Intensive Care Med (2000) 26: 1414±1420

DOI 10.1007/s001340000641 RE VIEW

D. J. Nyman End-of-life decision making

C. L. Sprung
in the intensive care unit

Received: 16 July 1999

comes being reported [4]. By the late 1970s/early 1980s,
Accepted: 8 July 2000 attempts at CPR had become standard practice in hospi-
Published online: 14 September 2000 talised patients suffering a cardiorespiratory arrest, irre-
 Springer-Verlag 2000 spective of their underlying pathology [4]. Unfortunate-
ly, only a small percentage (0±25 %) of patients having
)
D. J. Nyman ( )
Department of Anesthesiology
in-hospital cardiac arrest survive to leave hospital [4, 5,
6, 7]. Long-term survival was estimated at 15 % [8].
and Intensive Care, CPR is particularly unsuccessful in the elderly and in
Sha'are Tzedek Medical Center, those with diseases such as pneumonia, metastatic can-
POB 3235, Jerusalem 91031, Israel
E-mail: kaldeb@netvision.net.il
cer or multiple organ failure [9]. CPR is highly invasive,
Phone: +9 72-2-6 55 51 04 with great potential for causing harm [5]. If a viable car-
Fax: +9 72-2-6 66 60 03 diac rhythm is restored, patients may spend the last days
or weeks of their life in the ICU with depressed neuro-
C. L. Sprung logical function [9], undergoing painful and dehumanis-
Department of Anesthesiology ing procedures with little medical benefit [4]. Physicians
and Critical Care Medicine,
Hadassah Hebrew University
have come to realise that CPR is not necessarily in the
Medical Center best interests of all patients undergoing cardiac arrest
[4]; physicians, patients, and their families must all
come to terms with the inability of medicine to postpone
death indefinitely [4].
Twenty-five years ago most patients died in an ICU
Introduction
after undergoing CPR [10]. In 1990, 83 % of European
As medical technology advances, patients with chronic respondents withheld and 63 % withdrew life support
or terminal diseases, and their physicians, are confront- [11]. Despite being illegal, euthanasia was part of the
ing decisions regarding high-technology therapies and practice of 36 % of respondents. Withdrawal of all sup-
procedures that may offer small or no benefit to their port was usually preferred to euthanasia, especially by
quantity or quality of life [1]. When physicians take Catholic doctors. In 1996, the forgoing of life-sustaining
into consideration the good of society and the good of therapy was reported to take place in 5±22 % of patients
their patients, they may have to make extremely diffi- admitted to an ICU [10]. In 1997, a US study demon-
cult decisions with regard to continuation of life support strated an increase from 51 % to 90 % of deaths preced-
in the face of poor prognosis and limited space in the in- ed by recommendations to forgo life-sustaining treat-
tensive care unit (ICU) [2]. ment [12]; CPR was carried out in only 10 % of deaths.
There are great differences between countries [10] and
between states within the US [13]. This recent report of
practice in different ICUs in 38 of the US states found
Historical perspective
that of 6903 deaths, 393 patients were brain dead; 1544
Closed chest cardiac massage [3] and, subsequently, car- (23 %) received full ICU care including failed CPR;
diopulmonary resuscitation (CPR) were developed for 1430 (22 %) received full ICU care excluding CPR; 797
the treatment of sudden death, with successful out- (10 %) had life support withheld; and 2139 (38 %) had

life support withdrawn. There was wide variation in
practice among ICUs, with ranges of 4±79 %, 0±83 %,
0±67 %, and 0±79 % in the four respective categories de-
scribed above. The authors concluded that limitation of
life support prior to death is the predominant practice
in American ICUs, though there was wide variation in
end-of-life decision making between different ICUs.
The `do not resuscitate' order
To ensure that CPR is not performed in certain patients,
the ªdo not resuscitateº (DNR) order was developed
[4]. When a patient is designated as ªdo not resuscitateº,
other therapies may or may not be withheld or with-
drawn concurrently. The DNR order should only mean
the withholding of CPR and not the withdrawal or with-
holding of monitoring, other drugs, food, and fluids un-
less this has been discussed and explicitly agreed to [8].
