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


This commissioned working paper was discussed at the Council's September 2004 meeting. The views expressed here do not represent the official views of the Council or of the United States Government.

Commissioned Working Paper

When Persons with Dementia Get Sick:
The Care-Giver's Dilemma

Eric Cohen

When a loved one suffers from dementia, it often (almost inevitably) falls on care-giving family members or attending physicians to make decisions about proper care. Of course, some individuals with dementia have "advance directives" - like voices from the past, struggling to give directions for that time (then imagined, now here) when one's powers of self-determination have faded.  But in the present, care-giving decisions fall largely or entirely on surrogates. And even the most detailed advance directives often require interpretation in light of complex circumstances, and sometimes seem to conflict with the welfare of the person now here as opposed to the interests of the person he or she once was.

In many cases, the surrogate has been making decisions on behalf of a family member with dementia for years - decisions about finances, driving, living alone, or entering a nursing home. These decisions involve not just how to care for the person with dementia, but also how to reorder one's own life to make such care-giving possible - whether leaving one's job, spending less time with one's children, or taking a parent into one's home. Eventually and often more-than-once, an acute care-giving decision arises: the person with dementia gets pneumonia, or suffers from a cardiac arrhythmia, or needs dialysis or tube-feeding or artificial respiration to stay alive. The life that a demented person now has hangs in the balance; the death that is surely coming may or may not be here. What should we do as loving care-takers, and how should we decide? What boundaries, aims, and factors should shape such decisions? And what should society do to set ethical boundaries and protect personal choice, while also confronting the reality that care for the elderly is one social good among many?

This paper is partly occasioned by Robert Binstock's suggestion at the last meeting that the Council take up the question of how and when to treat supervening illnesses in persons with moderate or severe dementia. Medical ethics has long given consideration to the question of forgoing life-saving interventions, and even "pro-life" moralists have held that life-sustaining treatments were optional at best if (but only if) they were either too burdensome or inefficacious.

But as we confront the obligations of caring for persons with dementia or the prospect of suffering dementia ourselves, it may be time to re-examine traditional approaches and current practices in the care-giving realm. Many of our ethical categories and bedside protocols were devised before vast numbers of people could be expected to live many years without the capacity to look after themselves or even guide their own care. In the past, most people known or thought to be dying were likely to die very soon, regardless of what doctors did to intervene. Today, however, people can live many years in increasingly enfeebled and debilitated conditions, and if they are dying, it is difficult to know how long they might last. More and more people suffer from chronic and progressive illnesses, including dementia, and medicine is at the same time more able to intervene successfully to reverse supervening acute conditions (such as infections). Given these new realities, the ethics of forgoing life-sustaining interventions, especially when an individual is not obviously in the last weeks or months of life, seems to require a fresh look - even if the older ways of thinking should turn out to be still applicable and sound.

The focus here is on the dilemmas of caring for persons with dementia - a situation that mixes the responsibilities of long-term care with acute medical decisions at various stages of the disease. This is not a discussion of the ethics of euthanasia or assisted suicide when one becomes enfeebled or demented, but of the ethics of intervening against secondary illnesses. The paper is about whether and when and how vigorously to treat, and why? i

These decisions take place within certain ethical boundaries that set outer limits to what is and is not acceptable. But within these boundaries, they are almost always decisions requiring the discernment, prudence, and sound judgment of loving proxies, seeking the proper aim in the light of many relevant factors. This is a realm where fixed rules can offer only limited guidance; it requires asking many questions with no obvious or universally applicable answers; it requires practical wisdom in circumstances that are emotionally demanding and ethically complex. But it is also a realm where not anything goes, and where some rough guidelines can perhaps ensure that loving care-givers act in truly loving ways.

The proper aim, I argue, is always "best care" for this patient, under these circumstances, at this time. It means a covenant between the person with dementia and those entrusted with his or her care, a covenant embedded within a society that must always confront its own limits, and embedded within the larger human drama of decline, dependence, and renewal. Not far in the background of this discussion is the question of whether a world in which dementia is a universal fact of family life requires fundamentally new ways of thinking about our ethical obligations and about the ties between the generations.

Such a discussion risks the impatience of those who wish for straightforward answers, or who wish to let everyone decide for himself while fleeing the responsibility to guide the perplexed who must choose. The paper will perhaps ask many more questions than it answers. But it will also suggest that not every care-giving decision is equally good, even if discerning the good is not often easy.

I.  Framing the Question

It makes sense to begin with the decisions as they present themselves in concrete cases: Should a person with early-stage dementia - still aware enough to know what is happening and what is coming - decide to forgo dialysis? And should his family intervene in favor of treatment? Should a person with middle-stage dementia - confused, dependent, but generally happy - undergo surgery to install a pacemaker? And should it matter if the person with dementia is not happy but miserable? Should a person with late-stage dementia - incapable of walking on her own, recognizing her spouse, or controlling basic bodily functions - be transferred to the hospital to treat pneumonia? ii

Every situation is different: the patient is a little happier or a little healthier in one case than in another; the burdens of care-giving are greater on one family than another; the patient's earlier wishes are more or less clear. But judging wisely in particular cases requires some understanding of the kind of dilemma that care-givers face, and the nature of the choices open to them.  This is a task we undertake here in three parts.

First, what are the boundaries of acceptable care-giving?  When should seeking or forgoing treatment be ethically off-limits? As Rebecca Dresser has written, "Just as there are legal limits on the range of choices parents may make for their children, so should there be limits on the range of choices competent persons have concerning their future treatment and the choices available to proxy decision-makers in general. In short, we as a society must take a moral and legal stand on which categories of incompetent patients must be treated and which may have treatment forgone."

