Least worst death: selective refusal of treatment

In recent years "right-to-die" movements have brought into the public consciousness something most physicians have long known: that in some hopeless medical conditions, heroic efforts to extend life may no longer be humane, and the physician must be prepared to allow the patient to die. Physician responses to patients' requests for "natural death" or "death with dignity" have been, in general, sensitive and compassionate. But the successes of the right-to-die movement have had a bitterly ironic result: institutional and legal protections for "natural death" have, in some cases, actually made it more painful to die.

I, ''''''"' yo." "nght-to-dio" movo­ments have brought into the public con­sciousness something most physicians have long known: that in some hopeless medical conditions, heroic efforts to ex­tend life may no longer be humane, and the physician must be prepared to allow the. patient to die. Physician responses to patients' requests for "natural death" or "<lea th with dignity" have been, in gen- M. PABST BATfIN, PH.D., is associate prOfessor oJphilosophy at the University oj Utah. Work Jor this article, conducted at the Veterans Administration Medical Cen­~ r, Salt Lake City, was sponsored by the tah Endowment Jor the Humanities, Hu­lrIanist- in-Residence Programs. The Hastings Center eral, sensitive and compassionate. But the successes of the right-to-die movement have had a bitterly ironic result: institu­tional and legal protections for "natural death" have, in some cases, actually made it more painful to die. There is just one legally protected mech­anism for achieving natural death: refusal of medical treatment. It is available to both competent and incompetent patients. In the United States, the competent patient is le­gally entitled to refuse medical treatment of any sort on any personal or religious grounds, except perhaps where the inter­ests of minor children are involy~d. A number of court cases, including Quinlan, Saikewicz, Spring, and Eichner, l have es­tablished precedent in the treatment of an incompetent patient for a proxy refusal by a family member or guardian. In addition, eleven states now have specific legislation protecting the physician from legal action for failure to render treatment when a com­petent patient has executed a directive to be followed after he is no longer compe­tent. A durable power of attorney, ex­ecuted by the competent patient in favor of a trusted relative or friend, is also used to determine treatment choices after incom­petence occurs. An Earlier but Not Easier Death In the face of irreversible, terminal ill­ness, a patient may wish to die sooner but "naturally," without artificial prolongation of any kind. By doing so, the patient may believe he is choosing a death that is, as a contributor to the New England Journal oj Medicine has put it, "comfortable,.decent, 13 and peaceful'? "natural death," the pa­tient may assume, means a death that is easier than a medically prolonged one 3 That is why he is willing to undergo death earlier and that is why, he assumes, natural death is legally protected. But the patient may conceive of "natural death" as more than pain-free; he may assume that it will allow time for reviewing life and saying farewell to family and loved ones, for last rites or final words, for passing on hopes , wisdom, confessions , and blessings to the next generation. These ideas are of course heavily stereotyped; they are the product of literary and cultural traditions associated with conventional death-bed scenes, rein­forced by movies, books , and news sto­ries, religious models, and just plain wishful thinking. Even the very tenn "nat­ural" may have stereotyped connotations for the patient: something 'close to nature, uncontrived, and appropriate. As a result of these notions, the' patient often takes "natural death" to be a painless , con­sc; ious, dignified , cu lminative slipping­away. Now consider what sorts of death actu­ally occur under the rubric of "natural death." A patient su ffers a card iac arrest and is not resuscitated . Result: sudden un­consciousness, without pain, and death within a number of seconds. Or a patient has an infection that is not treated. Result: the unrestrained multiplication of micro­organisms, the production of toxins, inter­ference with organ function, hypotension, and death. On the way there may be fever, delirium, rigor or shaking, and light­headedness; death usually takes one or two days, depending on the organism in­volved. If the kidneys fail and dialysis or transplant is not undertaken, the patient is generally more conscious, but experiences nausea, vomiting, gastrointestinal hemor­rhage (evident in vomiting blood), in­ability to concentrate, neuromuscular irri­tability or