Withholding and withdrawing of life support are the
processes by which patients are either denied or have re-
moved from them various therapeutic modalities, with
the expectation that they may die as a result [2]. Treat-
ments that may be withheld or withdrawn apart from
CPR are mechanical ventilation, oxygen, positive end-
expiratory pressure (PEEP), vasopressors, blood trans-
fusions, surgery, antibiotics, dialysis, antiarrhythmic
drugs, nutrition, and fluid therapy [11].
In the US, a DNR decision was found to also limit
other therapies, although some patients designated
DNR did receive chemotherapy, surgery, and intensive
care [14, 15, 16, 17]. A Canadian study showed that with-
drawal of inotropic support and ventilation often ac-
companied the DNR order [18]. In intensive care the
writing of a DNR order may represent a turning point
in the therapeutic attitude towards the patient, and
may be the start of an end-of-life process of withdrawing
or withholding life support [2]. For this reason, some
physicians are afraid to classify a patient as DNR be-
cause the staff may change its aggressive attitude with
respect to other therapies. In Europe, DNR orders are
also often associated with withdrawal and withholding
of life-sustaining therapies other than CPR [19, 20].
Limiting treatments at the end of life
Decisions concerning forgoing life-sustaining treat-
ments must take into account preservation and quality
of life [21, 22]. Patients may choose to forgo treatment,
or clinicians may judge that major goals of therapy are
unachievable [21]. The medical justification for any
treatment is the benefit that arises from such treatment.
If the treatment has achieved those benefits or can no
longer reasonably be expected to do so, the treatment
loses its justification and may be forgone [21]. Physi-
1415
cians sometimes refer to these therapies as futile. A
healthcare professional has no obligation to offer, begin
or maintain a treatment which, in his or her best judge-
ment, will be physiologically futile [21, 23], unreason-
able or both, either voluntarily or in response to patient
or surrogate demands [9, 24]. In fact, most patients and
surrogates agree with reasonable recommendations
from the physician to forgo life-sustaining therapy [25].
There is, however, no accepted definition of futility; de-
scribing a therapy as futile incorporates a value judge-
ment that carrying out the therapy is ªnot worth itº.
The American Thoracic Society in 1991 defined a life-
sustaining intervention as futile ª...if reasoning and ex-
perience indicate that the intervention would be highly
unlikely to result in a meaningful survival for that pa-
tient.º [24]. Some believe that society should acknowl-
edge the patient's right to decide ends where futility be-
gins [26], and that it is inherently and unavoidably mis-
leading to offer a futile treatment [27]. The ethics com-
mittee of the Society of Critical Care Medicine pub-
lished a consensus statement in 1997 regarding futile
treatments [28]. This suggested that treatments that are
extremely unlikely to be beneficial, are extremely cost-
ly, or are of uncertain benefit may be considered inap-
propriate and inadvisable, but should not be labelled fu-
tile. Communities have a legitimate interest in allocat-
ing limited medical resources by excluding inadvisable
treatments [28].
Futility arguments
The decision not to provide a useless therapy requires
two sets of value judgements. Any assertion that a ther-
apy will be useless is a matter of probability, not certain-
ty [27], and trying to predict outcomes reliably from var-
ious prognostic scoring systems is notoriously unsuc-
cessful [29]. Physicians cannot predict with absolute cer-
tainty that a patient would not survive, for instance, a
cardiac arrest and resuscitation [27]. Secondly, useful-
ness or futility are judged only relative to an end [30].
Judging a therapy futile also implies a judgement limit-
ing the ends that the physician is obliged to adopt [27].
Each medical treatment should be considered from
the patient's perspective in the light of the overall bene-
fit that it may offer, the burdens that it may entail, and
the professional duties that are involved. One of the
most important factors is the likely success of the con-
tinued course of treatment in achieving its goals and an-
ticipated benefits that justified its initial use [21, 28]. A
treatment offering a reasonable expectation of physio-
logical benefit may be withheld in the case of a patient
suffering from terminal illness; treatments offering no
benefit and which serve to prolong the dying process
should not be used [21]. In particular, CPR should not
be offered to patients who are brain dead, vegetative,
1416
critically or terminally ill with little chance of recovery
and so unlikely to benefit from CPR [9]. Futility policies
should be developed by health professionals, patients,
and community leaders [25].
