The President's Council on Bioethics click here to skip navigation



Table of Contents

The President's Council on Bioethics
Washington, D.C.
September 2005

Ethical Caregiving: Principle and Prudence
in Hard Cases
Chapter 4

In this chapter, we move from general considerations and basic principles to consider ethical caregiving as it is actually practiced: in concrete human situations, with real patients, concerned family members, and an array of medical professionals intimately involved in providing care. Our goal in the following case discussions is to explore how devoted caregivers should approach treatment decisions in a range of complex clinical situations-always aiming at best care for the person now here, and always working within the ethical boundaries articulated above.

Every clinical case is unique, with innumerable medical and personal factors that need to be considered-factors that are always changing. With dementia, caregivers are rarely handling a single medical condition or personal problem; they are usually juggling multiple issues at once-ranging from ongoing ailments to worsening risk factors to acute diseases, and ranging from family struggles with day-to-day care to worries about finances to caregiver exhaustion and depression. In specific clinical cases, caregivers often need to make decisions based largely on probabilities-the probability that a given disease will progress without treatment, the probability that a treatment will work to the desired effect, the probability of complications and side effects, or the probability that the person's dementia will worsen quickly or slowly.

But ethically responsible caregivers also operate within certain firm moral boundaries-central among them the obligation never to seek a patient's death in making decisions about a person's care. This complexity requires us to think through many larger ethical issues in the context of particular cases-where the aim is always to care and never to kill, always to benefit the life the patient still has, even as the person's powers decline, dependence becomes total, life options are limited, and death looms ever closer.

In presenting these cases, we have three interrelated purposes. First, we wish to demonstrate by example what it means to think seriously about "best care," served by loving prudence, beyond the procedural solutions of "let each family decide" or "consult the advance directive." Second, we seek to show the great complexity and moral difficulty often involved in deciding how to provide best care, precisely because the best course of action is frequently far from obvious. And third, we aim to show how, even in hard cases where there is no obvious good choice, prudence can probe deeply how best to care for the particular individuals for whom we are responsible, all the while respecting those ethical boundaries that should be respected by everyone. At least as important as the "bottom line" in each case is the character of moral reflection and the manner of moral reasoning required to judge and act well.

In discussing cases of ethical caregiving, it is important to remember that we are addressing not only the professionals in hospitals, nursing homes, and hospices. We are also addressing audiences both more private and more public: family members and friends who participate in making decisions for loved ones; public policymakers, directors of caregiving facilities, and others who form the ideas and set the practices that shape how the more immediate caregivers think and act when called upon to do so. It is also necessary to remember that the "hands-on" caregivers, the principal actors in our caregiving cases, are not professional ethicists nor generally given to long discourses about the whys and wherefores of their recommendations and decisions. As human beings and members of our society, they have moral ideas and intuitions that guide them at the bedside; as professional caregivers, they have certain standards of care that they pledge to follow; as clinicians, they wrestle concretely-not at the level of abstract moral principle-with hard choices when values they hold dear are in tension. At the same time, however, it is also true that patterns of practice call for justification, especially when they are challenged by appeals to novel principles or demands for new procedures. Challenged or not, these patterns of professional care come to shape and reflect broader public judgments about how to think and what to do when the thinking and doing become very difficult. We trust that practitioners reading these case discussions will see the merit in trying to think things through and to offer discussable reasons for whatever course of action is under consideration. In matters this weighty, we need to generalize our experiences, in search for wisdom in the face of often tragic choices.

We offer two additional caveats: First, one central idea of this report is that caregivers in particular and society as a whole must never lose sight of the humanity of those persons with dementia entrusted to their care-which is to say, the unique histories and life stories that define the richness of the life now moving toward completion. Yet the following cases, presented in ways to permit exploration of moral problems that caregivers and patients may experience, risk losing the fine texture of the persons and lives they so minimally describe. We are aware that such an approach risks encouraging the very dehumanization of persons with dementia that we are in this report trying to combat, and we urge the reader to lean against any such apparent intimations or suggestions.

Second, these cases leave many precise clinical details undescribed: just as every life is unique, so is every medical situation. In trying to be somewhat paradigmatic, we risk losing or deforming the realities of the bedside, and risk leaving unexplored the clinical details that are often central in deciding what the best possible course of action truly is. With these caveats in mind, we nevertheless hope these case discussions might enrich our thinking about what "best care" really means in practice and how principle and prudence can operate together to do right by those for whom we are summoned to care.

We organize the discussion that follows thematically, by considering cases that highlight the following issues: (1) the distinction between deciding for oneself and deciding for others; (2) the significance of dementia and its stages; (3) the meaning of well-being for persons with dementia; (4) the significance of living wills; and (5) the well-being of caregivers themselves.

I. Deciding for Oneself AND Deciding for Others

We begin by exploring the moral differences (and similarities) between deciding for oneself in a still competent state and deciding for another who has become mentally incapacitated. Consider, in this vein, the following case:

A married man, in his sixties, has two adult children who live far away. He has cared for two parents who have suffered and eventually died from Alzheimer’s disease, and he begins to notice his own lapses of memory: forgetting to turn off the stove, losing track of the time, forgetting the names of colleagues, repeating himself in conversations. He suspects (and his doctor agrees) that he is probably in the very early stages of Alz-heimer’s disease, and he surmises from personal experience where he is heading and what his life will likely become. He knows he would prefer a shorter and still lucid life to a slow decline into total dependency—with the accompanying erosion of self-control and loss of shame, and with the possibility of behaving in sexually inappropriate ways as his father did or becoming incontinent and immobile as his mother did. For years, he has also been treated for high cholesterol and high blood pressure, and his doctors tell him that the blockage in his left-main coronary artery is se-vere enough that he is a candidate for bypass surgery. As his cognition begins to weaken, he is still aware of his overall condition. He decides to stop taking all his heart medication, and he makes clear to his doctor and family: no surgery—ever. How should his family and his doctor respond?

Whatever judgments we might make about the moral propriety of the individual's decision to forgo all treatment for heart disease, it would be a different matter if his family or doctor sought to deny him future care because of his probable diagnosis of early-stage Alzheimer's. Perhaps his reasons for declining treatment are sound; perhaps they are not; perhaps we can never know for certain. We will shortly explore the possible reasons why he does what he does, and consider the possible moral meanings of his choices. But clearly it would be morally misguided for others to deny treatment if the patient himself wanted it, and not only because of the coercion involved. In making such a decision, the individual's family and doctors would be sending the clear message that they would rather see him dead than suffer advancing dementia; that his coming life will be an unjustified burden to them and to society; and that he lacks all basic claims on care because of his probable diagnosis and looming decline. Such a denial of care is not only coercive, it is dehumanizing.

Denial of care, of course, represents an extreme. Further analysis would be required to show why it would be impermissible for family members or doctors to suggest to the patient, had he not thought of doing so himself, that he could, if he wished, stop taking his heart medications as a form of "Alzheimer's prevention." But neither denial nor discouragement of treatment is the issue here, but rather acquiescence: should the doctor and family endorse or accept the patient's decision to stop his medications and to refuse all further treatment of his cardiac condition?

In reality, of course, such a thought would not likely occur to them. In most such cases, the patient's family members and doctors would probably resist vigorously his decision to forgo all future treatment. They would urge him to continue his medications and perhaps even to undergo bypass surgery. Even the loving wife who admires her husband's nobility and self-command, or the family physician who knows what the individual went through in caring for his own parents, would likely balk at his seeking or inviting an avoidable and earlier death. But are there good reasons for the patient to resist their resistance, or for the individual to resist becoming a "patient" at all? What considerations determine whether his decision to forgo treatment is morally acceptable or morally sound? And should the range of morally acceptable choices be wider when the person decides to forgo treatment for himself, as opposed to when others decide to withhold treatment from and for him?

The most obvious difference concerns the absence or presence of direct coercion: The patient's decision is his own; it is not imposed on him against his will or without his full participation. The patient also makes what appears to be an informed decision, drawing upon his direct experience caring for his parents with Alzheimer's and extrapolating forward to what it might be like to live in their condition and to need the kind of care he once provided them. Not only is he not coerced, he knows his own inner life in a way even those closest to him arguably cannot. He knows what he values, what he thinks he can endure, and what he hopes the final chapters of his life story will look like.

It is, of course, possible that he has not reasoned this out, that he is reacting emotionally to the bad news, or that he is even, without knowing it, clinically depressed. But insofar as he has thought about his decision, any number of considerations might (alone or in concert) seem to him to warrant doing whatever he can to decrease the chances that he will have to endure the full course of Alzheimer's disease. Perhaps he knows that he does not want to die incapable of recognizing his wife in his final months, or that he does not want to die after years in a nursing home. One can also imagine the range of goods that his decision to forgo treatment hopes to achieve: Positive goods, such as leaving an education fund for his grandchildren rather than spending down all his assets on nursing home care, and negative goods, such as ensuring that he never behaves in ways that erode the code of honor and decency by which he has always lived. People with Alzheimer's sometimes do things that are normally considered shameful and offensive-such as making sexually inappropriate advances or undressing in public. Of course, the disease is the cause of the loss of inhibitions. Yet it is, in a crucial sense, still the person who acts, and seeking to avoid or preempt such behavior is a morally worthy goal, even if a particular means of doing so is morally questionable.

But the seemingly clear differences between deciding for oneself and deciding for others need to be examined and questioned. To what extent, we must ask, are such momentous decisions ever fully "one's own"? As Robert Burt explored in Taking Care of Strangers, people often make the choices that they believe others want for them.1 Perhaps the patient imagines that his wife and children do not want to care for him, and perhaps his decision to forgo treatment reflects his assessment of their hidden wishes more than his own desires. Perhaps he does not trust his doctor to be able to guide him through the illness. Rather than confirming these suspicions by acquiescing in his decision, his family, friends, and physician should explore these matters with him, explicitly and thoroughly.

