A Consensus-Based Approach To Providing Palliative Care to
Patients Who Lack Decision-Making Capacity
Jason H.T. Karlawish, MD; Timothy Quill, MD; and Diane E. Meier, MD, for the ACP–ASIM
End-of-Life Care Consensus Panel
Making palliative care decisions for a patient who lacks
decision-making capacity presents several challenges.
Other people, such as family and caregivers, must choose
for the patient. The goals and values of these decision
makers may conflict with those of each other and with
those of the patient, who now lacks the capacity to partic-
ipate in the decision. This paper presents a case study of a
patient with severe Alzheimer disease who has two com-
mon clinical problems: neurogenic dysphagia and aspira-
tion pneumonia. The case study describes a consensus-
based decision-making strategy that keeps what is known
about the patient’s wishes and values in the foreground
but also expects guidance from the physician and elicits
input from family members and other people who care for
and have knowledge about the patient. The steps of this
process, including key clinical prompts and potential tran-
sition statements, are outlined and described. The overall
goal of the case commentary is to demonstrate that phy-
sicians can guide a highly emotional and personal process
in a structured manner that has meaning for the patient,
family, physician, and other caregivers.
This paper is also available at http://www.acponline.org.
Ann Intern Med. 1999;130:835-840.
This paper was written by Jason H.T. Karlawish, MD; Timothy
Quill, MD; and Diane M. Meier, MD, and was developed for
the American College of Physicians–American Society of Inter-
nal Medicine (ACP–ASIM) End-of-Life Care Consensus Panel.
Members of the ACP–ASIM End-of-Life Care Consensus Panel
were Bernard Lo, MD (Chair); Janet Abrahm, MD; Susan Block,
MD; William Breitbart, MD; Ira R. Byock, MD; Kathy Faber-
Langendoen, MD; Lloyd W. Kitchens Jr., MD; Paul Lanken,
MD; Joanne Lynn, MD; Diane Meier, MD; Timothy Quill, MD;
George Thibault, MD; and James Tulsky, MD. Primary staff to
the Panel were Lois Snyder, JD (Project Director); Jason Karla-
wish, MD; and Karine Morin, LLM. This paper was reviewed
and approved by the Ethics and Human Rights Committee and
the Education Committee. Members of the Ethics and Human
Rights Committee were Risa Lavizzo-Mourey, MD (Chair);
Joanne Lynn, MD; Richard J. Carroll, MD; David A. Fleming,
MD; Steven H. Miles, MD; Gail J. Povar, MD; Susan H. Thomp-
son, MD; James Tulsky, MD; Alan L. Gordon, MD; Siang Y.
Tan, MD; and Lee J. Dunn Jr., LLM. Members of the Education
Committee were Faroque A. Kahn, MD (Chair); Michael A.
Ainsworth, MD; John B. Bass, MD; John R. Feussner, MD;
Donald E. Girard, MD; John J. Hoesing, MD; Faith T. Fitzger-
ald, MD; Alphonso Brown, MD; Jerome H. Carter, MD; Sandra
Adamson Fryhofer, MD; William J. Hall, MD; Rodney Horn-
bake, MD; Christine S. Hunter, MD; Mary E. Moore, MD; and
Kurt Kroenke, MD. This paper was approved by the Board of
Regents on 12 July 1998.
For current author addresses, see end of text.
As we grow older the world becomes stranger, the pattern more
complicated of dead and living.
–T.S. Eliot, “East Coker”
and has a husband and daughter. For the past 3 years
she has lived in a nursing home and requires assis-
tance with all of her basic activities of daily living. Mr.
B. visits her daily and feeds her lunch. In the past
several weeks, she has taken longer to finish small
portions. At times, she coughs while being fed. One
morning after breakfast, Mrs. B. develops agitation, a
cough, and a temperature of 100.3 °F.