Second, within these ethical boundaries, what should be the guiding aims of the loving or responsible care-giver? Several aims present themselves: doing what the patient would have wanted, if he were competent to say; doing what is best for the patient now, as she is now; doing what is best for the family as a whole; doing what is best for the larger society; maintaining life by every ethical means possible; limiting physical or psychic suffering; trying to ensure a good death or avoid a terrible one. We must examine these different possible aims, and puzzle through what they mean and what claim each of them properly has on surrogate decision-makers, especially when different human goods seem to be in tension.

Finally, what are the relevant factors in making concrete care decisions, and what weight should each of these factors be given? These factors include: the human meaning of dementia (as opposed to other forms of sickness or decline that make one permanently dependent on others); the stage of dementia; the experiential realities of the demented person; the current welfare or suffering of the person; the nature, burdens, and likely success of treatment; what life will be like after treatment; the proximity of a demented person to death; the possible effect of treatment on the trajectory of death; the character and values of the person with dementia before the disease; the burdens and blessings of continued care-giving for the family; and those larger obligations that set ethical limits on the realm of choice.

In the end, as said above, this paper suggests that care-giving decisions involve difficult discernments within certain ultimate limits, where the highest aim is "best care" - an idea that encompasses both the solidarity and obligations of the care-giver to persons with dementia, and the dignity and obligations of a demented person even in his days of dependency and decline. This does not mean that "precedent autonomy" - the expressed wishes of demented persons, before the onset of disease - retains no place in shaping decisions about care.  And it does not mean that we should ignore the kind of person our loved one was - noble, religious, intellectual, repentant - in deciding how best to care for him now. But it does mean that the place of precedent autonomy is finally limited by the embodied reality now before us: the person whose life hangs in the balance is not autonomous, but dependent. And the real-life decisions for a demented person's care have fallen to us as loving proxies.

II. Defining the Boundaries

The first task is defining the boundaries: What decisions or approaches to care-giving should be "off limits"? And what are the limitations of setting fixed rules for care-givers?

(1) Clearly, from an ethical perspective, autonomy should not be unlimited. As Leon Kass and others have argued, there is no absolute right to die; no right to suicide; no right to control completely the timing and circumstances of one's own death. We do not let healthy individuals donate all their organs, even if the individual's freely chosen death might save many others. And we don't simply allow depressed persons to starve themselves, even if they rationally claim that leaving this wretched life is what they truly desire. Moreover, autonomy is not only limited as a right, but also as a human reality. In the case of dementia, the patient is not autonomous but dependent; he relies on others to guide his care, even if such dependency extends no further than having his advance directive meticulously executed. The limits to autonomy in care decisions reflect the deeper reality that autonomy is limited by mortality itself. Being human means being biological, and being biological means that one's will is sometimes limited by the realities of being an embodied self. iii

But this general truth about the limits of autonomy is more puzzling when one considers specific care-giving cases, especially in a rights-based society like our own. An individual with early-stage dementia may have a legal right, while still partially competent, to forgo certain kinds of medical treatment, even if such treatment would extend life or improve health. He might choose to forgo the heart surgery he needs to unblock his arteries, or reject the antibiotics he needs if he gets pneumonia, or even stop eating in order to preempt the looming depravations of dementia. But is he ethically right to make such choices? Which ones, if any, and why?  What are the proper limits to his autonomy?

Forgoing heart surgery seems obviously more justifiable than forgoing antibiotics or food: the treatment for heart disease is more burdensome than taking pills or eating or being fed, and heart disease strikes us unwillingly in a way that starvation in this case does not. Moreover, it may be that getting heart surgery is the right choice for some individuals and the wrong choice for others at similar stages of decline. But the ethical justification for seeking or forgoing heart surgery requires an appeal to human goods or obligations that go beyond autonomy. It is not enough to say simply: "This is my choice." One must argue why one's choice for or against treatment is a good one.

The scope and limits of autonomy are even more puzzling in the moderate and late stages of dementia, where care-giving decisions are made entirely by surrogates.  And the dilemmas are not simply resolved or removed by the presence of an advance directive; in some cases, the dilemmas may be even more acute. For example: If an individual in middle-stage dementia - confused, dependent, but seemingly happy - contracts pneumonia, should his advance directive stipulating the rejection of all medical care be respected or over-ridden? The law surely protects the family and doctor who respect the advance directive, and it may even punish those who override it. But do care-givers behave ethically by letting the person with dementia go without treatment? Should they honor the person's prior wishes, as expressed when he was in a position to express his own interests and values? Or should they seek to benefit the life the person with dementia now has by overriding the advance directive and offering treatment?

In both of the above examples, the question is whether autonomy - in the form of present demands or anticipatory instructions - should be limited by other ethical goods and obligations. And in both cases, there is strong evidence that there should be some ethical limits to autonomy; that individual wishes should not reign supreme; that the privilege of choice does not justify all choices. And yet, defining the boundaries is no easy task. Aiming at death should be off limits, and rejecting non-burdensome treatments (or food) that extend life and improve health for the person now here seems ethically questionable, though perhaps not absolutely off limits. The individual or the care-giver might appeal, in such cases, to the desire to live (or see one's dependent loved one live) a certain kind of life in one's final days, where death is the unavoidable consequence of such a life but not the directly chosen goal. But the decision to forgo treatment on these grounds requires an appeal to goods or aims that go beyond autonomy - aims that can be examined critically in ways that a mere assertion of personal autonomy cannot be.

(2) It is also clear, from an ethical perspective, that proxy decision-making should not be unlimited, and that some decisions, while legally permissible, may not be ethically justifiable. The responsibility to make decisions for another person does not confer the right to make any decision whatsoever.  And even those who have a special claim to know what is best for their loved one - spouse on behalf of spouse, adult child on behalf of an aging parent - must shape their claims in accordance with certain universal obligations, even as they seek to shape their decisions in light of the particulars of their own case.