Decision making
In the US, most decisions regarding the forgoing of life-
sustaining treatments are made by family members act-
ing as surrogates using a substituted judgement or best-
interest standard [31]. Substituted judgement occurs
when a surrogate attempts to determine what the in-
competent patient would have decided had the patient
been able to choose. This can only be used if the patient
was at one time capable of developing preferences and
values, and left reliable evidence of those attitudes con-
cerning their current medical situation [32]. The `best-
interests' standard attempts to promote the good of the
individual as viewed by the shared values of society.
Such factors as the avoidance of death, relief of pain
and suffering, preservation or restoration of function-
ing, and quality and extent of life, as well as the impact
on the patient's family are usually taken into account
[33].
Dying patients and their families may have different
goals than health care providers, e.g., promoting com-
fort, maintaining or achieving the ability to communi-
cate, allowing death to occur with rapidity and certainty,
or allowing a distant relative to arrive before death [34].
Even the most doomed CPR attempt may have symbol-
ic or psychological significance for the patient or family
[27]. As Snider has said ª...the art of medicine still lies
in making the correct decision on the basis of incom-
plete information, and in finding an appropriate balance
among the often conflicting principles that guide medi-
cal practice.º [8].
Patient factors
Conditions associated with the forgoing of life-sustain-
ing treatments are: severe sepsis, shock, severe neuro-
logical, cardiac and/or respiratory dysfunction, multiple
organ failure, disorders unresponsive to maximal thera-
py, malignancies, severe underlying illnesses [35, 36], se-
vere head injury [37], and overall poor prognosis [22, 38,
39]. Older, sicker, and mentally impaired patients are
usually the ones to receive DNR orders [8]. The lack of
a private attending physician was an independent pre-
dictor for the withdrawal of life-sustaining interventions
in one study [40]. Characteristics that have been found
not to be associated with decisions to forgo treatments
are gender, race, social worth, type of health insurance,
and potential costs to society [41].
Physician factors
Physicians find it easier to limit resuscitative efforts such
as chest compressions, defibrillation or antiarrhythmic
therapy, than to withdraw life-sustaining treatments
such as mechanical ventilation, nutrition, hydration or
antibiotic therapy [42]. Physicians are more likely to
withdraw blood products than haemodialysis, intrave-
nous vasopressors, total parenteral nutrition, antibiot-
ics, mechanical ventilation, tube feeding, and intrave-
nous fluids [43]. Although fluids and nutrition have
been equated with other medical therapies, some health
providers believe intravenous fluids and nutrition (other
than TPN) are key components of patient care and
should not be equated with medical interventions [44].
Physicians have also been found to prefer to withdraw
forms of therapy supporting organs that failed for natu-
ral rather than iatrogenic reasons, to withdraw recently
instituted rather than long-standing interventions, to
withdraw forms of therapy resulting in immediate death
rather than delayed death, and to withdraw forms of
therapy resulting in delayed death when confronted
with diagnostic uncertainty [43]. Physicians also prefer
to withdraw forms of life support that are scarce, expen-
sive, invasive, artificial, unnatural, emotionally taxing,
and high technology [45]. They prefer not to withdraw
forms of therapy that require continuous rather inter-
mittent administration, and forms of therapy that cause
pain when withdrawn [45].
There is a concern that end-of-life decisions are mov-
ing from withdrawal of ventilators, nutrition, and hydra-
tion towards assisted suicide and active euthanasia pro-
grams [46]. A US study in 1995 found that 96 % of re-
spondents have withheld or withdrawn life-sustaining
treatment on the expectation of a patient's death, and
that most do so frequently in the course of a year [47].
In a recent survey of physicians in ten different special-
ties in the US, 11 % would, under current legal con-
straints, hasten a patient's death in some circumstances
by prescribing medication, while 7 % said they would
provide a lethal injection [48]. A greater number (36 %
and 24 %, respectively) would agree if these steps were
legal. Three percent reported that they had written at
least one prescription to be used to hasten death, 5 %
admitted that they had administered at least one lethal
injection, while 11 % had received a request for a lethal
injection.