In addition, they should ask him to consider how his decision might affect the web of relations he most values. Has he considered that his choice to forgo treatment might send the message that he does not really trust his family members to be devoted caregivers or that he does not believe they are capable of caring for him? And has he given adequate thought to the potential moment of crisis-the sudden heart attack-that his decision to forgo treatment makes more likely and perhaps deliberately invites? Does he expect his loved ones or neighbors simply to watch him die rather than call the paramedics? Has he considered what it will be like for his family to suffer the ordeal of rushing him to the hospital, hoping to save him, living through the experience of triage, fearing and perhaps losing the possibility of saying goodbye? If he survives the heart attack, he might end up in a severely debilitated condition, or else die a lingering death from chronic heart failure, with shortness of breath and swelling of the extremities, rather than dying suddenly as he might have wished. If he does indeed die suddenly, has he considered how his choice to forgo preventive treatment might affect the way his wife and children will remember him?

In seeking to avoid the worsening degradations of Alzheimer's disease, he probably worries deeply about being remembered in a distorted way, not as the person he has long been but as the person whose self is hollowed out by dementia. But what about the potential distortion of being remembered as someone who sought an earlier death by deliberate acts of omission, or someone whose actions helped precipitate a death in crisis? Would this be seen as a sign of his noble willingness to accept an earlier death to prevent becoming wholly dependent on others, or as a mark of his timidity in the face of a difficult death by dementia?

In reality, individuals, though free to choose, never simply decide solely by and for themselves, even when they are fully competent. The present patient's decisions reflect, in part, his image of himself in the eyes of others, and his decisions affect the course of their lives. More generally, our identity and values always take shape within a network of human relations. The individual may know himself from the inside, but his inside knowledge is shaped by his understanding of how others understand him. And his inside knowledge, never final or definitive, is always open to transformations of self-understanding in light of new circumstances, at least as long as he is still self-aware. A discussion with those affected by his decision or with his trusted physician could very well lead to such a transformation of outlook, and they should certainly attempt it.

Beyond doubt, the competent individual should be accorded greater freedom of choice in making treatment decisions than should surrogates deciding for others. It is always possible that the individual knows aspects of his own life story that no one else can know, and with this privileged perspective his wishes deserve heightened respect. And surely, competent individuals should never be coerced to accept treatments they do not want, even after attempts at thoughtful persuasion have failed.

But the competent individual's greater freedom to decide does not mean that every permissible decision is morally sound, or that his loved ones should simply accept his choices as necessarily right or necessarily final. Moreover, this freedom of action operates within certain widely accepted moral limits: even if the law allows it, an individual is not morally justified to act deliberately to achieve his self-destruction. And although the individual deciding for himself rightly exercises greater freedom than surrogates deciding for others, the moral criteria for deciding for oneself and deciding for others are actually quite similar. In both cases, what matters most is what we actually do in service to our goals and the reasons for seeking or forgoing particular treatments in particular circumstances. And in both situations, the person or the surrogate is making a decision about whether the life in question is worthy of continued care and about whether the life in question requires protracted preservation.

Let us return to the case itself, and what the doctor might say to the individual in this situation: "Yes, you are probably in the early stages of Alzheimer's. But the trajectory of that disease is long and gradual, you have much good life left to enjoy, and you can count on me to help you negotiate your way. I will be with you, through good times and bad. But that it is not now my most pressing concern. Your heart disease is getting worse. You need to continue your daily medications. You need to begin a stricter heart-healthy diet. You need to exercise more intensively. And you are a candidate for immediate bypass surgery, and I recommend that you have it very soon."

In speaking with the patient, doctors and family members must take seriously the moral reasons why the individual might seek a course of life that, although shorter than it need be, does not end in dementia. Perhaps he believes that suffering a heart attack is a more humanly fitting way to die-facing death with the powers of memory and self-awareness still largely intact, without the deformations that long-term care can sometimes impose on families. And perhaps he concludes that he has no moral obligation to accept medical care that seems likely to rob him of this more fitting death and condemn him instead to what he sees as a dehumanizing death by way of dementia. Why must he cede control of his destiny to the medical inventions and interventions of others? Why must he visit or call the cardiologist in the first place?

Yet there are sound moral answers to these questions. We are ethically free to accept a higher risk of death or expose ourselves to a foreseeable death in the course of choosing a particular kind of life-such as a soldier serving his country on the frontlines or a policemen working in a gang-infested neighborhood, where death by gunfire may be the likely cost of bravery, or a physician tending patients suffering from plague or performing surgery on patients with AIDS, where death by contagion may be the possible cost of compassionate care. But it is hard to see, in this case, how ceasing heart medication is essential for living a particular kind of life, a life in which being un-medicated is the necessary cost of living well in the present. It is hard to see the moral grounds for stopping such medication now (that is, upon learning the diagnosis of probable Alzheimer's)-assuming the drugs have no burdensome side effects, and assuming the economic burden of the drugs is minor.

The most likely reason for the patient to stop treatment now is to increase his chances of dying earlier, or of dying this way (heart disease) rather than that way (dementia). Taking heart medication does not preclude living now in the way he wants to live; the drugs are not a barrier to present goods or a direct cause of present distress. And unless an individual has an established pattern of resistance to taking such drugs-for example, for religious reasons, akin to those who reject blood transfusions-there seems to be no decisive moral difference between stopping medications in-process and deciding not to start medications that are newly indicated: in both cases, one forgoes treatment with an earlier death as the goal, or with an earlier death as the potential means of achieving other goals (such as never burdening loved ones or never becoming incontinent). Even if the individual rightly possesses the freedom of action to make such choices-that is, to reject needed medications-his loved ones are morally right to resist.i

It is, of course, possible that the person sees his choice not as expressing a desire to die sooner but as a desire to live out the remainder of his life in a certain way. He might even think or say something like this: "I am a person coming to the end of my life and its projects. I trust in God and I am reconciled to my fate. My obligations to my body are present but limited now. It is okay for me to decide about medical treatments in a way that makes it more likely that I'll live out my natural life in the family and home I love, aware of salvation and relationships. It is acceptable to all to make such choices. I am not a prisoner of medical advances that would condemn me to a more troubled course."ii

This is a more difficult argument to answer. Yet even here, the patient's family and his physician could rightly argue back, making the case for the goodness of the life still at hand, reminding him of the burdens that his choice might be inflicting on those who love him, and reassuring the patient that a commitment to treatment now need not mean that he will in the future become a prisoner of unwanted medical innovations when he comes closer to his death. Persuading the patient to change his mind about taking his medications still seems to be the course of loving prudence.

The duty to exercise more frequently or to eat more restrictively is a different matter. If life's primary goal were merely optimizing health and longevity, then good diet and regular exercise would be one's primary obligations. But we seek to be healthy in order to live well, not the other way around, and many of the things we rightly enjoy in life are-to say the least-not always conducive to maximizing longevity. It would be entirely understandable if our patient-mindful of his impending descent into dementia-would like to relax and enjoy life in his last years of lucidity more than he could under the heart-healthy regimen. Perhaps he wishes to spend more time reading and less time on the treadmill, or wants to enjoy large pasta dinners instead of dressing-free salads, or wants to continue his favorite pastime of shoveling snow even though his doctors recommend against it. In choosing to forgo the ideal regimen of diet and exercise, the individual is choosing one kind of life over another. He is choosing some goods (the pleasures of eating) over others (the greater chance of longevity). He is not simply acting so that he might die sooner, even if his actions make it more likely that he will.

The patient's choice to forgo bypass surgery also seems morally defensible, if not the only morally permissible choice under the circumstances. Bypass surgery is an experiential and physical ordeal-with serious risks, grave burdens, and various uncertainties. It disrupts personal life and family life. It requires hospitalization and a considerable period of recovery, in which being a patient is the core of one's existence for weeks and often months. It even carries a high risk of yielding a major drop in cognitive function. Although statistically it tends to extend healthy life, there are good reasons to opt against it, just as there are good reasons to opt for it. Another person, in the same situation, might rightly decide to have the surgery as soon as possible, so that he can recover and improve physically while his mental faculties are still relatively sound and the effects of Alzheimer's are still fairly limited. But the individual in this case is on solid moral ground in rejecting bypass surgery even if his doctors recommend it, and even if his family encourages him to have it.

Let us consider a slightly different case: What if, instead of being a person with heart disease, the individual was a (long-lived) type-1 diabetic, absolutely dependent on regular insulin shots? And what if, upon learning of his probable Alzheimer's, he decided to spend a few wonderful weeks with his family, say goodbye, and stop taking his shots, so that he lapses into a diabetic coma resulting in his imminent death. Unlike forgoing heart medication or rejecting heart surgery, this decision makes very likely a certain kind of death rather than simply altering the probable course of death. It ensures death now rather than awaiting a still unknowable death in the future. Although taking insulin shots is a marginally greater burden than taking heart medication, this marginal burden hardly justifies the decision to embrace death now by a deliberate act of omission. In this case, the individual's moral freedom should give way to the moral limits that rightly constrain that freedom. And although the competent individual cannot be forced to take insulin against his will, his family and doctors are justified in exerting every decent means to alter or reverse his death-seeking decision.

Consider yet another case: What if the patient with probable early-stage Alzheimer's gets diagnosed with a form of cancer that is treatable with a reasonable probability of cure but deadly if left untreated. The treatment of chemotherapy followed by radiation would require a disruption of his current life-regular trips to the hospital over an hour away, considerable discomfort and nausea, general fatigue-but not unendurable physical suffering. Untreated, the progression of the cancer would leave the patient largely unaffected for a while, with things getting worse toward the end, with a relatively swift decline and death in an estimated nine to eighteen months in the future. In this case, the individual sees two possible paths before him: the first path involves disrupting his current routine, possibly curing his cancer, and extending a life likely marked by worsening Alzheimer's; the second path involves enjoying the life he now has without disruption or burdensome treatments, living with a worsening cancer that will eventually kill him, and probably dying before his Alzheimer's hollows out all recognition of himself and his loved ones.

Unlike forgoing heart medications, a choice to forgo cancer treatment involves avoiding a burdensome treatment. And it involves accepting a relatively predictable progression toward death, rather than increasing the chances of having a deadly heart episode and thus inviting a future moment of crisis, for which his loved ones cannot really prepare. And unlike type-1 diabetes, an ailment that the hypothetical patient above had been treating throughout his life, with treatment already a part of his routine and not an impediment to it, the cancer diagnosis is new and the treatment will be a major disruption. In the end, this is a case where each individual should rightly decide for himself what is the best course-doing his best to make such decisions in concert with loved ones, and resisting the temptation to see cancer as simply a means to avoid the later ravages of advanced dementia. The decision to forgo cancer treatment in this case, if rightly made, will aim at a certain kind of life until the end, not at a certain kind of death sooner rather than later.