Palliative care is designed to relieve a patient’s
suffering in order to maximize dignity and quality of
life (1). Respect for a patient’s autonomous choices
is a foundation of western bioethics, but a patient
with severe dementia can no longer decide whether
to receive predominantly palliative care or to con-
tinue potentially life-prolonging therapy. Other peo-
ple, such as family members and caregivers, must
choose for the patient. The goals and values of
these decision makers may conflict with those of
each other and with those of the patient, who now
lacks the capacity to participate in the decision.
Conflicts are especially likely in decision making
about two common clinical problems: aspiration
pneumonia and neurogenic dysphagia. These prob-
lems engage deep values about feeding, starvation,
and the meaning of care for a patient with severe
dementia (2). Furthermore, a patient’s residence in
a nursing home introduces additional problems re-
lated to regulations governing the management of a
patient’s weight loss and nutrition (3). How can a
busy internist address these problems in a manner
that achieves consensus among decision makers?
This case study illustrates an approach to develop-
ing palliative care plans for patients with severe
At a meeting with Mr. B. and the daughter, the
physician explains that aspiration pneumonia is a
common problem in severe stages of dementia caused
by irreversible and progressive loss of the ability to
chew and swallow food. Mr. B. agrees that his wife
has had trouble eating, and he describes how on some
days she eats very little of even her favorite foods.
To prompt the family to tell more about their per-
ception of how Mrs. B. has changed, the physician
asks, “I know that Mrs. B. was diagnosed with Alz-
rs. B. is a 73-year-old woman with severe Alz-
heimer disease. She is a retired schoolteacher
©1999 American College of Physicians–American Society of Internal Medicine
heimer’s over 7 years ago, but I have only known her
for the last few months. Can you tell me how she
seems to you now, and how things have changed
compared to when she was first admitted?”
The family’s story begins with a summer vacation
cut short when Mrs. B. fell and was hospitalized for 10
days. She never fully recovered and subsequently ex-
perienced a progressive loss of function over the next 7
years. The physician responds to the story by saying, “I
think I have a better understanding about how things
have changed over the past few years. It sounds like
both you and she have done the best you could
through a difficult situation. You know that Mrs. B.
has an incurable, progressive, and ultimately fatal dis-
ease. I can’t say for sure when she’ll die of her Alz-
heimer’s disease, but given its severity, we shouldn’t be
surprised when she does. Even if she does recover
from this pneumonia, she will not recover her swal-
lowing function. Recognizing this, we ought to care for
her in a way that makes us confident that after she’s
gone, we can say she was treated with dignity and
“I use two principles to help think through the
decisions we face. First, consider your understanding
of what Mrs. B. would want if she could tell us.
Second, we should balance the burdens and benefits
of each option in terms of its ability to relieve her
suffering and maximize her dignity and the quality of
her remaining life.”
This narrative begins an emotional and meaning-
ful dialogue that should lead to a consensus about
the best way to care for Mrs. B. given her clinical
circumstances. This consensus-based approach is
distinct from a discussion in which individuals sim-
ply share their opinions, but there is no genuine
attempt to reconcile their differences. The process
and pitfalls of consensus building will vary depend-
ing on the patient’s clinical condition, the family
dynamics, and the level of preexisting trust or con-
flict in the physician–patient–family relationships.
Nonetheless, the physician can guide this dialogue
by using goals and steps that are outlined in Tables
1 and 2 and are discussed below.
The physician’s first step was to identify potential
decision makers. If Mrs. B. had completed an ad-
vance directive that included a durable power of
attorney for health care, that person would be her
main representative in decision making. However,
most patients in nursing homes do not have advance
directives (4), and even when they do, the contribu-
tion of others, such as family and formal caregivers,
can greatly assist the designated surrogate in deci-
sion making (5, 6). Even if the patient doesn’t have
an advance directive, the physician can turn to close
family members and others who know the patient
well with the assurance that the family’s standing to
serve as surrogate decision makers is established by
both case and statutory law (7). Except in cases in
which the patient has no family or has a family that
does not represent the patient’s best interests, the
physician is under no moral or legal authority to
seek a guardian (a forthcoming paper from the
American College of Physicians–American Society
of Internal Medicine End-of-Life Care Consensus
Panel will focus on legal issues) (7, 8).