But again, this general truth is more puzzling in concrete cases. For example: Should the spouse of a demented person have the ethical freedom to reject all treatment for his or her beloved, even when the patient is in the early- to middle-stages of dementia and likely far from death, and even when the treatments are relatively non-burdensome and have a high likelihood of success? Is the desire to live without the burdens of a demented loved one - or the claim that a demented loved one would rather not be a burden to himself or to me - an ethical justification for rejecting treatment? Conversely, should loved ones be allowed to demand aggressive treatment without limits, even if the treatment has little chance of success, or will cause the demented person significant pain, or will compromise the small pleasures that remain in his or her final days, or will make it more likely that he will suffer a more protracted, painful, or degraded decline and death? In both cases - the dangers of under-treatment and the dangers of over-treatment - proxy decision-making should be limited by other ethical goods and obligations. But defining these boundaries is, once again, no easy task.

(3) The limits to both patient autonomy and proxy-decision-making in the personal realm have corollaries in the social realm.  Clearly, a public policy that mandated all available treatments, where the only goal is extending life as long as possible, would seem to deform other goods and considerations - such as dying a peaceful death or living comfortably in one's final days.  Conversely, a public policy that mandated non-treatment of all demented persons once they lack certain capacities (e.g., speech, bowel control, mobility) would also seem to be misguided, by foreclosing the possibility of discerning those instances where treatment is warranted even in spite of meeting such criteria. (For example, perhaps a grandchild in the Peace Corps has not yet made it home to say goodbye, or perhaps a son has not reconciled with his sick but still living father.)

In the majority of cases where patients can no longer recognize loved ones or control their bowels, one might legitimately argue that anything more than the most rudimentary and non-burdensome curative treatment for disease should be rejected in favor of palliation. It may be, for such people, that enough is enough, and that the time for passing has come. But this does not mean that it would be wise for the state to make it illegal to give individuals who meet such criteria more intense treatments if particular circumstances call for it. There is a difference between prudential rules of thumb and the activity of prudence in concrete cases.  Moreover, one recognizes the need for societies to allocate public resources to a variety of civic goods, and one could imagine or even applaud a society that decided not to pay for expensive heart treatment for middle- to late-stage dementia patients. But establishing non-coverage criteria based on age or quality of life or stage of dementia is different from mandating non-treatment or aggressively discouraging treatment as an official policy.

Both public policies - mandatory treatment and mandatory non-treatment - seem to defend a partial truth: in the former case, the good of caring for the sick and the obligation to preserve life; in the latter case, the need for the old to make way for the young and for society to allocate public resources to a broad range of civic goods. But these partial truths are deformed by making them absolute, and by eliminating the possibility for prudent decisions in individual cases, where sometimes doing what is generally sound is not what is called for in doing what is best here and now.

(4) A full defense of certain ultimate limits on the range of treatment decisions for persons with dementia would require more explanation and argument, including a fuller case for why it is wrong to choose death. But the following limits seem to remain, for most people, both wise and acceptable, iv and it is possible to give a succinct account of their justification: (a.) No aiming to kill and no active killing, either on one's own behalf or as a proxy for loved ones. The decision to withhold treatment may foresee death as a possible or even probable result, but it cannot ethically aim at death directly; it must aim at another good, such as avoiding additional suffering or living one's final days with maximum dignity. Moreover, the freedom to avoid a bad death is not the same as the freedom to ensure a good death, and this limit against aiming at death encompasses a limit on the right to control the circumstances of death in a fully premeditated way. (b.) No treatment in cases where the treatment itself will impose excessive pain or suffering with limited prospects of success; the obligation to preserve life cannot become a warrant to inflict undue harm. (c.) No system of "capacity-based rationing" - i.e., a system that mandates the withdrawal of treatment without exceptions, on the grounds that individuals without certain capacities no longer have a right to medical care. Such a system would assume that such people are always an unjustifiable burden on others or that their lives no longer have significant worth - both questionable and arguably wrongful assumptions. Societies may decide not to pay for certain treatments for people with moderate or severe dementia. But non-coverage is not the same as mandating non-treatment or aggressively discouraging treatment as a matter of public policy.

(5) Having posited these outer limits, there is much work left to do. We must consider the proper scope of choice within these ethical boundaries; and we must consider the general aims and relevant particulars that should guide care-giving decisions on behalf of demented persons.  Clearly, there are some things that should never be done; there are some treatment or non-treatment decisions that would be wrong in all circumstances. But the challenge is discerning what to do in cases where the right decision is ethically puzzling and existentially hard.

This requires at least a brief comment on the nature of choice itself, and the meaning of what we do when we choose one course of care for persons with dementia rather than another. In short: Do individuals have the right to choose because different people have different values, and thus weigh different goods differently, such that what is right for one person is not right for another in the same or similar circumstance? Or do individuals bear the burden of choice because every situation is so particular, with so many different factors, that doing what is right always requires a choice; it is never simply obvious? In the former case, choice is really autonomy: the freedom to define ideals and seek to live in accordance with them, or even the freedom to do what feels right without justification. In the latter case, choice is really a matter of discernment: not the freedom to do whatever one happens to think is best, but the space to discern what is best or right, with due acknowledgement of the difficulty of seeing what "best" or "right" really is. These two ideas of choice are not the same: the first respects all choices (within certain boundaries) as equally justified, so long as they are conscientiously made; it is grounded in the notion that different people have different values or the idea that freedom itself is the highest good. The second notion sees some choices as more praiseworthy than others, and sees choice as the obligation to discern what is best in particular cases, and the obligation to accept the weight of doing what is right and good as essential to our well-being.  Needless to say, most people do not act with an explicit theory of choice; and perhaps most people simultaneously see their own values as universally right - "I did the right thing" - while remaining tolerant of those whose values lead to very different choices. This tension between universal obligations and tolerance for competing notions of the good is central to liberal society, and central to seeing the care-giver's dilemma in its fullness.