Finally, how would this ethical analysis change if the individual's dementia progressed to the point where surrogates had to make all medical decisions on his behalf? And what weight should be given to the preferences he expressed while he was still competent? The latter question we will take up more fully in the case below on advance directives. But in general, surrogates should make decisions that aim at best care for the person now here: They would have no clearly compelling reasons to withhold his heart medication, assuming there are no intolerable side effects and assuming the cost of the medications is not excessively burdensome to the family. They might even be morally justified in resuming the medications that the once-competent patient had previously chosen to stop. They should encourage those foods and activities that seem to promote and augment their loved one's happiness. They should almost certainly forgo heart bypass surgery, rejecting it as unduly burdensome, especially when the patient cannot be a full participant in his own treatment and recovery, and because by then his physical capacity to endure such a procedure has probably weakened along with his dementia. And they should surely continue insulin shots if the patient were to suffer from diabetes. To do otherwise is to seize an occasion for death, and thus to treat the life being cared for as unworthy of continued care. But they should not begin aggressive cancer treatments that the person himself knowingly chose to forgo.

II. The Significance of Age-Related Dementia and Its Stages

As we move now to consider the dilemmas of caring for those with dementia who can no longer direct their own care, we must first inquire more precisely about the significance of age-related dementia itself, and whether this particular condition differs from other forms of disability and dependence. That is to say: Is the presence of dementia a morally legitimate factor in denying or forgoing treatment? Is the person with dementia morally different from-and should she be treated differently than-the lucid elderly person with severe physical debilities, or the mentally incompetent younger person who did not lose his mental powers with age but never had them at all?

There is, of course, a difference between the never-competent person and the once-competent person, and this age-related transformation seems to matter, humanly speaking. The person who late in life develops dementia is no longer the wise father but the needy one; he is no longer the husband who remembers every wedding anniversary but the man who cannot recognize his wife; he is no longer the grandfather who takes the young to the zoo but the man who acts strangely in the nursing home. The question, however, is what these changes mean, if anything, for how we regard our loved ones as patients, and whether their psychic loss is a disability that makes their dependence morally different in kind.

In the case of total physical dependence, the person is no doubt also transformed, and his or her physical neediness may bring about significant psychic change. But even in a state of total dependence, such a person might still direct his or her own care, or at least be involved in many caregiving decisions. Moreover, the dependent elderly person who is still cognitively self-aware may still acknowledge his special attachment to the loved ones who now care for him-perhaps with a mix of gratitude for being helped, sorrow at lost pastimes, and shame at being so needy.

With advanced dementia, however, the situation is somewhat different. The beloved person is still present-still the same person in body, with a continuous life trajectory. But he is also, in a certain sense, increasingly absent. He is both always himself and no longer himself. And although the same thing might be said regarding all forms of decline and dependence-the great dancer in a wheelchair is also no longer as she once was, certainly not in the way that once defined her-cognitive loss seems to be existentially different. The great dancer in a wheelchair still knows she was once a great dancer (though whether such knowledge makes life better or worse, easier or harder, is not a simple question). The great mathematician, now demented, knows nothing of his former life.

In clinical situations, these puzzling theoretical questions of identity and personal continuity give way to urgent practical questions of how vigorously to treat the supervening or concurrent illnesses that the patient with dementia also suffers. There is no question but that physicians tend to be more aggressive in treating serious illnesses in thirty year olds than in eighty year olds, partly because younger patients generally tolerate treatments better, but also because everyone tacitly recognizes that the old patient is closer to the end and that, beyond good standard medical care, certain treatments that are probably regarded as obligatory in the young ("ordinary care") will probably be regarded as optional in the old ("extraordinary care"). But within this general approach to aggressive treatment at various stages of life and health, the relevant question before us now is whether the special changes in mentation and self-awareness that come with dementia should affect decisions about medical care. Does a person with dementia have a lesser claim on or lesser interest in essential medical treatment solely by virtue of his dementia?iii Is he less a "person," and thus more ready for life to be over? And if the goal is providing the "best care possible" for the person with dementia, how does dementia itself influence or shape what best care requires?

The general answer seems to be two-fold: As argued above, the fundamental equality of human beings suggests that human worth, in the caregiving setting, does not depend on possessing particular capacities or particular qualities. To discriminate against patients because they suffer dementia is to violate this first principle. At the same time, as dementia advances, the systems of the body begin to shut down; and different choices might be made as the patient gets closer to death, even when death is not yet imminent.iv

To consider these questions in concrete cases, we need to make three crucial distinctions: First, we need to distinguish between the different stages of dementia, and to ask whether a person's level of mental decline is in itself a relevant factor for shaping treatment decisions. Second, we need to distinguish between "good days" and "bad days," and pay close attention to the fluctuating cognitive capacities exhibited by many persons with dementia. And third, we need to distinguish between conditions and complications: Dementia is a condition that might be used (wrongfully) in itself as a moral ground to treat patients differently; but dementia is also a source of ethically relevant special complications that must be faced and evaluated by caregivers in the clinical setting, especially when dementia patients cannot understand the value or necessity of a given treatment.

To see how dementia might affect the decisionmaking of surrogates, let us consider how we should care for the following person, suffering from bacterial pneumonia, at five different stages of her life:

Are there any moral grounds for treating the person's bacterial pneumonia differently at these different points in her life, and if so, why? Let us assume she has no living will, and that she has always trusted her daughters to take good care of her as she grows old. The decisions (in her later years) regarding medical treatment thus rest entirely on the daughters' judgment about best care for their elderly mother, a judgment that will depend in part on the moral meaning they ascribe to her worsening dementia and her increasing nearness to death.

In stage 1, the girl's parents are clearly deciding on her behalf, and they have a clear moral obligation to treat her pneumonia. In stage 2, the woman is deciding for herself, and one can see no obvious reason to forgo treatment (if the patient is not opposed in principle to all medical interventions, as are adherents of Christian Science). In stage 3, the woman is still able to understand the need for treatment once it is explained to her, but she probably needs assistance getting to the doctor, following through on her medications, and navigating among different health care providers. She is no longer the sole director of her own care, but she is still a full and active participant in her own treatment. She may be frustrated by her fading powers and growing dependence on others, but she still values life, and the good of treatment seems clear.

By stage 4, all her decisions need to be made for her, and explanations of her sickness and treatment are beyond her comprehension. She is a needy recipient of care, and the sufferings of her body may prompt physical expressions of dismay akin to a child's cry. But she is no longer a knowing seeker of care, and those who actually care for her-daughters, doctors, and nurses-are indistinguishable to her. Yet in the case as described, her health is not in rapid decline. Hence, the only apparent grounds to deny her treatment for her pneumonia would be that life without memory of the past or control of the future is simply not life worth living or that there are even worse futures coming, as indeed there are. In other words, dementia as a degenerative and debilitating condition would be the reason to stand aside in the face of a disease that is easily treated but potentially deadly if not treated. To say no to treatment in this case is to declare that death is preferable to life with this degree of dementia (or worse), or that dementia means being a "non-person" ready to die. But does saying yes to treatment depend entirely on the fact that she still seems to get some experiential satisfactions from being alive, even if she does not experience the present in connection to her personal past? In other words: Does the worth of being alive depend decisively on the enjoyment one takes in being alive?

This becomes an issue in stage 5, when even the limited experiential pleasures or experiential interests of her earlier dementia seem to be fading or gone. The caregivers must now distinguish between enjoying the present with little sense of the past (as in stage 4), and having apparently little or no active sense of the present itself (as in stage 5). Does the woman in stage 5 still have a life that we can benefit by treatment, and do we benefit that life by curing the pneumonia that assaults it? Does she have-or, even if we cannot be sure, might she have-levels of awareness (even if physiological more than cognitive) that recognize and respond to changes in how her family and her doctors and nurses choose to treat her?

Unless there are complications that make the very activity of providing treatment a burden, or unless there are other health problems that suggest that her systems are shutting down and that she is irreversibly dying, the only ground for denying treatment seems (yet again) to be the judgment that her life should now come to an end. Non-treatment seems to entail the view that the life she now has is no life at all, meaningless and degrading to her and not worth the labors required to sustain it. Such labors are surely great for the caregivers-emotionally, physically, and financially. And the patient's life experiences and life options are severely limited. But it is also hard to see how deliberately acting with the intent or hope that her life come to an end, however well-meaning one's motives, is actually a form of best care for the person as she now exists. And it seems hard to imagine how one could withhold treatment from the patient in stage 5 without intending that she die.v

Looking beyond this particular case, any decision to treat a person with dementia differently simply because of the condition itself seems to require accepting one of the following propositions:   (a) that human dignity resides foremost in our sustained mental capacities, and that a human being with severe dementia is no longer a person, but a body without a person; (b) that the person's wishes before becoming mentally incapacitated should always be honored, and that if the once-competent person stipulated the cessation of medical care in the case of dementia, this particular act of self-definition should always be respected, no matter how much a demented person may still seem to have a life, and even a seemingly happy one; or (c) that the experience of becoming demented is so horrible-different in kind from being mentally incapacitated throughout life-that the progressive and irreversible loss of one's powers justifies non-treatment. More specifically, to treat patients with dementia differently from other classes of dependent elderly persons, one must believe that dementia is in itself a special kind of disability or suffering; that treating intervening illnesses is a way of perpetuating the person's degradation (that is, a kind of torture, not caregiving); that a person with dementia lacks dignity not because of his dependence on others for everyday needs but because of his changed cognitive condition; or that it is legitimate for an individual to believe that life with dementia lacks dignity, and thus permissible to direct his surrogates in advance to cease all care if he becomes demented. But for those who judge that being worthy of care and worthy of life does not depend on possessing certain cognitive capacities, such comparative judgments about the comparative worth of different lives seem morally misguided. Discrimination solely on the basis of dementia seems incompatible with the moral demands of equal respect and best care, an incompatibility that includes discrimination by a past self against a future self with dementia.