After identifying the decision makers and clarify-
ing Mrs. B.’s diagnoses, the physician then encour-
aged them to describe the course of Mrs. B.’s de-
mentia by asking the husband and daughter how
Mrs. B. has changed. The purpose of this dialogue
is to achieve a consensus among decision makers
about the patient’s current disease state, prognosis,
quality of life, and previously stated values (5). This
The Core Principles with Which To Plan
Palliative Care for Patients Who Lack
Structure the decision making as a consensus-building process grounded in
dialogue among proxy, other close family members, physician, and
The goal is to achieve consensus about diagnosis and prognosis; the
benefits and burdens of different treatment options; and the meaning of
emotionally charged terms, such as “starvation,” “suffering,” “quality of
life,” “feeding,” and “dying.”
Decisions should be based on: 1) the patient’s preferences and 2) a balance
of the burdens and benefits of each option in terms of its ability to relieve
suffering and maximize dignity and quality of life.
Palliative care should be offered whether life-prolonging measures are
initiated or the patient is treated with comfort measures only.
Steps to Providing Palliative Care to Patients
Who Lack Decision-Making Capacity
Step Representative Quote To Achieve Step
1. Identify the main
participants in the decision
“We need to make some decisions about
the care of your wife. Is everyone here
who could help us think through what
we should do?”
“Can you tell me how she’s changed,
how things have gone for all of you?”
2. A. Allow the participants
to narrate how the patient
has come to this stage of
B. In cases where the
physician has an extended
relationship with the
patient and family
3. Teach the decision makers
about the expected clinical
course of the patient’s
“I know I’ve been caring for your wife for
many years, but it helps me if you can
tell me how she’s changed, how things
have gone for each of you?”
“Your wife has an incurable, progressive,
and ultimately fatal disease. I can’t say
for sure when she’ll die of her
Alzheimer disease, but given its
severity, we shouldn’t be surprised
when she does.”
“We ought to care for her in a way that
makes us confident that after she’s
gone, we can say she was treated with
dignity and respect.”
“For patients like Mrs. B., feeding with a
tube does not significantly reduce the
risk for pneumonia. On the basis of my
experience, a speech therapist can give
us some useful hints on ways to feed
her that will allow her to continue to
eat by mouth.”
4. Advocate for the patient’s
quality of life and dignity
5. Provide guidance on the
basis of existing data and
18 May 1999
• Annals of Internal Medicine
• Volume 130
• Number 10
sharing of narratives may expose important differ-
ences in beliefs and understandings in any of these
domains that must be reconciled before consensus-
based decision making can proceed. The more the
physician understands these various perceptions, the
more likely he or she is to develop a plan that will
respect Mrs. B. as a person. Even when the physi-
cian has a long-term relationship with the family
and patient, this step in consensus building should
not be skipped. Research showing that physicians
often inadequately understand their patients’ pref-
erences for health care supports this point (9–11).
The next step in consensus building is to begin a
dialogue about prognosis and about the potential
role of palliative care. This was done with the
phrase that began “Unfortunately, Mrs. B. has an
incurable, progressive, and ultimately fatal dis-
ease. . . .” Although this physician believes strongly
in a palliative approach that includes pain and
symptom management and avoidance of invasive
treatments for patients with severe Alzheimer dis-
ease, he must respect that others may value an
approach in which available medical technology
must always be used to prolong life. Before a con-
sensus about what might be included in a palliative
approach for a particular patient is possible, the
physician must learn what “treated with dignity and
respect” means in this family.
Finally, the physician guided the family on ethical
standards for decision making: 1) Consider what is
known of the patient’s wishes and preferences given
her current condition (for example, a living will or
potentially relevant statements made when the pa-
tient was competent) and 2) balance the burdens
and benefits of each option in terms of its ability to
maximize Mrs. B.’s dignity and quality of life. De-
ciding how to care for Mrs. B. by using only a rigid
understanding of her past preferences may fail to
respect her present circumstances (12). Achieving
consensus about her current quality of life, although
a subjective and personal process, incorporates the
family’s, physician’s, and health care team’s percep-
tions of her circumstances into what is known about
her past preferences and values.