III. Understanding the Aims

Now we turn to an exploration of the aims that should guide care-giving decisions for persons with dementia. As will become clear, sometimes the following aims hold together and sometimes they are in tension. And for each aim broadly understood, there are many possible meanings and thus different courses of action. Specifically, care-givers must discern which aims have the greatest ethical claim: Is it the best interests of the person with dementia or the best interests of the family as a whole? Is it strict adherence to the terms of an advance directive or care in light of a demented person's new reality? Moreover, care-givers must discern what it means to act wisely on a given aim: for example, what it really means to do what is in the "best interests" of a person with dementia, even if one is willing to bear any burden, use any resource, and make any sacrifice to care for one's loved one.  In the section that follows, we examine different possible aims - both their wisdom and their limits, as well as the complexities of understanding what they demand in relation to one another.

(1) Respect for the Competent Person:

A prominent aim in our current formalized system of care-giving for persons with dementia is the obligation to respect the wishes of the competent person the patient once was. In theory, this idea can have two general meanings: at the extreme, it might mean that only competent persons command respect as persons, and that all demented individuals are objectively less human than competent ones. But in the main, it is the idea that all individuals should have the freedom to decide for themselves, while they are competent, whether life with dementia has any meaning or dignity, and what kinds of treatment should be pursued or rejected on their behalf. It is this second meaning that has been formalized in our system of surrogate care-giving, most especially in the form of advance directives.

There is a partial wisdom in this approach: It aims to honor the distinctness of individuals and to maintain a connection between the person one was when fully self-aware and the person one might become (or now is) when demented. It aims to set some limits on the types of decisions care-givers can make on behalf of persons with dementia, and perhaps to protect against the dangers of under-treatment or over-treatment. And it aims to offer some clear directions to future care-givers, who might tremble at the prospect of deciding the fate of a loved one without any guidance.

But in practice, respect for the wishes of the competent person once present - or respect for what he would have wished, even if he did not say so explicitly - is not by itself a sufficient guide for loving care-givers. In many cases, the wishes of the once-competent person are not clear: many people cannot fully envision what they would want in a situation they have never experienced; or they cannot speak for a self that does not yet exist; or they do not wish to confront the prospect of dementia while still in their prime. And oftentimes, the once-competent person had mixed and contradictory feelings about what kind of care he would want if he became demented - worrying about being a burden to others but wanting desperately to be loved, believing that life without memory is worthless but that death is worse.

It is also the case that one's interests change over time, and change dramatically with dementia. The distinguished scientist might believe, in his prime, that life without mental acumen will have no meaning - but it might turn out, once demented, that he is happy and at peace. The proud parent might believe that total dependence on his children is too undignified to be contemplated - but then lose such pride once dementia sets in, and happily take the care that his children happily extend him.

More broadly, as discussed above, self-determination has intrinsic limits in a civilized society: we cannot simply do with ourselves as we please; we should not simply choose the manner of our own death by a willful act. Indeed, even the competent person's wishes must be limited, because sometimes even one's wishes do an injustice to the value of one's own life, or to the concerns of one's loved ones, or to the norms of the broader society.

In the real world, the burdens of decision happen now, in circumstances that past directives cannot always imagine. These decisions fall on us - the care-giver. And we experience the loved one with dementia as they are now - perhaps saddened by the capacities that are gone, but also attached to the person still present, who is more to us than just a corpse, and more to us than a set of instructions. We cannot respect the competent person's past wishes without also taking into account the incompetent person's present needs; and we cannot make decisions only for the self that exists in memory, but also for the self that exists in embodied reality.

This does not mean that the character of the individual before the onset of dementia should be treated as irrelevant in deciding how to care for the person he now is: to treat the past as irrelevant would deny the continuous life trajectory of the person; it would deny, in a different way, the reality of the embodied self. But the challenge is not simply doing what he would have wanted, but seeing if there is any way to help the person with dementia live in accordance with the ideals he once self-consciously held, but in a condition where full self-awareness is no longer possible, and without transgressing certain ultimate boundaries. And more broadly, it requires that we look to aims and obligations beyond precedent autonomy in framing decisions for the incompetent person now entrusted in our care.

(2) The Good of the Family:

In deciding how to care for a loved one with dementia, a care-giver might aim at the good of the family as a whole, not just the wishes or well-being of the demented person. For example, a care-giving spouse might think in terms of a demented person's children and grandchildren, and how continued care and continued dependence affects their well-being. Or a care-giving child might think in terms of the good of her own family, and whether the obligation to sustain life must at some point give way to the flourishing of the next generation.

One approach to aiming at the good of the family might be strictly utilitarian, weighing the costs and benefits of sustaining a demented person's life for everyone in the family: Does the living presence of a parent or grandparent, in his or her current condition, give more joy than heartache? Is the cost of care to the family compensated by the continued life that this expense makes possible? And when does this cost-benefit calculus change?

But such narrow utilitarianism seems out of step with our society's approach to these questions and our best ideas of the family. Central to the nature of the family are the bonds of fidelity it entails, both for better and for worse. We do not (or should not) enter into and out of families like contracts, where the only purpose is our maximum pleasure. Rather, we live in and through families - as parents, children, grandparents, spouses - in a spirit of unconditional mutuality, recognizing that just as others care for us, we may one day care for them, and recognizing that mutuality is not always the same as reciprocity: sometimes in families, we must give more than we seem to get. By reducing an individual family member simply to the pleasure he or she gives other family members, we risk undermining the very idea of the family that such a utilitarian calculus aims to promote. We would undercut the idea of the family in the name of the family.