Preserving autonomy
One of the basic principles of medical ethics is autono-
my. Physicians have an ethical obligation to honour a
patient's autonomous decisions [49]. Patients have the
right to forgo life-sustaining treatments based on their
preferences and goals [45, 47]. In the ICU, autonomy

may be difficult to preserve when the patient is uncon-
scious, the staff has no knowledge of the patient's prior
wishes, and medical interventions are proceeding at a
rapid rate [2]. In order to maintain patient autonomy,
US physicians have been advised to obtain informed
consent from patients or surrogates regarding cardio-
pulmonary resuscitation (CPR) and the withholding
and withdrawing of other treatments [2, 26]. Decisions
to forgo life-sustaining treatments may be prospective,
for example, as in a `do not resuscitate' order for a fu-
ture cardiac arrest, or actual, for example, intubation
and ventilation for a current respiratory arrest [25]. De-
cisions to withdraw treatments are typically actual and
not prospective [10]. A 1997 survey of five tertiary care
US hospitals found that 23 % of seriously ill patients
had discussed preferences for CPR with their physicians
[50]. Of those who had not discussed preferences, 58 %
did not wish to do so, a quarter of whom did not want re-
suscitation. Patient factors independently associated
with not wanting to discuss preferences for CPR includ-
ed not being black, not having an advance directive, es-
timating an excellent prognosis, reporting fair to excel-
lent quality of life, and not desiring active involvement
in medical decisions. Factors independently associated
with wanting to discuss preferences for resuscitation
but not doing so, included being black and being young-
er.
The United States
Even in the US, where autonomy predominates, doctors
do not always follow patient or surrogate requests.
Many physicians (34 %) were found to continue life-sus-
taining treatment despite patient or surrogate wishes
that it be discontinued. Reasons given included the phy-
sician's beliefs that the patient had a reasonable chance
of recovery, that the family was not acting in the best in-
terests of the patient, and that withdrawing therapy may
be unethical or illegal, as well as the physician's fear of
malpractice litigation [47]. In addition, 42 % withheld
or withdrew life-sustaining treatment unilaterally as
they judged further intervention futile. Of these deci-
sions, 39 % were made without the knowledge or con-
sent of patients or their surrogates, and 3 % were made
despite their objections. The reality is that ICU patients
rarely take part in their own end-of-life decisions [10].
Other countries
The situation in other countries has been very different
to that in the US. More paternalistic behaviour is found
throughout Europe. A nation-wide study from Sweden
in 1990 [19] revealed that most physicians never dis-
cussed a DNR order with a patient, and that only a mi-
1417
nority of families were involved in these discussions. A
more recent paper found that discussions took place in
61 % of cases where treatments were limited [51]. There
is scanty documentation of decisions to forgo life-sus-
taining treatment in Sweden while documented deci-
sions are often indicated by means of rather vague ex-
pressions [22]. Reasons for not supplying life-sustaining
treatment related both to quantitative and qualitative
questions, but primarily poor prognosis [22]. Chrono-
logical age has been used as a criterion in some [52] but
not all [51] Swedish hospitals for making decisions that
forgo life-sustaining treatment [52]. An increasing fre-
quency of decisions was seen with increasing patient
age, probably reflecting a more compromised chronic
health status [51]. Age discrimination is a phenomenon
that has also been seen in the US [41, 53, 54]. The older
the patient, the more DNR orders are written. Patients
with different disease processes but similar prognoses
also had different rates of DNR orders [55]; those with
AIDS or metastatic lung cancer had more DNR orders
written than those with cirrhosis and heart failure.
The percentage of patients or families involved in
discussions regarding DNR orders was not even men-
tioned in one UK study, although the authors consid-
ered an ideal situation to include discussion with pa-
tients, their relatives, and other staff [56]. Another Brit-
ish study indicated that prognosis of patients considered
to be suitable for DNR was discussed with their rela-
tives in only 36 % of cases, and that DNR orders were
seldom included in the notes of patients in whom CPR
was considered inappropriate [57]. In a UK ICU popu-
lation, however, the patient's family was involved in dis-
cussions concerning DNR orders in 80 % of cases [17].