twitching, and eventually con­vulsions. Dying may take from days to weeks, unless such circumstances as high potassium levels intervene. Refusal of am­putation, although painless, is char­acterized by fever, chills, and foul­smelling tissues. Hypotension, characteris­tic of dehydration and many other states, is not painful but also not pleasant: the pa­tient cannot sit up or get out of bed, has a dry mouth and thick tongue, and may find it difficult to talk. An untreated respiratory death involves conscious air hunger. This means gasping, an increased breathing 14 rate, a panicked feeling of inability to get air in or out. Respiratory deaths may take only minutes; on the other hand, they may last for hours. If the patient refuses intra­venous fluids, he may become dehydrated. If he refuses surgery for cancer, an organ may rupture .. Refusal of treatment does ·Hot simply bring about death in a vacuum, so to speak; death always occurs from some specific cause. Many patients who are dying in these ways are either comatose or heavily se­dated. Such deaths do not allow for a pe­riod of conscious reflection at the end of life,nor do they permit farewell-saying, last rites, final words, or other features of the stereotypically " dignified" death. Even less likely to match the patient's conception of natural death are those cases in which the patient is still conscious and competent, but meets a death that is quite different than he had bargained for. Con­sider the bowel cancer patient with wide­spread metastases and a very poor prog­nosis who-perhaps partly out of consid­eration for the emotional and financial resources of his family-refuses surgery to reduce or bypass the tu mor. How, exactly, will he die? This patient is clearly within his legal rights in refusing surgery, but the physician knows what the outcome is very likely to be: obstruction of the intestinal tract will occur, the bowel wall will perfor­ate, the abdomen will become distended, there will be intractible vomiting (perhaps with a fecal character to the emesis), and the tumor will erode into adjacent areas, causing increased pain, hemorrhage, and sepsis. Narcotic sedation and companion drugs may be partially effective in control­ling pain, nausea, and vomiting, but this patient will not get the kind of death he thought he had bargained for. Yet , he was willing to shorten his life, to use the single legally protected mechanism-refusal of treatment-to achieve that "narural" death. Small wonder that many physicians' are skeptical of the "gains" made by the popular movements supporting the right to die. When the Right to Die Goes Wrong Several distinct factors contribute to the backfiring of the right-to-die cause . First , ._and perhaps the most obvious, the patient may misjudge his own siruation in refusing treatment or in executing a natural-death directive: his refusal may be pf(~cipitous and ill infonned, based more on fear than - ona settled desire to die _ Second, the-phy_ sici~~ 's response .to th~ , patient's request for death with dlgOlty may be insens'_ tive, rigid , or even punitive (though in m~ experience most physicians respond with compassion and wisdpm). Legal Con­s~ raints may also make natural death more difficult than might be hoped: safeguards often render natural-death requests and directives cumbersome to execute, and in any case, in a litigation-conscious society the physician will often take the most cau: tious route . But most important in the apparent back­firing of the right-to-die movement is the underlying ambiguity in the very concept of "natural death_ " Patients tend to think of the character of the experience they ex­pect to undergo-a death that is "comfort­able, decent, peaceful"-but all the law protects is the refusal of medical proce­dures. Even lawmakers sometimes confuse the two. The California and Kansas natu­ral- death laws claim to protect what they romantically describe as "the natural proc­ess of dying. ': North Carolina's statute says it protects the right to a "peaceful and natural" death_ But since these laws act.u­ally protect only refusal of treatment, ~. can hardly guarantee a peaceful, easy death . Thus", we see a widening gulf be­tween the intent of the law to protect the patient's final desires, and the outcomes if the law is actually followed. The physician is caught in between: he recognizes his pa­tient's right to die peacefully, naturally, 'and with whatever dignity is possible, but foresees the unfortunate results that may come about when the patient exercises this right as the law pennits. Of course, if the symptoms or pain be­come unbearable the patient may change his mind. The patient who earlier wished not to be " hooked up on rubes" now be­gins to expeJience difficulty in breathing or swallowing, and finds that a tracheotomy