In case situations like this one, in which the necessary medical intervention is modest, likely to be effective,vi and does not require a struggle with the patient or a dramatic disruption of the patient's existence, treatment seems morally appropriate. But as the patient deteriorates and medical interventions become more complex and disruptive, that moral assessment will likely shift. We must ask whether aggressively treating the patient who can no longer be an active partner in her own recovery may turn the person entirely into an object, to the point where acting on the person no longer benefits the minimal life the person still has, but makes the person into a mere receptacle of technical intervention and the doctor into a mere technician. In cases like uncomplicated bacterial pneumonia, where treatment probably does not require a significant disruption of the patient's routine or physical coercion, the ethical demands of best care clearly point toward treatment. But in cases in which treatment would require sedation, physical restraint, frequent re-locations, or other complicating factors, dementia may become a factor leaning against particular treatments: not because being demented lessens one's moral claim on life, but because the treatment itself adds to the un-consenting and un-comprehending patient's miseries, burdens, or degradations. We will consider these complicated cases more fully below, as we consider what it means to promote the patient's well-being in the clinical context.

III. The Well-Being of the Patient

In the collection of cases considered in this section, we explore in greater detail how concrete treatment decisions are related to the personal welfare of persons with dementia. But if caregiving aims to serve the well-being of the patient as person, we need some operative idea of what well-being means, and how the well-being of a person with dementia is both similar to and different from well-being for persons in very different physical and mental conditions. We also need in advance a more detailed account of the different types of treatment-with the various upsides, downsides, burdens, and uncertainties of different medical interventions. Therefore, as a prelude to the cases themselves, we will briefly consider (a) dementia and well-being and (b) the nature of treatment and the trajectory of illness.

A. Dementia and Well-Being

For most of life, we live both immediately and reflectively. We experience the physical and emotional ups and downs of the present moment-the pleasures of eating a favorite food and the pain of burning a finger on the stove. We smile when we see a newborn child and shudder when we see a stranger get hurt. But our well-being also depends on how we see ourselves critically and self-consciously-on aspirations fulfilled and unfulfilled, on human attachments made and broken, on values affirmed and compromised.

With dementia, however, the reflective dimension of well-being fades and the person lives ever more immediately. The mature complexity of human relations is gradually replaced by the raw physical neediness more characteristic of an infant and the spontaneous (and not always appropriate) reactions to the presence of others and to their words and deeds. Life informed by thought about one's personal past and hopes for one's future gives way to life as a sequence of disconnected present experiences. As the individual's ability to see himself critically and self-consciously declines, the more complex psychological dimensions of well-being are eclipsed by the simpler psychophysical ones.

Responsible caregivers for persons with dementia aim at a series of overlapping human goods: the comfort of the person; the pleasures of activity as long as activity is still possible; a sense of connection to others; a stable, safe, reliable, and nurturing environment; and the good of being itself, of being alive rather than dead, even in a diminished condition. This does not mean, as we have often noted, that keeping the person alive is the caregiver's highest purpose. The probability or even certainty of an earlier death must often be accepted when life-extension involves painful or misery-inducing treatments. But for responsible caregivers, death itself will never become the means to relieving suffering, and life itself will never be treated as the burden to be relieved. We cannot, after all, serve the well-being of a person's life by seeking his death.

Promoting well-being for persons with dementia often presents special challenges in the clinical context-distinct from caring for independent persons at the height of their powers or physically incapacitated persons who are still cognitively able. Consider, for example, what it is like for a cognitively healthy person to have blood drawn by venipuncture: He knows beforehand why this procedure is necessary-such as the medical value of diagnostic tests or the civic value of donating blood. His understanding of the experience gives him a reason to endure or reinterpret what is, in the immediate sense, unpleasant and painful-indeed, a violation of his body. This understanding does not make the needle hurt any less, but it makes the pain humanly comprehensible.

A person with advanced dementia, however, no longer has the capacity for such understanding. Drawing blood may serve the person's present interests-as critically assessed and understood by others-but a person with dementia cannot understand how or why. He experiences the needle simply as a threat, a violation, a source of pain. His immediate experience is uncorrected by reflective understanding. Skillful caregivers might sometimes find ways to have drawing blood "make sense" within the patient's impaired cognitive reality. But in many cases, the violation will be experienced as senseless, because it will be experienced uncritically-as an act committed against rather than for the patient's well-being.

The point, once generalized, becomes crucial for ethical analysis and prudent judgment: Certain treatments that are not excessively burdensome for comprehending patients may be excessively burdensome for patients with dementia. Dementia itself does not make the person less worthy of treatment, but the complications created by the person's lack of understanding often change the judgment of how best to serve his well-being.

Finally, although concern for bodily well-being is central to caring for the person with dementia, caregivers take care not only of the body but also of the whole person who is and lives as and through his body. They are caretakers of his personal past as well as his bodily present, of the life he has lived in addition to the life he is now living. In this light, caregivers always need to consider the significance of past values and past character-such as the ideas of meaningful life that the person once held, or the ideals of family flourishing and personal honor that he lived by when he was still an ideal-forming and ideal-seeking being. The person whose needs are eventually largely bodily also possesses enduring moral interests, even if he is no longer aware of them.

This insight cuts in multiple directions: It means that even those who might be oblivious to our care still have an interest in our care, because they still possess the dignity of being human; and it means that even those who no longer know the ideals they once held still have an interest in those ideals-an interest entrusted to us. This does not mean that we must always allow the past self to speak decisively for the present self, or that we should judge the worth of the present life entirely by the standards of a person's past existence. But it does mean that caregivers, in caring for the unique person the patient now is, must honor the unique person the patient once was-for example, by being less aggressive in pursuing optional treatments for a person who all his life treasured self-reliance above all else, or, conversely, by being more aggressive in pursuing optional treatments for a person who always held that any vestige of life was incalculably precious.vii What such honor means and demands in any concrete case is often a puzzling question, especially when persons with dementia get sick in other ways, and especially when caregivers must begin thinking about not only the patient's continuing life but the patient's approaching death.

B. The Nature of Treatment and the Trajectory of Illness

To think clearly about how certain treatments or non-treatments might affect the well-being of persons with dementia in the clinical setting, we need a clear way of thinking about the complex character of "treatment" itself-with all its benefits, uncertainties, tradeoffs, and imperfections. The professional caregiver always begins with a diagnosis of the current situation: What is wrong with the patient, what are the options, and what will probably or likely happen if we leave the patient untreated? The diagnostician inquires about the present ("Is the patient suffering now?") and imagines the future ("How will the patient's condition change if we do or do not intervene?").viii Perhaps without treatment the patient would die imminently and certainly; perhaps he would face persistent or worsening discomfort and suffering; perhaps he would be at greater risk of more severe complications or earlier death in the near future; or perhaps these greater risks will amount to nothing.

Once we have diagnosed the current problem or problems, we then ask what kinds of treatment are possible: Can we cure the given illness entirely-as we may reasonably hope to do with a urinary tract infection using antibiotics? Can we manage the illness with ongoing treatments-as with kidney failure using dialysis? Can we slow down the progression of the illness-as with radiation treatments for some forms of cancer? Can we ameliorate the symptoms without altering the course of the illness itself? Does the treatment involve a one-time intervention, a battery of interventions for a fixed period of time, or a permanent regimen of interventions until death? Does it involve a long and difficult recovery? Will it create temporary or permanent disabilities? What are the likelihood of success and the likelihood of complications? Answers to all these questions, whether given explicitly or tacitly, are germane to any decision to treat or not.

Different treatments obviously entail different upsides, downsides, and levels of uncertainty. The best treatments involve no pain or disruption with a high probability of cure; the worst involve physically invasive procedures with little chance of improvement. Some treatments involve virtually no burdens-such as taking pills with minimal side effects. Others involve increasing levels of burden-from minor surgery with a speedy recovery, to radiation therapy that leaves the patient weak and nauseated, to interventions that result in permanent disabilities or deformities. For persons who are elderly, the burdens of treatment are often more severe. And for persons with age-related dementia, there are the additional problems of confusion and disruption-because the patient cannot understand what is happening and why, and because altering the patient's familiar routines is often especially terrifying.

Ultimately, caregivers must compare the burdens, consequences, and potential complications of the treatment itself against the burdens, consequences, and potential complications of non-treatment; and they must compare the likely realities of life after treatment against the likely realities of life without treatment.ix This is where the true moral complexity of the clinical situation presents itself, especially because persons with dementia who get sick are often already sick in other ways. In many cases, the life of the patient is already filled with many physical burdens. The goal of treatment is always to reduce those burdens, never to add to them. But the dilemma, in many cases, is whether adding days, months, or years to such a burden-filled life-even if the life-extending treatment effectively eliminates a discrete threat-is morally sound. Good care often means accepting that the life now present is a life with burdens that cannot be eliminated by our actions. The dilemma comes when we must decide whether to extend a burden-filled life in cases in which the life-extending treatment itself is not a burden or not a very great burden.

Especially perplexing are circumstances in which two illnesses are present simultaneously: an illness that would probably kill the patient relatively quickly and with minimal suffering (like bacterial sepsis), and an illness already under way and beyond cure, that will likely lead to severe physical distress on the way to certain death (like pancreatic cancer with metastases). In such a case, would the intervention against the curable disease bring suffering greater than the disease itself, by dramatically increasing the chance of additional suffering and a more difficult death? On the one hand, caregivers seek to benefit the life the patient still has by treating current problems; on the other hand, caregivers seek to avoid the prolongation of suffering and the most painful kinds of death. Do we benefit the life a person still has by extending life with severe burdens? Do we benefit the life a person still has by intervening in ways that cause new disabilities in the process of eliminating present dangers to life itself? Is the moral calculus different for elderly persons who suffer from age-related dementia compared with mentally incapacitated persons at earlier phases of life or physically incapacitated persons whose cognitive powers are still intact? These kinds of questions-and more broadly, the relationship between treatment and well-being for persons with dementia-can be considered only in concrete cases, with due regard for both the moral principles that should guide all caregiving decisions and the irreducible particularity of every clinical situation.

C. Three Complex Cases

To probe these issues, we will consider three types of cases in this section: The first explores the relevance of a patient's subjective well-being and outward behavior toward others; the second considers the burden of certain treatments and the significance of a patient's resistance to treatment; and the third considers the meaning of concurrent deadly illnesses, where a treatable supervening disease would likely bring a quicker and more peaceful death.