Missing from these recommendations to the fam-
ily is an exclusive appeal to futility as grounds for
decision making. Futility refers to the claim that no
desirable benefit can be achieved by potentially life-
prolonging treatment (13). Physicians frequently cite
futility as their reason for terminating further treat-
ment (14). Although the concept of “medical futil-
ity” exists to communicate extremely poor progno-
sis, it can also inadvertently convey an unequivocal,
unilateral, and negative judgment about a patient’s
quality of life without leading to an explicit discus-
sion of these issues from the differing perspectives
of physicians and families. Furthermore, too fre-
quently, the term connotes that “nothing more can
be done” for a patient, that further intervention
would be meaningless, or that her life is of no
current value. Therefore, a physician who relies ex-
clusively on futility as a reason to pursue palliative
care can obscure an honest discussion of how peo-
ple understand and value the patient’s continued
existence and of the range of possibilities for pal-
liative and life-prolonging interventions.
Mr. B. and the daughter agree that Mrs. B. never
expressed clear preferences about how she should be
treated. The physician reassures them that they can
still work on a plan. He outlines the options, including
transfer to the hospital or staying at the nursing home
with or without antibiotic therapy, and explains his
view of their benefits and burdens: “A helpful way to
think through these choices is to come to some con-
sensus upon her current quality of life, and then de-
cide what options will best maximize it.”
In the ensuing discussion, Mrs. B.’s daughter and
the physician state that they feel that Mrs. B.’s quality
of life is poor because she cannot communicate or
move around, but the husband disagrees. “I know she
still hears me and understands me. I can see it in her
eyes, when I bring her food from home.” The daughter
begins to cry and says that her mother would not want
to live the way she is. Again, the husband disagrees:
“You’re not there every day! I am!” This common
conflict challenges the consensus-building process.
The physician looks for common ground by asking
them: “If she gets worse, if she can’t recognize you or
starts to suffer more, do we all agree that we should
focus purely on her comfort and that even antibiotics
would be too invasive?” The husband starts to cry and
is comforted by the daughter. They both nod in agree-
The physician then proposes a compromise plan.
“Mr. B., your visits are important. I recommend that
we keep her here at the nursing home where you can
visit her as much as possible; we can simultaneously
provide antibiotics and try our best to keep her com-
fortable. If she deteriorates, our focus can shift exclu-
sively to relieving her symptoms and minimizing her
immediate suffering. I strongly recommend against
providing cardiopulmonary resuscitation should she
have a cardiac arrest, since I don’t think it would
work and it would only add to her suffering.”
The family agrees that the plan strikes a proper
balance between benefits and burdens, giving her a
chance of recovery without subjecting her to a foreign
environment or overly harsh treatments.
This decision-making process exposed two com-
mon features of caring for patients with severe de-
mentia. First, clear information about the patient’s
wishes is typically unavailable. Second, decision
makers often have differing assessments of the pa-
tient’s preferences and quality of life (15). In this
18 May 1999
• Annals of Internal Medicine
• Volume 130
• Number 10
case, the daughter and the physician thought that
Mrs. B. had a poor quality of life and were con-
cerned that hospitalization and even antibiotics
might further decrease her quality of life. The hus-
band disagreed. They all were genuinely trying to
act in Mrs. B.’s best interests without clear informa-
tion about her preferences. They achieved compro-
mise with the decision to give Mrs. B. a therapeutic
trial of antibiotics at the nursing home.