But even if one affirms the bonds of fidelity and mutuality, matters are not so simple. For example, a care-giver might ask how the person with dementia himself might seek to benefit the family, even in his demented state. Or a care-giver might reflect upon the proper balance between caring for the old and nurturing the young, and discern that a particular ailment at a particular stage of dementia, while treatable, is also nature's way of restoring the balance, to be accepted rather than opposed. Or one might weigh the significance to the young of seeing firsthand what it means to care for the old, or the significance of being reminded what a blessing health really is, or the significance of being taught what it means to be mortal.

There is no doubt that caring for a demented person imposes real burdens on the family, both on the middle-aged and the young. But it is not always so easy to discern the meaning of those burdens, or to envision what would happen to the family bond if the old were left to die in the name of the young, or the weak left to die in the name of flourishing.

(3) The Good of Society:

Most individuals, when making decisions about how to care for loved ones with dementia, will not decide in terms of what is good for society.  But it is also the case that family members are not the only actors in this drama: there are the doctors and hospitals who make recommendations about best care; there is society as a whole, which pays a large proportion of health-care and long-term care costs for the elderly, and which must weigh these goods against other civic goods; and there are the fundamental values of society, such as nurturing the young, securing the equality of all, and protecting the vulnerable from harm. Thus while individuals may not aim at the good of society in making decisions, society as a whole will establish the conditions that influence those decisions - including the influence of law and culture on the ethical intuitions of the individuals who bear the responsibility of care. 

Moreover, just as families face hard economic decisions and trade-offs, so does society as a whole: Wealthy societies may be able to build both nursing homes and nursery schools, but even the wealthiest societies cannot do everything imaginable for everyone; they cannot maximize the well-being of all individuals all of the time. What kind of society would we be if we left the elderly with dementia (and their families) to fend wholly for themselves? But what kind of society would we be if public subsidies for long-term care or heroic medical interventions were financed by cuts in educating the young or securing the peace?

To the broad question of society's contract with the elderly, one can envision a broad range of answers: Some argue that society should enact certain objective non-treatment standards or public rationing, on the grounds that we should not divert valuable health-care resources to care for those with a low quality of life and little time left.  Others argue that respect for the dignity and equality of all persons is the central value that makes society possible, and that the supreme good of society is to ensure that this value is publicly affirmed rather than undermined by age-based or capacity-based rationing of health-care. Still others believe that the best way for society to deal with such ethically fraught matters is through a procedural solution - allowing individuals and families maximum possible freedom to decide what is best for themselves, while preserving tolerance for different worldviews and civic peace within the society as a whole. 

But none of these answers, by themselves, is wholly satisfactory. There is a partial wisdom in seeking to set social limits on treatment: This policy confronts the reality of mortality, the passing of the generations, and the interconnectedness of all social goods; and it acknowledges the potential danger of squelching the young in caring for elderly persons whose dependence becomes a long second childhood. But such a policy also risks dehumanizing persons with dementia - treating them as public burdens, as non-members of society, whose claim on us is limited by their very weakness and dependence. Such a policy might directly or indirectly impose death even in cases where simple medical treatment would benefit the life a demented person still has.

There is partial wisdom in seeking a procedural solution, one that defers care-giving decisions as much as possible to individual families. And yet, as discussed above, a policy of autonomy without limits gives individuals maximum freedom but offers little guidance to patients and care-givers about how to live that freedom, and it fails to confront the potentially dehumanizing excesses of autonomy itself - including the possibility of seeing one's own life as less valuable than it truly is. Moreover, a policy built on self-determination alone offers little guidance about what society owes its dependent members, and what its dependent members owe society.

The third view - that society has an obligation to respect the full dignity of persons with dementia and should not see dementia as a reason for unequal treatment - seems closest to the truth as an abstract principle, but also wholly inadequate by itself as a practical guide to policy. It still leaves us with the difficult discernment of deciding what is truly in the best interests of patients. And it offers little guidance about the best ways of affirming the dignity of demented persons in a society where resources are always limited. It says nothing about what obligations persons with dementia may have to the society of which they are part, including the possible obligation to "stand aside" for others who have not yet lived a full life, rather than seek every possible remedy in their final days or in an extended decline. Respecting the dignity of persons with dementia does not settle the hard question of whether society should pay, for example, for coronary by-pass surgery for someone with moderate to severe dementia, or whether dementia (as opposed to age and level of dependence) should be given any special weight in making such a decision. These types of hard questions remain for prudent care-givers and prudent citizens to decide.

(4) The Good of the Person Now Here:

At the heart of being a care-giver and surrogate decision-maker is trying to see the world as the demented person sees it, and perhaps occasionally seeing the worth of the person's life, even when life with dementia may no longer seem to be a benefit to the person himself. This is, of course, no easy task, but it is also not an impossible one. As Rebecca Dresser has argued, we can do a much better job than we presently do understanding the phenomenology of persons with dementia; we can pay more attention to those experiences that make little sense to us - such as reading books without paying attention to the order of the pages - but often give persons with dementia experiential pleasure. It also means seeing the demented person not simply as an individual with interests for himself, but as a life whose value transcends the self's own interests. But even as we try to empathize with persons who suffer dementia or defend the possibility of being demented and happy, we cannot lose sight of the deprivations that often or eventually come with this disease: the loss of modesty and self-control; the outbursts of anger; the physical discomfort; the need for feeding tubes or physical restraints. Seeing the good of life with moderate or severe dementia often means seeing the worth of a life that had been stripped of many worthwhile and dignified things.