1. The Relevance of Subjective Well-Being.

Medically, both these patients have a similar heart condition requiring a similar treatment. Without a pacemaker, the danger of a cardiac arrest in the near future increases dramatically, and with it the likelihood of sooner and sudden death, or of a non-deadly but debilitating episode that leaves the patient in a considerably worse condition. But the two persons are also very different in the condition of their lives: one appears content and one does not, one is agreeable and one is difficult, one brings mostly happiness to others and the other brings mostly misery. Do these existential differences matter in deciding how or whether to treat?

For Patient #1, the moral presumption in favor of installing the pacemaker seems fairly strong, assuming the procedure itself is relatively non-burdensome and does not somehow disrupt the person's seemingly pleasant existence by permanently altering his everyday routine. With a pacemaker, this seems like a modest concern. But what about more demanding medical interventions that would disrupt the patient's happy existence?

Because the clinical possibilities are so wide-ranging, it is hard to generalize. If the medical condition in need of treatment is not painful in itself or life-threatening now, and if the treatment would likely disrupt the patient's current happiness, there are sound moral reasons to forgo treatment, even if doing so increases the likelihood of an earlier death. In such a case, the caregivers choose one kind of life (present happiness, with the possibility of a shorter lifespan) instead of another kind of life (present ordeal, with the possibility of a longer lifespan). With the pacemaker, the medical intervention seems morally desirable because the burden of treatment and the likely disruption are likely to be fairly minimal. With more burdensome treatments-such as chemotherapy or coronary-bypass surgery-the moral calculus clearly shifts, making treatment or non-treatment equally defensible options depending on the circumstances, or making non-treatment the morally preferable option in cases (such as the examples just given) where the burdens of treatment are severe. A dementia patient cannot comprehend the value or nobility of enduring painful interventions in the present for the sake of future goods, and this lack of understanding may in some cases provide grounds for forgoing certain treatments, either because of the burdensomeness of imposing an unwanted treatment on the patient or because of the danger of turning the patient into a mere object.

For Patient #2, the situation is more complex-especially if the individual strongly resists all efforts to undergo the procedure. We will consider the significance of a patient's resistance to treatment in our next case; for now, let us focus on the meaning of the patient's current discontent and bad behavior, and whether it deserves to be given any weight in making a treatment decision. The first question is whether the patient's apparent discontent and bad behavior are being caused by the caregiving environment itself or by poorly treated psychiatric problems. It may be that the behavior itself can be modified by better care. But let us assume, in this case, that Patient #2's situation cannot be dramatically improved short of constant heavy sedation. What then?

If the question for caregivers is how to benefit the life the patient still has, is installing the pacemaker a benefit to him? Certainly it is a medical benefit: it improves his bodily condition and decreases the likelihood of repeat episodes. But is it a benefit to the patient as a person? In this case, the person seems to experience life with dementia as a special kind of misery, and he certainly causes significant misery for others. Why intervene to help prolong such a life, if doing so means prolonging this misery? Let us assume that he was, before dementia, a thoughtful and decorous gentleman. Do we benefit the life the patient still has by prolonging his degradation and bad behavior? Can one choose not to treat out of fidelity to the person we once knew-that is, out of the loving desire not to prolong the deformation of his past character that merciless nature has now imposed upon him? Or would we be abandoning the patient if we treat his unhappiness and bad behavior as a reason to invite an earlier death, as disqualifying conditions for our ongoing care?

If the caregiver elected not to install a pacemaker, he would not be seizing an occasion to produce certain death (in the way, for example, that stopping insulin injections for a person with type-1 diabetes would). But the caregiver would deliberately leave the person in a condition where an earlier death of a certain sort is probably more likely. We cannot know for certain how the patient will change as dementia progresses; we cannot know for certain whether his apparent misery will get worse, or whether his mood might improve by changing his environment, or whether dementia will advance to the stage where his bad behavior will be eclipsed by further cognitive and physical decline. And we must ask whether our failure to treat now will only make his cardiac condition worse in the days ahead-with more cardiac episodes, more emergency situations, more falls from dizziness and perhaps permanent injuries as a result. We must ask whether non-treatment is also a kind of endangerment.

This case puts the dilemma of caregiving sharply before us: Prolonging life seems like prolonging personal misery; forgoing treatment seems like courting death. The strongest moral grounds for forgoing treatment seem to be the fear that the ordeal of even this minor procedure would make his catastrophic reactions and uncontrollable behavior even worse upon his return. One could argue that the complications caused by his dementia make treatment an excessive burden for this patient. But let us ask, by comparison, whether we would install a pacemaker in a similarly miserable, delusional, and disruptive psychiatric patient in his thirties, who could only be controlled with the use of heavy sedatives. If the answer is yes for the young psychiatric patient, then the reason not to treat Patient #2 would seem to turn on the age-related dementia itself: that is, on the grounds that he has lived a full life, that his disease is undoubtedly irreversible and progressive, that the process of becoming demented is an affront to his past dignity, or that he is ready to die precisely because he is both old and demented.

Perhaps the difference, in this case, is that the young psychiatric patient is physically strong enough to endure the ordeal of forced treatment, while the elderly patient is not. Or perhaps one would treat the younger psychiatric patient because there is a greater chance that medical progress will provide a future cure for his mental condition. These factors might make the two cases clinically and morally distinct.

But in the end, those who decide not to treat Patient #2 would need to think carefully about what they were really doing and why they were moved so to act. They would need to consider the unpredictable-and perhaps horrible-consequences of not treating: for example, a non-deadly heart episode that causes the patient to fall down and break his hip. Of course, they would also need to consider the human consequences of electing treatment-such as extending a life that appears miserable, a life that ends in a long dehumanizing decline. They might believe that they are under no moral obligation to act so as to make a person's fate worse. In this case, caregivers might feel a powerful moral intuition to forgo treatment, but it is hard to articulate the grounds for doing so without tacitly or explicitly declaring that the impairments of Alzheimer's disease are, by themselves, a legitimate moral reason to aim at death as the means of ending the individual's unhappiness.

As with Patient #1, one could surely envision many medical interventions for Patient #2 that would not be morally obligatory, including those that might be necessary to prevent an imminent and certain death. Certainly, one could decline a battery of treatments that required long-term physical restraints-treatments that necessarily treat the patient as someone trapped in a kind of medical servitude. But in this particular case, the moral analysis seems to lead in favor of installing the pacemaker in both patients-with sober recognition that doing so for Patient #2 may extend a life that exists with little or no present contentment. Often, with regret and anguish, we must opt for affirming what seems to be miserable or undignified life over death, and that choice should be at least the default position, in both law and prudence. In some situations, we are forced to recognize what Vergil called "the tears of things."2 Recognizing this, we must aim for the choice that does not define individuals in distress as better off dead. And in this case, the patient's very misdeeds and apparent discontent are proof that there is life left in him, and that death is probably not imminent.

2. Burdensome Treatment and Patient Resistance.

This case permits us to examine the relevance, for deciding about best care, of the burdens of receiving treatment-both the physical burdens of the given treatment itself and the burdensome effect that administering treatment might have on the patient's lived experience. It also permits us to look at the significance of patient resistance as a factor in deciding about treatments that might otherwise be medically and ethically indicated.

Before coming to the question about continuing hemodialysis, we first review the decision to initiate it: Was this decision compatible with ethical caregiving, an example of best care? The answer is not obvious.

On the one hand, it is likely that dialysis would preserve the patient's life for an indefinite period of time. In the absence of treatment, her kidney disease is probably deadly, sooner rather than later. In purely physical terms, hemodialysis is less taxing than, say, major surgery or chemotherapy. It would almost certainly be medically recommended for a younger patient with renal failure, or for a non-frail person of similar age who was not suffering from Alzheimer's disease.

On the other hand, dialysis is not happy treatment. Studies show that the majority of patients do not think that their life is good on dialysis (even when they are glad to be alive because of it). Sometimes too much fluid is taken off, sometimes too little. The shunts clot and get infected. Although it saves lives short-term in patients with end-stage renal disease, it often causes death from heart disease in a few years. Virtually no patient wants to stay on dialysis, and some patients, tired of waiting for a kidney transplant that often never comes, elect to discontinue their treatment rather than continue living under its burdens.

Unlike mentally competent patients facing the prospect of hemodialysis, a patient with moderate dementia cannot knowingly elect to seek its benefits despite its burdens, or to forgo those benefits because of those burdens. In one sense, our dementia patient's burden of treatment might in fact be less than usual: owing to her lack of self-awareness, she might not suffer the psychic distress that often accompanies living through the ordeal of never-ending treatment, contemplating death hovering over her all the time. In another sense, however, because she cannot choose or understand the treatment or its purpose, she cannot willingly understand or accept its physical discomfort or the alteration of her relatively comfortable and familiar daily routine at home.

In the present case, the physician apparently regarded hemodialysis as optional for the patient, extraordinary rather than obligatory, thinking the benefits did not outweigh the burdens in her case. Ethically, we think that this is a defensible judgment. But her family members, eager not to lose her and perhaps hoping also to keep her alive for a special family occasion already planned for the following year, opted for treatment, and the doctor readily obliged. Ethically, we think that this too is a defensible judgment. Both the doctor and the family caregivers were seeking to benefit the life the patient still had; neither was pronouncing judgment on its worthiness.

But after a trial of treatment, people have cause to reconsider. The treatment, especially the human ordeal it entails and the physical coercion it seems to require, turns out to be much more burdensome than anticipated. Should it now simply be stopped? Or are there things to be considered and tried first before reaching such a conclusion?

One proper avenue is to see whether there might be ways to reduce the burdens and to dissolve the patient's resistance to treatment-ways to make the necessity of treatment "make sense" to the patient or at least seem less burdensome, or ways to make the patient a more willing (or less resistant) participant in her own care. If the resistance is caused, for example, by the way the patient is being treated by her caregivers at the dialysis center, a change in their behavior or approach might ease the current ordeal. If the patient might now also be suffering from depression, antidepressant medication might improve the patient's behavior. Because traveling by car to and from the treatment is especially distressing, peritoneal dialysis-less effective than hemodialysis but administrable in the home-might well be considered. Alternatively, one might think of moving the patient into an assisted living facility where dialysis is available onsite, and where perhaps a new, stable life routine can be established-though one should not underestimate the stress such a move might well produce.