Dialogue is essential for achieving consensus on a
course of action that is responsive to both past and
present patient realities, as well as to the concerns
and priorities of the family (5, 6, 16). The goal of
dialogue is not to provoke conflict but to clarify
common ground and differences and lead to better
appreciation of the meaning of the decision for the
patient and her family (5, 17). This approach to
decision making is grounded in narrative theory that
unifies the clinical and moral dimensions of medi-
cine (18, 19). Clinical medicine is grounded in a
series of stories told and interpreted from a variety
of perspectives. The physician usually interprets
these stories, using the science of clinical medicine,
to develop a diagnostic and therapeutic plan. These
same stories can be simultaneously used to under-
stand the patient’s values, goals, and meanings of
illness, which should guide the personal and moral
sides of the same process. Narrative theory was put
into practice when the physician prompted Mr. B.
and his daughter to describe their perception of
Mrs. B.’s current condition, as well as how she has
changed over recent years. The physician also
shared his own perceptions. Reconciling the percep-
tions conveyed in these stories decisively shaped the
process of medical decision making on Mrs. B.’s
This theory has some limitations. Consensus oc-
curs in the context of choices. However, in the care
of patients with severe dementia who live in nursing
homes, local customs, beliefs, and systems of care
can limit reasonable choices. For example, long-
term care regulations are often wrongly believed to
require that all residents with neurogenic dysphagia
receive artificial nutrition and hydration. Surrogate
decision-making laws are often misinterpreted to
require a legally designated guardian for noncom-
petent patients who lack an advance directive (3).
Although a few state laws (in Missouri and New
York) require a high degree of proof of a surrogate
decision maker’s knowledge of a patient’s wishes to
allow the surrogate to withhold or withdraw artifi-
cial hydration or nutrition from noncompetent pa-
tients, most states support the legal right of surro-
gate decision makers to refuse any and all unwanted
medical treatments when this decision is based on a
consensus of the patient’s wishes and best interests
(3). In addition to legal inconsistencies and misper-
ceptions, both health care systems and local com-
munity practices powerfully influence choices and
decisions. For example, in the same community, two
otherwise high-quality nursing homes may have dra-
matically different rates of enteral feeding for pa-
tients with severe dementia. Research shows that
large national variations in the rates of dying at
home correlate with regional inpatient bed availabil-
ity, not with patient or family preferences (20).
An additional practical concern of a busy inter-
nist is that these dialogues take time. No empirical
data compare the time requirements of this method
with those of other decision-making strategies. But
the investments in mutual understanding and trust
building should ultimately improve decision making,
promote higher-quality care, and prevent conflict as
the patient’s illness progresses.
“What happens if I feed her?” asks Mr. B. “Isn’t
she going to choke or get a worse pneumonia? Should
we feed her by a tube or in the vein?”
Mr. B.’s questions describe a typical decision-
making cascade. Because oral feeding is thought to
put the patient at risk for aspiration pneumonia, it
is stopped, and mechanical feeding, either intrave-
nously or enterally, is started so that the patient
does not aspirate or starve.
In addition to exploring the benefits and burdens
of these technical interventions, the physician’s duty
is to teach the family about palliative care options
for patients with neurogenic dysphagia caused by
severe dementia who develop aspiration pneumonia.
The decision to use or not to use a feeding tube is
not a prima facie determinant of the quality of
palliative care for a patient with severe dementia.
Although no randomized comparison of oral versus
enteral feeding has been reported, the available
studies suggest that enteral feeding does not signif-
icantly reduce, and may even increase, the risk for
aspiration pneumonia (21, 22). In addition, enteral
feeding may not prevent weight loss or the progres-
sion of pressure ulcers (23, 24) and is associated
with substantial 1-year mortality rates (25). Finally,
a longitudinal cohort study (26) suggests that neu-
rogenic dysphagia and common nutritional compli-
cations of severe dementia can be managed by care-
ful oral feeding. Given these data, the decision to
use a feeding tube is best viewed as a personal
choice based on deep values about the meaning of
using a feeding tube or of continued oral feeding.
The physician explains that a feeding tube will al-
low the delivery of adequate nutrition and hydration
for a woman of Mrs. B.’s age and weight, but it may
not improve the quality or length of her life or prevent
further aspiration. It may also deprive her of the tastes,
smells, and touches of normal eating. He appeals to
the family to recall how they just made the decision to
care for Mrs. B.’s pneumonia. The key issue was the
18 May 1999
• Annals of Internal Medicine
• Volume 130
• Number 10
pleasure Mrs. B. received from her husband’s daily
visits and the food that he gave her.