Approaching care-giving decisions in this way focuses on a core question: How do we benefit the life this person now has - even if it is not the life a demented person would have chosen for himself? This straightforward question does not always have a straightforward answer. To benefit the life a loved one now has means discerning the true good of a demented person - and more particularly, this demented person, with this history, these attachments, in this condition. Surely physical well-being is central to the obligation of best care, and so finding ways to ease suffering and treat physical ailments as they arise - broken hips, the flu, urinary tract infection - would have a strong claim. But there are other potential claims as well, going beyond the good of the body. Does one benefit the life a person now has by trying to prevent an excessively painful death or a death in triage, even if it means accepting an earlier death? Does one benefit the life the person now has by allowing him to live out his final days with at least some remnant of past nobility, thus choosing (for example) non-treatment for a middle-stage person with pneumonia, whose wits may be gone but whose self-control over the body remains intact, and whose presence still awakens reverence rather than horror in his grandchildren? Does one benefit the life a person now has by choosing life even if it means a greater duration of physical suffering, thus making the person a cross-bearing witness, present to be loved unconditionally by his spouse?

Surely, different people will see the good of the person with dementia in different terms and thus make different decisions; they will value different factors differently. But all care-givers and potential patients must be careful to avoid two potential errors: doing what is best only for the care-giver in the false name of what is good for the patient; and refusing in advance to be a burden to loved ones for their sake, when in fact they would rather bear the burdens of care than the burdens of not caring. Spirit and motivation matter in these decisions, and two decisions that look very similar may have a profoundly different meaning when seen in their totality.

(5) Excursus: Aiming at Death or Choosing a Life?:

Before moving to consider the particular factors that should be weighed in concrete care-giving decisions, it might be worth pausing to consider a puzzling case:

A noble doctor, who devotes his career to saving life and devotes himself above all to the flourishing of his family, is diagnosed with Alzheimer's. A proud and noble man, he knows firsthand the ravages of this disease on families - the way it interferes with grandchildren's little league games, the way it interferes with the marital life of one's adult children, the way it interferes with the professional obligations of the next generation. And he shudders at the thought of being a burden to his children - children he knows would be capable of providing him loving care with equanimity, but whose lives and attention to primary duties would also be disrupted by doing so. And he shudders at the prospect that his wealth will go to pay for his long-term care rather than for the education of his grandchildren. And he shudders at the prospect of standing before the young as a broken man rather than a man of self-possession and wisdom.

A few days after being diagnosed with Alzheimer's, the man comes down with pneumonia, and seeking to avert the years of degradation and decline that would surely follow, seeking to live his last days nobly, believing he has lived a good and full life - he refuses treatment, and he dies a few weeks later.

How should we regard this man's decision? Did he aim at death, and therefore treat his own life with less dignity than he should? Or did he aim at living a certain kind of life - a noble life - and simply accept death as the cost of his own nobility? Was his decision not to treat the pneumonia really any different than taking hemlock? Would the noble act require hiding the decision not to seek treatment, and preserving the appearance of dying without recourse? Does his decision, if known in full, send a message to his wife and children that he does not believe they have the inner resources to care for him? Or does it send a message that he loves them too much to be a burden? Does his desire to die as a noble man in fact make him seem small - afraid of his looming dementia and dependence? What if the motive of the decision was self-giving - seeking to leave his wealth to educate his grandchildren rather than spending it all to care for himself in his final years? Would the decision be different if it were made a few years later (in the middle-stages of dementia) by his wife, who knew his wishes and loved him for his nobility? Does the man possess an ethical right to reject treatment that his wife does not have? Finally: Should we admire this man or see his behavior as misguided?

IV. Weighing the Particulars

In the third part of our analysis, we move from the boundaries and aims of treatment decisions to the relevant particulars of caring for this person at this time.  Prudence (or practical wisdom) seeks to do the best thing possible in the light of the circumstances. The previous section dealt with the aims; this one deals with the circumstances. The circumstances do not by themselves determine anything - rather, with his eye on the desired goal of "best care," the care-giver seeks to discern what each concrete case demands.

This section deliberately raises more questions than it answers. It lists the many factors that should be taken into account in discerning how best to care for a person with dementia, and probes how these different realities should be weighed.  In making such a list, one does not endorse the "checklist" approach to such decisions, as if each factor could be given a score, and as if some algorithm could tell us what to do without doubt, anguish, or ambiguity. Matters, alas, are not so simple.

(1) The Meaning of Dementia: Perhaps the first relevant consideration is the meaning of age-related dementia itself, and whether this particular condition differs from other forms of dependence - such as the non-demented elderly person without mobility or bladder control, or the incompetent younger person who did not lose mental powers with age but never had them at all. There does seem to be a human difference between the never-competent person and the once-competent person, and this age-related transformation seems to matter humanly speaking: He 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 no longer recognizes the woman he married; she is no longer the grandmother who takes the young to the zoo but the grandmother who acts funny 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 makes their dependence different in kind.

In the case of total physical dependence, the person is no doubt also transformed, and his very 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 some care-giving decisions. Moreover, the dependent person who is not demented may still acknowledge, if in small ways, the 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 and so out-of-control. With 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 while one might say the same with all forms of decline and dependence - the great dancer in a wheelchair is also no longer himself - mental loss seems to be different. The great dancer in a wheelchair still knows he was once a great dancer; whether such knowledge makes life better or worse, easier or harder, is not a simple question.