It is possible that some combination of new medications and new caregiving strategies will still make hemodialysis possible without imposing an excessive burden and without radically altering existing caregiving arrangements. It is possible that the patient will tolerate peritoneal dialysis at home. But if these new strategies do not help the situation, it may be that ceasing treatment altogether and restoring (if possible) her pre-treatment routine is what best care requires, even knowing that the consequence will be an earlier death.

The decision to cease treatment can never be made lightly, especially in a case where death is hardly imminent so long as treatment continues. But in cases such as this where patient resistance makes the very activity of getting treated a great burden-not simply physically in terms of pain, but humanly in terms of the patient's overall well-being-the decision to cease treatment and accept an earlier death seems morally permissible once other alternatives fail, and it may even be the best choice among a range of imperfect options. Forgoing treatment might allow the patient to live out her last days with the happiness that her relatively stable and peaceful home life still permits-with minimum disruption or struggle, and without being treated as a mere object of care requiring permanent coercion.

As in all cases, the obligation of caregivers here is to ensure that they do not choose the path that benefits their own lives and well-being at the cost of the patient's life and well-being. Every avenue of easing the burden of treatment should be pursued before coming to the conclusion that treatment in this case, for this person, is an excessive burden that should not be continued. But in the end, there is a highly principled and prudent case for forgoing treatment: One is seeking to serve the patient's well-being here and now. One is not using death as the means to end suffering; one is not seeking death deliberately by one's own actions or omissions; one is simply seeking the best life possible under the circumstances, even if the best life now possible is a shorter life.

We would probably think differently about the problem of burden and resistance if the patient were not an elderly person with dementia but a resistant child or a middle-aged psychiatric patient who believed being taken to dialysis was a form of despotic control. Caregivers will and probably should give patient resistance different weight in these alternative cases, and there seem to be at least some moral grounds for choosing differently based on the person's age and underlying condition. In the case of the child, one imagines that he will one day appreciate the parental coercion that kept him alive. In the case of the psychiatric patient, his life trajectory is longer than the incapacitated person with age-related dementia, and the possibility of benefiting the patient's life in novel ways remains a greater possibility. But for the elderly dementia patient, it is almost certain that the person's dementia-and thus her inability to understand the need for dialysis-will never get better and only get worse. And there is little hope for improvement from medical progress.

The potentially greater obligation to treat the psychiatric patient is not because the younger person's life has greater dignity than the older person's life; it is because the different clinical conditions of the two individuals alter what it means to benefit the life each patient currently has. In both cases, the bar for forgoing treatment should be set high-but even higher in the case of the psychiatric patient, because the possibility of a significant future improvement is also being forgone if and when hemodialysis is stopped.

The ethical dilemma in all these cases-where the patient resists for reasons that do not make sense or for reasons that do not necessarily serve the patient's best interests-is deciding what weight to give such resistance. There are some who believe that resistance on the part of dementia patients is autonomy's subconscious way of expressing itself-that is to say, it is the incapacitated person's way of expressing a will to die or the willingness to surrender to the inevitable. But more likely, this resistance is the activity of an individual living purely and spontaneously in the present-unable to fathom past wishes, unable to fathom the future consequences of non-treatment, and fearing in some instinctual way what it will mean to lose the experienced goods of the present, such as staying at home where things are familiar and not scary.

In such cases, the incapacitated person's well-being is entrusted to surrogates acting on her behalf. The ethical dilemma is that intervening medically on behalf of the patient seems to require acting against the patient, and in some cases the present burden of a given intervention-an intervention that is burdensome in part because of how it is seen by individuals who cannot see things clearly-outweighs the future benefit. In such cases-and the specific case here presented seems to be one of them-caregivers can morally choose a path that ends in an earlier death, because such a path is the best life possible for the patient under the circumstances. But they should always seek the fullest range of life-sustaining alternatives that do not wreak undue havoc on the goods the patient still has in the present.

3. The Trajectory of Death.

This case forces us to confront a very difficult question: Do we benefit the life a person still has by extending a life marked by terrible physical suffering and with worse physical suffering still to come? More broadly, how do we balance the obligation to heal the sick with the obligation to ease the pain of those who suffer? Thinking clearly about this case requires assessing the progression of her cancer, the nature of her suffering (both here-and-now and looming ahead), and the significance of the person's dementia, which forecloses the patient's freedom to direct her own care without eliminating her experience of physical misery. It also confronts us concretely with the moral difference between the burdens we produce by our acts of medical intervention and the (possibly even greater) burdens we (indirectly) make possible by extending an exposed life to the looming miseries of a disease-in-process.

The patient in this case is clearly heading toward death but probably not yet irretrievably dying. As the disease progresses and death looms closer, the case for not treating the pneumonia surely seems more compelling. At some point, we might say that the patient has become a "whole ensemble of diseases," that death is too close to fight any longer, that treating a particular treatable disease no longer benefits the patient in her totality as a person. With possibly six months left to live, one must be hesitant about seizing an occasion for death by leaving the pneumonia untreated. But as six months becomes six weeks or six days, the case for surrender strengthens, especially as the debility of the cancer gets worse.x

Obviously, at every stage of care, physicians should do what is medically possible to control, lessen, or ameliorate the patient's physical pain. In most cases, pain relief (or partial relief) is possible, if one pursues it diligently. But in some cases, short of permanent sedation that itself brings increased risks of earlier mortality, adequate pain relief is not possible. And so it is worth taking up the case precisely as written: with death from cancer looming three to six months ahead, with the person suffering from advancing dementia, with pain present now and likely to get significantly worse day by day, and with a decision in the hands of caregivers about whether to treat a concurrent bacterial pneumonia.

As in the case described earlier, using antibiotics to treat her pneumonia would most likely be effective-though it is important to recognize, first, that any readily treatable pneumonia has a substantial chance to resolve itself spontaneously, and second, that any pneumonia likely to be lethal is potentially dangerous even if treated.xi And in the immediate, physical sense, the treatment itself is non-burdensome: it does not involve moving or violating the patient, or causing her new kinds of discomfort or distress. As caregivers, however, we must think about the future as well as the present, about future burdens as well as present needs. In treating her pneumonia-even if we assume it is the right thing to do-we make it more likely that she will endure more future pain than she would were the pneumonia to take her life. We cannot avoid that stark reality.

Accordingly, compelled to choose whether to give antibiotics, we must wonder whether doing so is really doing good-especially for a patient with advancing dementia whose existence is defined almost entirely by physical experience. The dementia patient lacks the self-conscious loves and longings that might give a competent person a reason to fight on, a reason to endure terrible physical suffering, a reason to live because there is something knowable to live for. The demented person may reach a point-if she is not already there-where the pain is so awful and her cognitive capacities are so limited that life with cancer is sheer and incomprehensible misery, with nothing positive in her subjective experience to compensate or sustain her.

So why might we decide to treat? As caregivers, we do not want to refuse treatment as a way of getting the patient to die. (Indeed, were that our plan of action, we might have difficulty explaining why we should not simply end her life now, pneumonia or no pneumonia, and spare her the ordeal of the next three to six months.xii) Thus, even assuming the best and most merciful of motives, we should still draw back from acting in a way intended to bring about her death.

But there is more to be said. In such difficult circumstances it is always right to be suspicious of our own motives. The human psyche is extraordinarily complex, and what we tell ourselves is mercy can sometimes be something quite different. Made uneasy by this woman's dementia, seeing in her a sign of our own vulnerability and mortality, suffering at our own inability to relieve her suffering as much as we would like, we may prefer that she simply go away as soon as possible.

If we treat her pneumonia, this does not mean that we want her to suffer or that we cruelly impose suffering upon her. If we have been her caregivers up to this point, we might well hope-or pray-that she die soon and suffer no more. Nevertheless, part of honoring our commitment to the equal dignity and worth of every human being, irrespective of condition or circumstance, is to accept the demands of solidarity even in suffering.

Our most fundamental commitment is not to minimize suffering but to maximize care-never to abandon care for another human being. In many, perhaps most, instances these two obligations amount to the same thing. But there may be occasions-and this case may be such an occasion-in which maximizing care, never abandoning care, means that there is at least some suffering we cannot relieve.

The reasons to treat depend, finally, on a judgment that this woman is not yet irretrievably dying-though, to be sure, she is terminally ill. Six months is not a short time; it allows for her life to be touched by loved ones (though her capacity for interaction is already severely limited by her dementia), or it gives time to enjoy whatever simple pleasures (if any) still mean something to her. At some point-hard to calculate-this ceases to be the case. The range of life's possibilities narrows so greatly that forgoing treatment is not minimizing care but lovingly accepting the natural conclusion of a life. Even when we argue for treatment, therefore, we should recognize the limited scope of the argument in this instance. Perhaps we are not yet at the time for stopping such treatments, but we are surely approaching it, and it would be foolhardy to make assured judgments about it.

How might we make the argument against treating her pneumonia-even now when she may have as much as six months to live? We might judge that this woman has come to a point where we can find few or no ways to benefit any longer the life she has. We might conclude that we should let nature take its course and acquiesce in her dying, being present with her and comforting her as best we are able. We might conclude that extending her life, even if medically possible, is not humanly wise. This would not be an attempt to get her to die, justified by an appeal to merciful motives; we would not think ourselves to have failed as caregivers should she survive the pneumonia without treatment. On the contrary, it would be the best way we could find-in what are admittedly very difficult circumstances-to remain committed to her life and well-being. And even as we leave her pneumonia untreated, we should do everything we still can to care for the person in her final trial; forgoing this particular treatment does not mean forgoing all treatments, since some interventions may still make her existing life better even as death looms closer.

A further reason not to treat lies in what we might think of as a humanistic understanding of medicine. The point of medicine is not to treat or cure a disease, such as pneumonia. The point of medicine is to treat human beings-persons who are ill and suffering. What we want to treat here is not so much pneumonia as it is this woman-and, hence, we deal with the pneumonia as a way of caring for her, not simply in order to overcome a disease. Yet, she has now become, as we put it earlier, an "ensemble of diseases." If antibiotics are not useless for treating the pneumonia, they may now be useless for treating her. If we can find no ways to benefit the life she has, then we should step back from attempts to treat. Here again, however, we should not be too confident of our judgments. We should be wary of how mixed our motives may be, and we should be reluctant to decide that we have reached the point where treatment is no longer a benefit to the living person still with us. Nevertheless, our patient must always remain the woman, not simply the diseases that afflict her.