The husband becomes agitated. “I just can’t starve
her, watch her starve, if that’s what you’re saying.” He
begins to weep.
“But Dad, she eats what she wants. No one’s saying
don’t feed her. It’s just don’t force-feed her.”
“But how will we know she’s hungry?” he asks.
The physician offers a suggestion. “Her eating and
your feeding her is one of the few meaningful activities
that she has left. I think you both agree that we ought
to at least allow her to try some food by mouth.
Perhaps a speech therapist can give you some useful
hints on feeding your wife. The issue of how best to
feed your wife doesn’t need to be made now. Let’s
take a few days and see how she does with both the
pneumonia and the feeding. In between, don’t hesitate
to call me with any questions.”
The dialogue has again reached a point of con-
flict. Table 3 lists some general ways to achieve
common ground. In this case, the physician presents
information about some of the burdens of enteral
feeding, and some of the benefits of continued oral
feeding that had not been considered by the family.
He then proposes a short-term trial of oral feeding,
with the help of a speech therapist, while simulta-
neously not precluding the possibility of enteral
feeding in the future. In negotiating terms, he has
proposed a compromise solution that avoids a
power conflict and allows everyone’s interests to be
Two days later, the husband and daughter return.
The husband explains, “I met with that therapist and
she showed me how to get her to eat a whole con-
tainer of sherbet that she likes. We talked with our
minister. I think we’re going to keep it natural like that
Mrs. B. gradually recovers from the pneumonia
after several days that entail considerable physical dis-
tress (cough, respiratory distress, fever, and agitation).
In addition to the oral antibiotics, her symptoms were
palliated with opioids, humidified oxygen, nebulizers,
antipyretic agents, and low-dose antipsychotic agents.
Several months pass. Mrs. B. is now unable to use
a straw. Her intake of spoon-fed fluids and food is
scant. Her husband worries that she will die of star-
vation. He and the daughter reapproach the physician
for advice about how to proceed.
The physician, Mr. B., Mrs. B.’s daughter, and their
minister review Mrs. B.’s life and the recent events.
After some discussion the physician says: “Her Alzhei-
mer disease has progressed to the point where she’s
dying. I believe we ought to come up with a plan that
minimizes her immediate suffering and maximizes her
dignity and her quality of life.”
A vigorous discussion follows. Eating was the one
meaningful activity that remained in Mrs. B.’s life, but
artificial feeding cannot serve the same function. With-
out that activity, the husband now believes that her
quality of life is too poor to warrant the burdens of a
feeding tube. Their minister agrees that a feeding tube
would only prolong her suffering at this point. A plan
is agreed upon to provide comfort measures only. She
will be offered tastes and smells of her favorite foods
and drinks; mouth and skin hygiene; lots of human
contact, including repositioning and massage; but no
enteral or intravenous nutrition or hydration.
After 7 days, Mrs. B. dies.
This case presents several common challenges to
the practice of palliative care for patients with se-
vere dementia. Mrs. B. could not speak for herself
and, like most Americans, had not completed an
advance directive. Even if she had, the preferences
she expressed when competent may well have been
indeterminate guides for managing her actual prob-
lems. She could not tell us whether she was suffer-
ing or describe her quality of life. The meaning of
her signs and symptoms had to be interpreted. Fur-
thermore, as a wife, mother, patient, and resident of
a nursing home, she lived in a diverse community
that had different views about what ought to be
done for her. Finally, her problems included demen-
tia complicated by aspiration pneumonia and severe
dysphagia. No randomized, controlled studies guide
the provision of oral versus enteral feeding under
these circumstances (21–24). Despite this lack of
evidence, those who care for and care about these
patients must make decisions.