But the question for us is whether the special changes that come with dementia should affect decisions about medical care: Is a demented person really closer to death because the self-conscious self is no longer present?  If not closer to death in time, is he more ready to die, or even partially dead already? Does a person with dementia have a lesser claim to or a lesser interest in life-sustaining treatment by virtue of his dementia? And if so, why? v

(2) The Stage of Dementia: Early, Middle, and Late: A second, related consideration is the stage of dementia - both in its effect on the person's general level of dependence and its effect on the person's mental capacities. The first part of the question turns on the meaning of dependence itself, not the special character of dementia: As a disease like Alzheimer's progresses, so does the patient's level of dependence on others for the activities of everyday life. This raises a broader question: To what extent, if any, should care-giving decisions turn on a person's degree of neediness? For example: Do people who are still mobile and still in control of basic bodily functions have a greater moral claim to life-sustaining treatment than those who lack mobility and bladder control? If so, why?

The second consideration relates more specifically to the meaning of different levels of mental decline: Should those who lack certain cognitive capacities but still know themselves be treated differently than those who no longer know themselves or their loved ones? And should those who no longer recognize loved ones but still possess the powers of speech be treated differently from those who can no longer speak?  In short: What is the significance, if any, of the stage of dementia for care-giving decisions? And what account should care-givers take of the fluctuating capacity of many demented persons?

(3) The Welfare of the Demented Person: One must consider the general welfare of the demented person who is ill - both physical and existential. Is the person suffering terrible pain that can only be treated by heavy sedatives? Is the pain chronic or periodic? Has the patient lost control of normal bodily functions - speech, mobility, bladder control? And if so, how does the patient seem to regard this loss of function?

Beyond physical welfare, what is the existential condition of the patient?  Does he seem peaceful or troubled? Does he seem to get any pleasure from his interactions with others - such as visits from loved ones - even if he does not know who they are? Does he show even the simplest sign of basic physical pleasure in response to a warm sponge or stroked head? Conversely, is the patient depressed, erratic, angry, violent, or withdrawn? Does the patient's welfare improve with certain kinds of attention or care - such as participating in activities, being visited by an old friend, or being spoken to regularly by a nurse rather than strapped down in a lonely room? And if there is an illness in need of treatment, is the illness the likely cause of the person's current distress? In the fullest sense, what kind of life does the patient now have? And is the current happiness of the patient really relevant to deciding how or whether to treat life-threatening illness?

(4) The Nature and Likely Effects of Treatment:

What is the nature of any potential treatment? Will it impose immediate burdens on the patient, and are they burdens he is equipped to endure physically and mentally? What is the likelihood of success and the nature of recovery? Will there be potential side effects that will adversely affect patient welfare even if the treatment is successful? Will the treatment threaten the goods the patient now has? How will the treatment, if successful, potentially affect the nature of the patient's death?

Also: Is there a difference between treatments that require surgery and those that rely only on medication? And is there a difference between medical care and tube-feeding? If so, why?

(5) Patient Response to Treatment: What significance should we accord to a patient's visceral and adverse response to medical care? Should we take this as a sign to stop treatment? Or is the negative response simply the result of the patient not knowing what is happening to him, and not understanding that the temporary burdens of treatment will benefit the life he has?

(6) Is the Patient Dying?/Proximity to Death: Is the dying process clearly underway, such that any treatment (such as chemotherapy) might delay death but not cure the underlying illness? And how close to death must one likely be to be called "dying"?  Are all patients with Alzheimer's already dying, or must death be more imminent for this prognosis to have a significant effect on treatment decisions?

(7) Advance Directives: Does the patient have an advance directive? Does it deal precisely with the current circumstances? And are there sometimes ethical or existential grounds for overriding it?

(8) The Unique Person - Character and Values: Did the person with dementia make his wishes clear, before becoming demented or in the very early stages of dementia, about how he or she would like to be cared for later on? What weight should care-givers give to such past wishes, especially when they seem to conflict with the current patient's best interests? More deeply, how should the character of the demented person throughout life or in his prime - noble, repentant, intellectual, humble, selfish, sweet, self-sacrificing - affect judgments about best care for him now?

(9) Family Circumstances and Loving Attachments: How has the patient's dementia affected his closest human attachments - spouse, children, grandchildren, friends? Is there someone here willing and seeking to care for him? Is the family ready to say goodbye? Is the patient's care causing excessive personal or financial burdens on the family as a whole? Is it interfering with the flourishing of the young?

(10) Larger Ethical Obligations and Boundaries: How does the decision about proper care for this patient, in these circumstances, at this time, accord with larger ethical obligations binding on everyone - such as the obligation never to aim at death, the good of all the generations, and the inherent value of every life?

V. Discerning the Good

The aim of this paper is not to settle the question of what to do when a person with dementia gets seriously ill, but to articulate a way of thinking about the care-giver's many dilemmas, as well as the dilemmas faced by everyone who might one day suffer dementia.  The argument is that while certain ethical boundaries limit the realm of choice, care-giving decisions are largely matters of loving prudence, where the aim is best care for the person with dementia, in light of myriad ethical, social, medical, familial, and existential realities. But the fact of choice does not mean that all choices are equally justifiable. And the care-giver's obligation is to discern the good - and not simply live one's out preferences - in circumstances where seeing the good is often very hard.

Yet having asked so many questions, it might seem evasive not to offer some concrete, if provisional, answers, given here simply as examples of how prudence within limits might operate. The following suggestions are presented not as the opinion of the staff as a whole or of the Council, but with a view to spurring discussion about possible guidelines in this difficult area:

(1) One boundary seems absolutely inviolable: killing the demented or helping the demented to kill themselves is always wrong, certainly as a matter of public policy. Doing so would deny the obligation of the well to care for the sick, the duty of the strong to care for the weak, and the responsibility to choose life and let death choose us. It would be foolish to call assisted suicide an easy way out, since the temptation to engage in this practice arises most often in cases where much suffering and heartache has already been endured. There may even be selfless motives for choosing death, such as sparing our children long years of misery as care-givers. And there may be seemingly compassionate motives for giving a suffering loved one a final dose of relief. But in the end, it is highly doubtful that aiming at death is a right we truly have. And it seems more likely that suicide or assisted suicide will burden survivors with guilt rather than grant them a merciful reprieve.

