the prescription and not upon the intent and act of the patient in requesting and ingesting it. Just as the laceration of Ms. Shorter's uterus could not have resulted in her death from exsanguination unless she refused a transfusion, a physician's prescription can have no lethal effect unless the patient chooses to ingest it.
Conclusion
From the analysis of the cases considered in this article, we have seen that the legal culpability of an individual does not arise merely from engaging in an act with a particular consequence or doing so with a specific intent. Something else is required, that is, a prior determination that bringing about that consequence under such circumstances is morally wrong or violative of a legitimate public policy. In the context of patients at or near the end of life, the wrongness of a hastened death cannot be determined merely by analyzing the causal mechanism of that death. Perhaps, to some degree, it can be illuminated by examining the intent of the actor. However, as I have noted previously, when the act takes place in the context of the physician–patient relationship, and particularly if the circumstances involve what we call physician-assisted suicide, the intent and the acts of both individuals in that relationship are implicated in bringing about the final outcome. Consistent with the analysis of intent and causation that I have offered concerning patients confronting death from grave or terminal medical conditions, it is not simply physician-assisted suicide that may reasonably be found to constitute “conduct that directly, purposefully, and unambiguously inflicts death” in Burt's phraseology. That same characterization can be reasonably applied to extubating a ventilator-dependent patient, removing the gastrostomy tube from Nancy Cruzan or Terri Schiavo, and providing terminal sedation to patients with intractable suffering. In all of these situations there are legitimate issues regarding the directness, purposefulness, or ambiguity of these actions on the part of healthcare professionals.20
The most cogent arguments of this sort can be found in the following selections: Beauchamp TL, Childress JF. Principles of Biomedical Ethics, 4th ed. New York: Oxford University Press; 1994:228–41; Brock DW. A critique of three objections to physician-assisted suicide. Ethics 1999;109:519–47; McMahan J. The Ethics of Killing—Problems at the Margins of Life. New York: Oxford University Press; 2002:455–85.
Finally, the sense of wrongness of the outcome cannot be ascertained merely from our “persistent contrapuntal conviction that death as such is inherently wrong—a logical and moral error” to which Burt refers. Moreover, we cannot right that purported wrong by arbitrarily declaring and then coercively enforcing an “end to autonomy” at the end of life. Even to attempt such a radical “transvaluation of values” in the Nietzschean sense would be to perpetrate a still graver moral wrong. For patients confronting imminent death, the meaningful choice is no longer between death and continued life, but rather, as those who care for the dying know all too well, between a (relatively speaking) good death and a bad one. But to say this is to go beyond the contention of George Annas that “a painful death is a presumptively mismanaged death.”21
Annas GG. How we lie. Hastings Center Report 1995;25:S12–S14.
In the pertinent passage, Ivan Ilyich poses a fundamentally existential question of authenticity: “What if in reality my whole life has been wrong?” Tolstoy L. The Death of Ivan Ilyich and Other Stories. Edmonds R, trans. New York: Penguin; 1960:157.
Dworkin R. Life's Dominion—An Argument about Abortion, Euthanasia, and Individual Freedom. New York: Alfred A. Knopf; 1993:217.