This case illustrates a palliative care strategy
grounded in the theory that decisions for patients
such as Mrs. B. are the result of dialogue and
consensus building. The physician’s initial invest-
ment of time may minimize the time and effort
needed for future decisions. The physician’s duty is
to teach all participants that Mrs. B. has a chronic,
irreversible, and ultimately fatal disease but also to
learn from these participants about Mrs. B.’s values
and quality of life. This frames decisions about hos-
pitalization, antibiotics, and enteral nutritional sup-
What To Do If No Consensus Emerges
Postpone the decision making and recommend that the participants take
time to think about and discuss key issues.
Understand and separate from each person’s perspective the goals of
medical care and the treatment choices to achieve these goals.
Invent new solutions (for example, a time-limited trial rather than an
Avoid power struggles or personalizing the conflict.
Call in a third party (for example, trusted clergy, ethics consult, or palliative
Don’t violate fundamental values of the patient, family, or physician.
18 May 1999
• Annals of Internal Medicine
• Volume 130
• Number 10
port as medical choices that ultimately shape the Download full-text
way she will live the last phase of her life.
As a result of this dialogue, a family may decide
that continued efforts to prolong life are critical
regardless of the severity of the patient’s disease.
Principles of negotiation are often useful under
these circumstances (Table 3). In general, physicians
should guide the process because of their familiarity
with medical processes and prognoses, but the sur-
rogate decision makers must try to represent the
patient’s voice. Differences should be explored
through dialogue that focuses on the patient’s best
interests and seeks common ground. Except when
decisions seem to clearly violate the patient’s best
interests or prior wishes, the family has the final say
in representing the patient in decision making. Fam-
ilies have to live with themselves and their role in
these decisions long after the patient has died.
Conversely, some medical practitioners or long-
term care institutions see their primary job as pro-
longing life under all clinical circumstances because
of religious principles or personal training. These
practitioners or institutions should make their phi-
losophy known from the outset, especially if they
feel obligated to override the values and wishes of
patients and families, because they will be unable to
pursue this consensus-based approach.
The meaning of suffering is personal and subjec-
tive. Evaluating the suffering of patients with de-
mentia who cannot speak for themselves is inher-
ently challenging. Even when suffering is recognized,
its relief may be relegated in the pursuit of another
goal, such as the preservation of life at all costs or
hope for a miraculous cure. Yet through the process
of repeatedly listening to the perspectives of each
participant and involving the participants in a con-
sensus-based interaction, decisions that respect the
patient’s dignity and quality of life can generally be
Acknowledgment: The American College of Physicians–American
Society of Internal Medicine thanks the Greenwall Foundation
for its support of the development of this paper.
Grant Support: Dr. Karlawish is supported by a Brookdale Foun-
dation National Fellowship and National Institute on Aging grant
1K01-AG00931-01. Dr. Meier is an Open Society Institute Fac-
ulty Scholar of the Project on Death in America.
Requests for Reprints: Lois Snyder, JD, Center for Ethics and
Professionalism, American College of Physicians–American Soci-
ety of Internal Medicine, 190 N. Independence Mall West, Phil-
adelphia, PA 19106.
Current Author Addresses: Dr. Karlawish: Institute on Aging, Uni-
versity of Pennsylvania, 3615 Chestnut Street, Philadelphia, PA
Dr. Quill: The University of Rochester and the Genesee Hospi-
tal, Department of Medicine, 224 Alexander Street, Rochester,
Dr. Meier: Departments of Geriatrics and Medicine, Mount Si-
nai School of Medicine, 1 Gustave L. Levy Place, Box 1070, New
York, NY 10029.
1. Doyle D, Hanks GW, MacDonald N. Oxford Textbook of Palliative Care. 2d
ed. New York: Oxford Univ Pr; 1998.
2. Lo B, Dornbrand L. Guiding the hand that feeds. Caring for the demented
elderly. N Engl J Med. 1984;311:402-4.
3. Meisel A. Barriers to forgoing nutrition and hydration in nursing homes. Am J
Law Med. 1995;21:335-82.