(2) But the dilemmas of caring for persons with dementia when they get sick are more complicated than decisions about assisted suicide; they are decisions about care. And the long-standing categories of the end-of-life debate may be inadequate for thinking about the problems of extended decline and dependence, where the person with dementia can spend years with no self-recognition of the self she once was. Surely, this new reality requires an acceptance of being depended upon by others for care-giving decisions and an acceptance of depending on others for care-giving decisions. It should dispel any illusions of being autonomous until the very end.

(3) We should have little trouble overriding an advance directive if we believed doing so would benefit the person entrusted in our care. (As an aside: I cannot imagine ever having an advance directive myself, since I would hope, perhaps foolishly, that a lifetime of being in a family has prepared my loved ones for making decisions on my behalf - decisions I am perfectly willing to leave to them.)

(4) One can see the nobility in refusing treatment for a life-threatening disease that strikes very early in one's dementia, when a person is still aware of what is coming and might choose to accept death as the cost of living a certain kind of life until the very end. But the noble deed may ultimately miss its mark: In the name of freeing loved ones from burdens, it might burden loved ones with inescapable questions about their family member's lack of trust in them. And in an effort to choose death with dignity, it might reveal a lack of courage in the face of un-chosen decline. That said, the ethical case for forgoing treatment for the sake of nobility appears to strengthen as the disease progresses, and it may be that a loving spouse has just as much authority to decide when enough is enough as the person himself before the full onset of dementia.

(5) The happiness of the person with dementia is a legitimate factor in deciding how aggressively to treat, and the best care for demented persons who are violent or depressed may mean intervening less aggressively (including accepting death) when illness strikes.  But one must also worry that certain treatments aimed at extending the lives of demented persons who are happy - such as heart surgery for an individual with life-threatening blockage in their coronary arteries - might end up destroying the very happiness it aims to extend; and that a few weeks or months of happy dementia followed by a heart attack is better than a painful operation that leads to a longer life of endless misery due to the trauma of the treatment itself. Yet happiness is not a fixed attribute, and that the unhappiness of persons with dementia may sometimes be due more to our betrayal than their own unavoidable despair

(6) We can imagine a care-giving adult son with a special daughter - either specially gifted or demanding special attention due to a disability - taking his own daughter's well-being into account when deciding how aggressively to treat his parent with moderate or severe dementia. The flourishing of the young daughter should not be the reason for forgoing treatment, but her well-being might rightly tip the scales in hard cases - without the care-giver ever explicitly saying so.

(7) We might well support setting limits on the types of treatments covered by Medicare for persons suffering severe dementia. But we should also protect the right of families to seek such treatments for themselves, if they believe there are sound reasons for doing so, and if doing so does not maliciously burden the sick person in the name of extending their life. At the same time, we recognize that such a policy creates a new problem, giving the wealthy (who don't need Medicare) more choices than the poor (who rely entirely on Medicare).

(8) Finally, on the question of whether dementia matters in deciding how to care for persons with dementia, we are uncertain. In short: Should we regard a person who is physically dependent for the activities of life differently if they are also mentally unaware? The presence of the body is too real to deny the presence of the person. But the loss of the soul, the memory, and the possibility of mutual recognition - this is too real to deny the loss of something profound about the demented self.  Whether this loss means the person is more ready to die or whether it means that the care-giver has a special obligation to care (an obligation to dance both parts of the dance), we are torn.

This list of conclusions is hardly exhaustive, and such opinions cannot be properly considered in the abstract. If anything, however, they suggest the difficult judgments we face in an era of mixed blessings and new burdens, where loving one's elders and dying with dignity may be harder than ever.


1. Of course, this analysis does not begin in a vacuum: There is significant case law governing what individuals, families, and doctors can and cannot do; there are hospital policies and established protocols; and families and doctors are making such life-and-death decisions all the time, relying on a mix of procedural guideposts and human intuitions to muddle through, trying to balance myriad pressures, feelings, and values to do what is right.
2. These concrete dilemmas point to larger questions: Is the choice for non-treatment that leads to death a "right," properly understood?  Or is there sometimes an obligation to seek treatment, both as a patient and for a loved one? Is the rejection of life-sustaining treatment by nature a choice for death? Or might it be a choice for living a certain kind of life, even to the very end? Or a choice to die a more peaceful death? Or a choice to avoid yet more days, weeks, months, or even years of suffering? It is these kinds of questions that we consider more fully below, with a particular focus on caring for demented person.
3. Of course, the above claim about the limits of autonomy is one of the deepest and most disputed questions in bioethics, and we cannot do justice to the complexities here. For an argument against the right to die and in favor of ethical limits to autonomy, see Leon Kass, "Is There a Right to Die?," Hastings Center Report, January-February 1993.
4. The claims that follow could no doubt be disputed. There is one state (Oregon) that explicitly allows assisted suicide, thus allowing aiming to kill with a compassionate or merciful intent. But such states and such views still seem to be in the minority.
5. Treating the demented person differently seems to require accepting one of the following propositions: (a) that human dignity resides foremost in our sustained mental capacities, and that a person with severe dementia is no longer a self, but a body without a self; (b) that the person's wishes before becoming incompetent should always be respected, 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 happy one; or (c) that the experience of becoming demented is so horrible - different in kind from being incompetent throughout life - that the loss of one's powers justifies non-treatment. In short, to treat a demented person differently than the dependent person who is not demented, one must believe that dementia inflicts a special kind of suffering, such that it would be wrong to extend it; or that a demented person lacks dignity not simply because of his dependence on others for everyday needs, but because of his changed mental 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.

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