This is surely a hard and puzzling case with no easy answers. It appears to be a case in which the distinction between "active killing" and "letting die" might take on real significance: which is to say, we appear to possess the moral discretion, as caregivers, to let pneumonia take its course, but (as always) never the moral freedom to kill the patient directly.

In trying to decide what to do in such a hard case, the particularity of the patient as a person might also take on greater significance. If the patient's children lived far away and wanted to be with her before she dies, that may be a reason to treat. If the patient was or is a religious person with a redemptive understanding of suffering, this fact might move loving caregivers to treat. Conversely, if the patient had an especially low tolerance for pain, or believed that generosity toward the next generation means accepting death after a long life whenever death arrives, then permitting pneumonia to be this "old woman's friend" might be the way to do our best as her caregivers. Ideally, loving family members and caring doctors would come together to make a prudential judgment in such a case, a judgment that would depend not only on the particularity of the clinical situation but also the particularity of the patient as a person with a past, a present, and an (albeit short) future life.

In the end, we believe that the moral argument could go either way in this case, depending in part on many personal and medical particulars, those just mentioned among them. Even those who lean toward treating the pneumonia should recognize that the wisdom of treatment diminishes as the cancer and its accompanying pain worsen and as death looms closer. This is a case in which, notwithstanding which side gets the better of the moral argument, it seems wise to leave family members and doctors some leeway to decide, lovingly and prudently, when to surrender to nature-but not the right to actively kill, even with supposedly well-meaning motives. As we struggle with our awesome responsibilities in heart-wrenching circumstances, we should remember that even our perfect desire to see another's suffering end does not make us the absolute master over that suffering or the authoritative arbiter of that person's life. Sometimes, the best we can do is stand with those who suffer and make sure they know and feel that they are not alone.

IV. The Significance of Living Wills

In caring for the person the patient now is, caregivers should never ignore the wishes and values of the person the patient once was. This requires respecting more than their written instruments; it obliges caregivers to consider the reasons that animated the person who wrote them. It would be important, in this case, to know whatever we could about why the patient chose to write an advance directive of this kind, with blanket instructions to forgo all invasive treatments: perhaps it was the desire never to burden his daughter; or the belief that life with dementia would not be worth living; or the desire not to exhaust all his resources on end-of-life care; or the fear that he would lose control of his behavior and act in inappropriate ways; or the worry that extending life indefinitely would be physically painful; or the belief that he should be allowed to die peacefully rather than being hooked up to machines. But absent such knowledge of his reasons-and it is often not easy to come by, even for close family members-caregivers are obliged to take seriously the instructions as such.

Seriously, but not slavishly. For past wishes, as we have explored, are not always morally decisive in such cases. They are a crucial point of consideration, but not the only or even the most important one. No individual can foresee every future circumstance in his or her life; an individual's best interests and true needs can change over time; and medical situations are so complex that we can only judge wisely what to do case-by-case and in the moment. Only by making an all-encompassing determination that his life with (more than minimal) dementia would never be worth sustaining might the competent individual rule out in advance all future treatments should he become demented. But such a blanket assertion about the worth of a future self denies the intrinsic dignity of embodied life even when one's cognition is impaired; it discriminates against an imaginary future self long before the true well-being of that future self is really imaginable. And although there are genuine reasons to tremble at the prospect of life with advanced dementia-and to wonder whether even non-burdensome treatments simply impose prolonged degradation-responsible caregivers should not acquiesce in denying the worth of the person entrusted to their care, just as competent individuals should resist making a priori negative judgments about the disabled person they might become.

The primary moral obligation of caregivers is to serve the well-being of the patient now here, and to ask not only what the patient would have wanted but what we owe the person who lies before us. This means paying attention to an advance instruction directive if one exists, but not following its orders regardless of all other circumstances. If the patient seems miserable because of his dementia, there might be greater cause to respect his living will. If he physically resists treatment, there might be cause to forgo further interventions and accept an earlier death. Yet, neither of these conditions is present in the case as we have it. When the individual in question wrote his living will, perhaps he could not imagine the desire of his daughter to care for him, or the fact that he would be so cheerful with dementia, or that he would be willing to endure surgery of this kind. All of these factors, so difficult to anticipate beforehand, should make us hesitate simply to apply his instructions in a mechanical and thoughtless manner, as if we were not ourselves reflective moral agents responsible to do the best we can for the man now before us. In this case, there seem to be clear and compelling reasons for caregivers to override the terms of the advance instruction directive and proceed with treatment, understood by both the daughter and the doctors as the best way possible to benefit the life the patient still has.xiii

The same would be true in an opposite kind of case: where the advance directive (written in the past) demands every possible treatment to keep the person alive, but where all possible interventions (in the present) would be excessively burdensome or would prolong a painful process of dying. In such cases, caregivers would also have significant moral grounds to override an advance directive and forgo certain interventions.

Our case might be further complicated by certain economic considerations. Suppose that the daughter knows that her father wrote the living will primarily as a means of preventing the spending down of assets he wanted to leave to his children and grandchildren. Yet the daughter, although mindful of his intentions, would much rather spend her father's money (her own prospective inheritance), and even her own money, to ensure that he receive the best medical care. Under these circumstances, she would be justified in exercising discretion, overriding his wishes where treatment is clearly beneficial and not exorbitantly expensive, although perhaps avoiding costly treatments in circumstances in which treatment is ethically optional.

In raising these complications, we hardly mean to suggest that the presence of a living will is irrelevant in deciding the best course of care. In many cases, there will be no reason to resist its instructions, either because they accord with what lovingly prudent caregivers would recommend anyway or because they recommend a course of treatment (or non-treatment) that is among the justifiable options. In other cases, caregivers will disagree about what to do, especially when the demands of best care are ambiguous, with compelling reasons either to treat or not to treat. In cases such as the one presented here-involving a not excessively burdensome treatment, a cheerful and physically strong patient, and the fact that forgoing treatment will lead to a painful and imminent death-prudent judgment points toward overriding the living will, even when caregivers disagree. But in many cases, where the best interests of the patient are not so clear and where caregivers disagree, there is solid moral ground to defer to the living will.

Consider, for example, the two cases discussed earlier: the patient with kidney trouble who physically resists being taken to treatment and the patient with deadly cancer who contracts pneumonia. In those cases, caregivers face wrenching choices about what best care requires, and they might rightly turn to an advance instruction directive, if it exists, for guidance in deciding what to do. In such complicated circumstances, where loving caregivers are not certain what best care requires, an advance instruction directive might provide the moral authority or needed permission to fight on or let go, knowing how hard either path will surely be. But, as we have emphasized throughout, the authority of living wills should always be limited by the obligation to serve the patient now living among us, by always seeking the best care possible under the circumstances.

V. The Well-Being of Caregivers

In this case, the demands of caring well for a person with dementia have taken an enormous toll on the caregiver's family; and the decision to transfer the patient to a nursing home may have altered the trajectory of her final days-though exactly how we can never know for sure. How do we judge the behavior of the caregiving daughter, the husband and sons, and the attending physician?

The true caregiver always aims to do his or her best for the person being cared for, especially when the dependence of the needy person is absolute. But the moral obligation of best care always exists alongside many other roles and obligations: for example, one's role as a nurturer and provider for young children; as an employee at a company; as a volunteer in the community; as a spouse for better and for worse; as a person with aspirations of one's own. Caregivers are not caregivers only. Of course, they should always strive to make decisions aimed at the well-being of the person with dementia-asking what is best for the patient, not what is best for the caregiver. But caregivers should also see the obligations of caregiving in light of life as a whole, with its many attachments, many burdens, and many purposes. This is why caregivers rightly ask: What do I as a caregiver owe him or her as patient and person? For this is a question that encompasses both the unchanging worth of the life being cared for and the genuine limitations of being human caregivers. Those with unlimited resources might owe their loved ones more in terms of medical care; those nations with greater wealth might owe their citizens more public assistance and support. Prudence requires honestly making such assessments, both as individuals and as a community.

As a simple rule of thumb, caregivers should do the best they can do; they are never compelled to do what they cannot do, but they are obligated to see how much they can do without deforming or destroying their entire lives. But in practice, this rule of thumb rarely leads to any fixed rules, because every person faces different demands and has different capacities. And inevitably, we cannot do our best simultaneously in every area of our life: that is to say, we cannot do our best for everyone all the time; we cannot be there for everyone all the time; we cannot devote resources to everyone equally all the time.

To be a caregiver is to confront not only the limitations of the person with dementia who relies upon us entirely, but our own limitations as human beings who are more than just caregivers or who are caregivers in multiple ways for multiple people. In doing so, we need to avoid two kinds of dangers: the danger of not working hard enough for the person with dementia who needs us, by saying too early or too easily that the burdens of care are too great; and the danger of betraying everything and everyone else in one's life, by saying too often and too persistently that the burdens of caregiving can be shouldered alone, or that caregiving for the person with dementia always trumps every other obligation or good.

Life, of course, comes in many phases, with obligations and opportunities that shift over time: As a mother of a newborn child, an adult daughter may be more limited in the time she can devote to caring for a parent with dementia; as an adult son struggling to provide his teenage children with a college education, he may be more limited in the resources he can give to his parent with dementia. Alternatively, the successful corporate lawyer may need to put his or her career on hold when a parent becomes dependent; the phase of life in which the pleasures and obligations of career are dominant may need to give way, for a time, to a period of life during which caregiving becomes dominant. In all situations, the presence of myriad obligations requires prudential discernment about which obligations place the greatest claim on our time, our resources, and our very lives at particular moments. This challenge presents itself to caregivers day after day, not only when persons with dementia get sick in other ways, and not only when life-and-death decisions need to be made about medical care.