will relieve his distress . The bowel cancer patient experiences severe discomfort from obstruction, and gives permission for de­compression or reductive surgery after all. In some cases, the family may engineer the change of heart because they find dying too hard to watch. Health care personnel may view these reversals with satisfaction: "See," they may say, "he really wants to live after all ." But such reversals cannol.. always be interpreted as a triumph of the will to live; they may also be an indication that refusing treatment makes dying tOO hard. The Hastings Center Report, April 1983 - Options for an Easier Death How can the physician honor the dying patient's wish for a peaceful, conscious, and culminative death? There is more than one option . Such a death can come about whenever the patient is conscious and pain-free; he can reflect and, if family, clergy, or friends are summoned at the time, he will be able to communicate as he wishes. Given these conditions, death can be brought on in var­ious direct ways. For instance, the physi­cian can administer a lethal quantity of an appropriate drug. Or the patient on severe dietary restrictions can violate his diet: the kidney-failure patient , for instance, for whom high potassium levels are fatal, can simply overeat on avocados. These ways of producing death are, of course, active euthanasia , or assisted or unassisted sui­cide. For many patients, such a death would count as "natural" and would sat­isfy the expectations under which they had chosen to die rather than to continue an in­tolerable existence. But for many patients (and for many physicians as well) a death that involves deliberate killing is morally wrong . Such a patient could never assent to an actively caused death, and even though it might be physically calm, it could hardly be emotionally or psychologi­cally peaceful. This-is not to say that active euthanasia or assisted suicide are morally wrong, but rather that the force of some patients' moral views about them pre­cludes using such practices to achieve the kind of death they want. Furthennore, many physicians are unwilling to shoulder the legal risk such practices may seem to involve. . But active killing aside, the physician can do much to grant the dying patient the humane death he has chosen by using the sole legally protected mechanism that safe­guards the right to die: refusal of treat­ment. This mechanism need not always backfire. For in almost any tenninal condi­tion, death can occur in various ways, and there are many possible outcomes of the patient's present condition . The patient Who is dying of emphysema could die of respiratory failure, but could also die of cardiac arrest or untreated pulmonary in­fection. The patient who is suffering from bowel cancer could die of peritonitis fol­lOwing rupture of the bowel, but could also die of dehydration, of pulmonary infec­tion, of acid-base imbalance, of electrolyte deficiency, or of an arrhythmia . The Hastings Center Death, Too Soon or Too Late A death can occur too soon or 100 late. Further, there is the question of how it will come about. There are dyings that areslow ariq agonizing ahd dyings that are gentle and graceful. It's tbe latter sort to which one aspires, but there is a substan­tial risk of being overtaken by froward circumstances. It seems reasonabJe that if death is impending, one should want to exert some influence on the mode of one's dying. Just as one wants to be able to influence the major e,;,etlts that shape and constitute a life at earlier stages, one may want to avoid the indignity of having to witness and endure a final stage not as an effective agent, but merely a deteriorating object. As the poet Rilke observes, we have a. tendency to associate a certain sort of end with a specific disease: it is the "official death" for that sort of illness. But there are many other ways of dying than the official death, and the physician can take advan­tage of these . Infection and cancer, for in­stance , are old friends ; there is increased frequency of infection in the immuno­compromised host. Other secondary con­ditions, like dehydration or metabolic derangement, may set in. Of course certain conditions typically occur a little earlier, others a little later, in the ordinary course of a tenninal disease, and some are a mat­ter of chance. The crucial point is that cer­tain conditions will produce a death that is more comfortable, more decent, more pre­dictable, and more permitting of conscious and peaceful experience than others. Some are better, if the patient has to die at all, and some are worse. Which mode of death claims the patient depends in part on cir­cumstance and in part on the physician's response to conditions that occur. What the patient who rejects active euthanasia or assisted suicide