4. Janofsky JS, Rovner BW. Prevalence of advance directives and guardianship
in nursing home patients. J Geriatr Psychiatry Neurol. 1993;6:214-6.
5. Kuczewski MG. Reconceiving the family. The process of consent in medical
decisionmaking. Hastings Cent Rep. 1996;26:30-7.
6. Brock DW. What is the moral authority of family members to act as surro-
gates for incompetent patients? Milbank Q. 1996;74:599-618.
7. Meisel A. The Right to Die. 2d ed. New York: Wiley; 1996.
8. Buchanan A, Brock DW. Deciding for Others: The Ethics of Surrogate De-
cision Making. Cambridge: Cambridge Univ Pr; 1989.
9. Fischer G, Tulsky JA, Rose MR, Siminoff LA, Arnold RM. Patient knowl-
edge and physician predictions of treatment preferences after discussion of
advance directives. J Gen Intern Med. 1998;13:447-54.
10. Virmani J, Schneiderman LJ, Kaplan RM. Relationship of advance direc-
tives to physician-patient communication. Arch Intern Med. 1994;154:909-
11. A controlled trial to improve care for seriously ill hospitalized patients. The
Study to Understand Prognoses and Preferences for Outcomes and Risks of
Treatments (SUPPORT). The SUPPORT Principal Investigators. JAMA. 1995;
12. Dresser R. Dworkin on dementia. Elegant theory, questionable policy. Hast-
ings Cent Rep. 1995;25:32-8.
13. Schneiderman LJ, Jecker NS, Jonsen AR. Medical futility: its meaning and
ethical implications. Ann Intern Med. 1990;112:949-54.
14. Prendergast TJ, Luce JM. Increasing incidence of withholding and with-
drawal of life support from the critically ill. Am J Respir Crit Care Med.
15. Sulmasy DP, Terry PB, Weisman CS, Miller DJ, Stallings RY, et al. The
accuracy of substituted judgments in patients with terminal diagnoses. Ann
Intern Med. 1998;128:621-9.
16. Hurley AC, Volicer L, Rempusheski VF, Fry ST. Reaching consensus: the
process of recommending treatment decisions for Alzheimer’s patients. ANS
Adv Nurs Sci. 1995;18:33-43.
17. Quill TE, Brody H. Physician recommendations and patient autonomy: find-
ing a balance between physician power and patient choice. Ann Intern Med.
18. Bruner J. Acts of Meaning. Cambridge, MA: Harvard Univ Pr; 1990.
19. Hunter KM. Doctor’s Stories: The Narrative Structure of Medical Knowledge.
Princeton, NJ: Princeton Univ Pr; 1991.
20. Pritchard RS, Fisher ES, Teno JM, Sharp SM, Reding DJ, Knaus WA, et
al. Influence of patient preferences and local health system characteristics on
the place of death. SUPPORT Investigators. Study to Understand Prognoses
and Preferences for Risks and Outcomes of Treatment. J Am Geriatr Soc.
21. Ahronheim JC. Nutrition and hydration in the terminal patient. Clin Geriatr
22. Finucane TE, Bynum JP. Use of tube feeding to prevent aspiration pneu-
monia. Lancet. 1996;348:1421-4.
23. Henderson CT, Trumbore LS, Mobarhan S, Benya R, Miles T. Prolonged
tube feeding in long-term care: nutritional status and clinical outcomes. J Am
Coll Nutr. 1992;11:309-25.
24. Finucane TE. Malnutrition, tube feeding and pressure sores: data are incom-
plete. J Am Geriatr Soc. 1995;43:447-51.
25. Grant MD, Rudberg MA, Brody JA. Gastrostomy placement and mortality
among hospitalized Medicare beneficiaries. JAMA. 1998;279:1973-6.
26. Volicer L, Seltzer B, Rheaume Y, Karner J, Glennon M, Riley ME, et al.
Eating difficulties in patients with probable dementia of the Alzheimer type. J
Geriatr Psychiatry Neurol. 1989;2:188-95.
18 May 1999
• Annals of Internal Medicine
• Volume 130
• Number 10