In this case, the daughter is heroically trying to care for her mother-a mother who could not survive, living at home, without her care. Her husband and children also need her as a mother and a wife; and though they could endure without her, they increasingly feel abandoned, and increasingly see the grandmother/mother-in-law as too great a burden to keep at home. All individuals, in this case, need to ask themselves probing questions: The caregiving daughter should ask whether she is paying attention to the genuine needs of her children and husband: Is she showing interest in their lives? Is she available to them as a source of guidance or strength in hard times at work or at school? Is she paying attention to the education and moral development of her children? The husband and children need to ask whether they are being selfish: Are their own problems and needs so great that they cannot endure them a bit longer, and comfort their mother/wife by supporting her in her role as caregiver? Can these needs wait until later-when the life of the grandmother/mother-in-law has ended, and this phase of the caregiving daughter's life has ended? Are they meeting the obligations of fidelity toward the wife/mother they miss so much?

These burdens and tensions are often not only emotional but economic: Should the funds set aside for the children's college tuition be used to pay for an in-home health care aid or for a private assisted living facility? What if the choice is between an inferior nursing home covered by Medicaid and superior nursing care paid for out of the tuition fund? Or what if the choice is between working longer hours to pay for optimal care and working fewer hours to provide care oneself?

Without question, caregivers should never seek the death of the person being cared for in the name of other familial goods. But they will almost certainly make caregiving decisions in light of other family realities-such as whether to keep the person at home or put her in a nursing home, whether to keep working and hire outside help or leave work to care for a dependent parent themselves, whether to scale back children's extracurricular activities to spend more time at the nursing home or to carry on as usual. In making such decisions, both families and societies face the most fundamental kind of ethical dilemma: a dependent person with immeasurable worth who depends on finite people with many other obligations. On the one hand, we wish to ensure that care for persons with dementia is the best possible; as persons of incalculable value, they deserve no less. On the other hand, we need to confront the fact that best care for the person with dementia may mean less-than-ideal care for everyone else who needs us, and it may mean doing less than our best in those activities of life where we aspire to excellence. Because the person with dementia often cannot survive without our care and often teeters between life and death, we are moved to put everything else aside to meet the demands of caregiving. We live, with them, in a perpetual state of near-crisis or real crisis. At the same time, because the person with dementia seems to have such a diminished life, we might be tempted "to put them aside," so to speak, so that everything else can flourish. The moral challenge is finding the best possible balance among these competing obligations, while operating within the fundamental moral boundaries articulated above: never seeking someone's death in the name of other goods, always doing the best one can for the person entrusted to one's care.

Finally, the physician in this case understandably wants to help the family as a whole, and he may have offered sound advice in recommending that the patient be moved to a nursing care facility. From a strictly therapeutic perspective, it may be that the patient has been at home too long, without ready access to regular professional care. But it also seems far too simple to say, as the physician does, that the patient "would not recognize the change." As a matter of cognitive understanding, this is surely true: the patient no longer recognizes her daughter or the home she lives in. But although it may not be obvious or evident to observers, persons with dementia often do "recognize" when the kind of care they are receiving changes-not cognitively, but experientially or physiologically. Moving the patient to a care facility may have been the right thing to do, and perhaps the caregiving daughter should have done so even earlier. But we also cannot ignore the fact that doing so sometimes means a degradation, not an improvement, in the quality of care; and it sometimes leads to a shortening of the person's life-not by aiming at her death or leaving life-threatening illnesses deliberately untreated, but simply by weakening her basic "urge to live" by leaving her in the care of strangers or in a foreign environment. This fear is what makes putting a loved one in a nursing home so difficult for many families, especially those without the economic resources to pick whatever care facility they want. The gravest decision faced by many families is often not whether to treat or not, but whether caring for a family member at home is the morally sound thing to do-for the patient, the caregiver, and the caregiver's entire family.

*          *          *          *          *

The cases presented in this chapter illustrate the difficulties conscientious caregivers face in trying to provide the best care possible for people who can no longer care for themselves. Our discussions have explored what ethical caregiving might require in such cases, suggesting ways of thinking and acting sufficient to the task of benefiting patients with dementia in the clinical context, often in circumstances that admit no perfect options or happy outcomes. Although the goal is clear-to serve the true welfare of the patient here and now-loving prudence must often struggle to find the right means of serving it. Caregivers should always seek to benefit the life the patient still has, even when they elect to stand aside because further interventions would only make matters worse and even when they recognize that surrender is the path of loving care.

To deal with the ethical complexities found under even the best of circumstances, our discussions throughout have assumed the presence of competent and caring physicians and loving and engaged family members, all of them trying to do their best. Needless to say, these ideal conditions are not always present, perhaps not even usually present. We are mindful of the temptations to treat patients with dementia as second-class human beings or "non-persons"; we are aware that doctors are often impatient or inattentive, that nurses are overworked, and that family members are easily tempted to put their own interests first, sometimes even rationalizing betrayal of the patient as the course of mercy. For these reasons we worry lest our counsel of "allowing to die" in a few of these difficult cases may serve to encourage less than loving caregivers to betray and abandon their charges or encourage society at large to move down the slippery slope toward the moral pit of euthanasia. Nevertheless, it remains our obligation to develop the outlines of what ethical caregiving requires here and now, in the hope that such knowledge can itself help protect against the dangers that lie ahead and enable us all to treat the vulnerable among us according to the better angels of our nature.



i. To be sure, such resistance has its costs and, hence, its decent limits. The family should not resist to the point of driving themselves apart or of inflicting strife all around.

ii. We owe this insight and example to Dr. Joanne Lynn.

iii. Alzheimer's disease is, of course, more than the dementia through which it most prominently manifests itself. As we indicated earlier, it is a progressive and ultimately fatal disease-unlike the multi-infarct vascular dementias that may or may not be part of a fatal picture. This means that differential treatment of patients with Alzheimer's might be based not on the presence of dementia as such but on the whole disease process. This fact becomes more relevant in thinking about treatment decisions in the middle and late phases of the disease.

iv. Compare the situation of advanced Alzheimer's with two other clinical cases: (a) the young person in a persistent vegetative state and (b) the middle-aged person in late-stage terminal cancer. In the case of the persistent vegetative state, the person lacks all cognitive capacity, but she is not suffering from a degenerative condition. In many cases, the patient is stable; the body is not shutting down, even if it requires artificial feeding to sustain it. In the cancer case, the body may be breaking down, but the persistence of self-awareness means that the struggle for life still continues in a unique sense; the personal will to fight on exists in contradistinction to the realities of bodily breakdown. With advanced Alzheimer's disease, however, there is neither stability as the body begins to fail nor the presence of a self-conscious patient struggling for life. This unique combination of cognitive and physical breakdown, this clinical instability combined with advanced age, raises questions as to whether the dying process has already begun and whether death should continue to be opposed. Perhaps some treatments are better seen not as preservation of life but as prolongations of the dying process, even if such prolongations last for a while.

v. In order to focus only on the relevance of the degree of dementia, the case as presented supposes a simple case of community-acquired pneumonia, manifested in cough, fever, shortness of breath and general misery, but in the presence of adequate blood pressure and good enough cardiac and renal function. But if the patient were in shock or suffering kidney failure, and the pneumonia were just one manifestation of generalized breakdown and a harbinger of more or less imminent death, the benefits of treating the pneumonia would be highly uncertain. Antibiotic treatment would be less likely to succeed; and even if successful, the patient might be relegated to spending the rest of her life in the intensive care unit. Under these circumstances of vital organ system failure, treatment of the pneumonia would become optional, not only for the person in stage 5 but perhaps even for the person in stage 4.

vi.Accustomed as we are to the miracle cures of antibiotics, unimaginable by our grandparents, we are prone to overlook the fact that, despite the high success rates, even the treatment of the common pneumococcal pneumonia is not guaranteed to succeed. It can be followed by lung abscess, bacterial meningitis or brain abscess, septicemia with shock, empyema, endocarditis, and acute respiratory distress syndrome. Some of these complications of treatment would require invasive, even heroic, measures. The simple treatment of simple pneumonia is not necessarily always simple.

vii. There are, of course, limits: Honoring the unique person the patient once was does not mean that we would be obliged to give hemlock to Dr. Kevorkian.

viii. It is hard to imagine a situation in which it is ethically sound for caregivers to tolerate physical suffering in the present when such suffering can be ameliorated with treatment or palliation-unless the medical intervention itself will cause suffering or impairments that are worse than the underlying ailment. What to do in cases where the burden of treatment in the short-term is very great but the long-term benefits are significant is a complicated question. Indeed, caregivers must make very complex judgments about the relationship between present goods and future harms and between present harms and future goods. For competent patients directing their own care, one could imagine avoiding treatments or forms of palliation that compromise self-awareness, and one could imagine patients who might prefer to suffer physically, to the degree such suffering is tolerable, in order not to affect their psychic state. But in dementia cases, such self-awareness is already compromised, partially or completely, and thus the obligation to ease physical suffering seems preeminent in most imaginable circumstances.

ix. Though the account here focuses on the professional caregivers, especially the doctors, these people must also be able to explain these matters to family members or others who, as proxies for the patient, have the ultimate responsibility for deciding what should be done.

x. Of course, one of the great dilemmas faced by caregivers is the inability to know with certainty how much longer a given patient will live, or whether death is finally arriving, no matter what we do to fend it off. Such uncertainty could lead us in either direction regarding treatment. But it should induce a certain modesty before the awesome choices forced upon us in this dance with death, even as we unavoidably must make them despite our imperfect knowledge.

xi. Antibiotic treatment usually increases the odds of curing pneumonia by some middling percentage, roughly 30 to 60 percent; neither treatment nor non-treatment have certain yes or no effects.

xii. But on this point, see our earlier discussion in Chapter 3 of the partial moral relevance of the distinction between deliberately killing and allowing to die.

xiii. Those who disagree with this judgment because they oppose overriding the living will should ask themselves whether they would also object to surgical pinning of a fractured hip, should the patient fall and break it, leaving him in pain and unlikely to walk again.



1. Robert Burt, Taking Care of Strangers: The Rule of Law in Doctor-Patient Relations. New York: Free Press, 1979, p. 11.

2. This passage appears in Vergil's Aeneid, I.462: "Sunt lacrimae rerum et mentem mortalia tangunt." ("These are the tears of things, our mortality cuts to the heart.")


Next Chapter Right Arrow

  - The President's Council on Bioethics -  
Home Site Map Disclaimers Privacy Notice Accessibility NBAC HHS