may realistically hope for is this: the least worst death among those that could naturally occur. Not all un­avoidable surrenders need involve rout; in the face of inevitable death, the physician becomes strategist, the deviser of plans for how to meet death most favorably. He does so, of course, at the request of the patient, or, if the patient is not compe­tent, the patient's guardian or kin. Patient autonomy is crucial in the notion of natural death . The physician could of course pro­duce deat~ by simply failing to offer a par- Samuel Gorovitz, Doctors' Dilemma: Moral Conflict and Medical Care, Macmillan, 1982, p. 153. ticular treatment to the patient. But to fail to offer treatment that might prolong life, at least when this does not compromise limited or very expensive resources to which other patients have claims, would violate the most fundamental principles of medical practice; some patients do not want " natural death," regardless of the physical suffering or dependency that pro­longation of life may entail. A scenario in which natural death is ac­complished by the patient's selective re­fusal of treatment has one major advantage over active euthanasia and assisted suicide: refusal of treatment is clearly pennilled and protected by law. Unfortunately, how­ever, most patients do not have the special­ized medical knowledge to use this self­protective mechanism intelligently. Few are aware that some kinds of refusal of treatment will beller serve their desires for a "natural death" than others. And few pa­tients realize that refusal of treatment can be selective. Although many patients with life-threatening illness are receiving mul­tiple kinds of therapy, from surgery to nu­tritional support, most assume that it is only the major procedures (like surgery) that can be refused . (This misconception is perhaps perpetuated by the standard prac­tice of obtaining specific consent for major procedures, like surgery, but not for minor, ongoing ones.) Then, too, patients may be unable to distinguish therapeutic from palliative procedures. And they may not understand the interaction between one therapy and another. In short, most pa­tients do not have enough medical knowl­edge to foresee the consequences of refusing treatment on a selective basis; it is this that the physician must supply. It is already morally and legally recog­nized that informed consent to a procedure involves explicit disclosure, both about the risks and outcomes of the proposed proce­dure and about the risks and outcomes of aJternative possible procedures. Some courts, as in Quackenbush ,4 have also rec­ognized the patient's right to explicit dis­closure about the outcomes of refusing the proposed treatment. But though it is cru­cial in making a genuinely informed deci­sion , the patient's right to information about the risks and outcomes of alternative kinds of refusal has not yet been recog­nized. So, for instance, in order to make a genuinely informed choice , the bowel can­cer patient with concomitant infection will need to know about the outcomes of each of the principal options: accepting both bowel surgery and antibiotics; accepting antibiotics but not surgery; accepting sur­gery but not antibiotics; or accepting nei­ther. The case may of course be more complex , but the principle remains: To rec­ognize the patient's right to autonomous choice in matters concerning the treatment of his own body , the physician must pro­vide information about all the legal options open to him, not just information sufficient to choose between accepting or rejecting a single proposed procedure. One caveat: It sometimes occurs that physicians disclose the dismal probable consequences of refusing treatment in order to coerce patients into accepting the treatment they propose . This may be par­ticularly common in surgery that will result in ostomy of the bowel. The patient is given a graphic description of the impend­ing abdominal catastrophe- impaction, rupture, distention, hemorrhage, sepsis, and death . He thus consents readily to the surgery proposed . The paternalistic physi­cian may find this maneuver appropriate, particularly since ostomy surgery is often refused out of vanity , depression, or on fa­talistic grounds . But the physician who frightens a patient into accepting a proce. dure by describing the awful consequences of refusal is not honoring the patient's right to informed, autonomous choice: he has not described the various choices the pa­tient could make, but only the worst. Supplying the knowledge a patient needs in order to choose the least worst death need not require enormous amounts of ad­ditional energy or time on the part of the physician; it can be incorporated into the 16 usual informed consent disclosures. If the patient is unable to accommodate the med­ical details, or instructs the physician to do what he thinks is best, the physician may use his own judgment in ordering and re­fraining from ordering treatment. If the pa­tient clearly prefers to accept less life in hopes of an easy death, the physician should act in a way that will allow the least worst death to occur. In principle, how­ever, the competent patient, and the proxy deciders for an incompetent patient, are entitled to explicit disclosure about all the alternatives for medical care. Physicians in burn units are already experienced in tell­ing patients with very severe burns , where survival is unprecedented, what the out­come is likely to be if aggressive treatment is undertaken or if it is not~eath in both cases, but under quite different conditions . Their expertise in these delicate matters might be most useful here. Informed re­fusal is just as much the patient's right as informed consent. The role of the physician as strategist of naturar death may be even more crucial in longer-term degenerative illnesses, where both physician and patient have far more advance warning that the patient's condi­tion will deteriorate, and far more opportu-, nity to work together in determining the conditions of the ultimate death. Of course, the first interest of both physician and patient will be strategies for maximiz­ing the good life left. Nevertheless, many patients with long-term, eventually termi­nal illnesses, like multiple sclerosis, Hunt­ington's chorea, diabetes, or chronic renal failure, may educate themselves considera­bly about the expected courses of their ill­nesses, and may display a good deal of anxiety about the end stages. This is par­ticularly true in hereditary conditions where the patient may have watched a par­ent or relative die of the disease. But it is precisely in these conditions that the physi­cian's opportunity may be greatest for hu­mane guidance in the unavoidable matter of dying. He can help the patient to under­stand what the long-term options are in re­fusing treatment while he is competent , or help him to execute a natural-death direc­tive or durable power of attorney that spells out the particulars of treatment re­fusal after he becomes Incompetent. Of course, some diseases are complex, and not easy to explain. Patients are not always capable of listening very well," es­pecially to unattractive possibilities con­cerning their own ends . And physicians - are sometimes reluctant to acknowledge that their efforts to sustain life will even_ tuaJly fail. Providing such information may also seem to undermine whatever hope the physician can nourish in the pa­tient. But the very fact that the patient's demise is still far in the future makes it possible for the physician to describe vari­ous scenarios of how that death could oc­cur, and at the same time give the patient control over which of them wiU actually happen. Not aU patients will choose the same strategies of ending, nor is there any reason that they should . What may COunt as the "least worst" death to one person may be the most feared form of death to another. The physician may be able to in­crease the patient's psychologicaJ comfort immensely by giving him a way of meet­ing an unavoidable death on his Own terms. In both acute and long-term terminal ill­nesses, the key to good strategy is flexibil­ity in considering all the possibilities at hand . These alternatives need not include active euthanasia or suicide measures of any kind, direct or indirect. To take advan­tage of the best of the naturally occurring alternatives is not to cause the patient's death, which will happen anyway , but to guide him away from the usual, frequently worst, end. In the current enthusiasm for " natural death" it is not patient autonomy that dis­mays physicians. What does dismay them is the way in which respect for patient au- j lonomy can lead to cruel results. The cure for that dismay lies in the real ization that the physician can contribute to the genuine honoring of the patient's autonomy and rights, assuring him of "natural death" in the way in which the patient understands it, and still remain within the confines of good medical practice and the law. REFERENCES 'In re Quinlan, 355 A. 2d 647 (N.l. 1976); Superillfendellf of Belchertown v. Saikewicz, 370 N.E. 2d 417 (Mass. 1977); In re Spring, Mass. App ., 399 N.E. 2d 493; In re Eichner, 73 A.D. 2d 431 (2nd Dept. 1980) , 2S.S. Spencer, '''Code' or 'No Code: A Nonlegal Opinion," New England Journal oj Medicine 300 (1979) , 138-140. . 3See Dallas M. High's analysis of the vanous senses of the term " natural death" in ordmary language , in " [s ' Natural Death' an Illusion?", Hastings Cellfer Report , August 1978, pp. 37-42. 4/11 re Quackenbush , 156 N.J. Super. 282, 353 A. 2d 785 (1978). The Hastings Center Report, April 1983