In a Dutch study only 14 % of patients were involved in
the DNR decision (32 % of all competent patients)
[58]. Another Dutch study from a geriatric department
indicated that only 3 % of patients and 24 % of families
were involved in DNR decisions [59].
In a 1999 survey of European attitudes, 80 % of re-
spondents believed that written DNR orders should be
applied but only 58 % did so [60]. Ninety-three percent
of physicians withheld treatment from patients with no
hope of meaningful life, but withdrawal was less com-
mon (77 %). Forty percent of respondents said they
would deliberately administer large doses of drugs to
these patients until death ensued. In another survey of
European intensive care doctors, Vincent found they
were not always completely honest with their patients
[61]. Doctors from the Netherlands were more likely to
give complete information, whereas doctors from
Greece, Spain, and Italy were less likely to do so.
1418
Cultural differences
Different practices may be required for patients of dif-
ferent religions and cultures. A paper from Hong Kong
found that DNR orders were infrequently employed
and rarely given as written directives [62]. It was ex-
tremely unusual for doctors to discuss with patients ei-
ther advance directives or code status. In the Chinese
culture, to tell someone of their impending death is con-
sidered not only rude, but dangerous. Decisions regard-
ing the forgoing of life-sustaining treatments are usually
made by the doctor in consultation with the family rath-
er than the patient [62]. Due to limited availability of
ICU beds in Hong Kong, patients are not always offered
the option of advanced life support; if they have been
put on life-support, the decision to withdraw or with-
hold further aggressive treatment is usually initiated by
the physicians with the consent of the family. Among
Navajo Americans, discussing negative information is
also in conflict with their view of life and may be harm-
ful [63]. In many Bantu languages (Central Africa)
there is no term for ªpersonº, so that respect for autono-
my is viewed in the context of wider social responsibility
[64]. Korean Americans and Mexican Americans also
prefer a family-oriented approach to medical decision
making as opposed to African and European Americans
who prefer a patient-autonomy approach [65]. In Israel,
withholding treatments is common, but withdrawing
treatments is limited to brain dead patients, or to pa-
tients where the medication has no physiological effect
[35]. Fluids and nutrition are considered different to
other medical interventions and are not withheld or
withdrawn [35]. Jewish ethics differentiate between ac-
tive and passive actions [35]. In a dying patient, death
may not be hastened by withdrawing therapy although
life need not be prolonged; withholding therapy is there-
fore possible.
Can we improve the situation?
The US SUPPORT study tried to improve end-of-life
decision making and to reduce the frequency of a me-
chanically supported, painful, and prolonged process of
dying [66]. The first phase found that only 47 % of physi-
cians knew when their patients preferred to avoid CPR,
46 % of DNR orders were written within 2 days of
death, 38 % of patients who died spent at least 10 days
in an ICU, and for 50 % of conscious patients who died
in the hospital, family members reported moderate to
severe pain at least half the time. Physicians did not im-
plement patients' refusals of interventions. When pa-
tients wanted CPR withheld, a DNR order was not writ-
ten in about 50 % of cases. In an effort to improve these
results, the second phase of the study consisted of inten-
sive interventions aimed at providing physicians with in-
formation concerning patient desires and objective as-
sessments of prognosis. Unfortunately, patients experi-
enced no improvement in any of the previous findings
nor was there reduced use of hospital resources [66].
Despite the greater importance placed on patient auton-
omy in the US compared to Europe, physician behav-
iour does not always demonstrate respect for patient
wishes. Physicians have been found to show great vari-
ability in choosing a level of care for given scenarios
[67].
The effect of an advance directive was also investi-
gated by the SUPPORT study group [68]. In a popula-
tion of seriously ill patients, advance directives did not
substantially enhance physician-patient communication
nor decision making about resuscitation [68], nor did in-
creasing the documentation of pre-existing advance di-
rectives reduce hospital resource use [69]. This lack of
effect was not altered by the Patient Self-Determination
Act nor by the extra efforts performed in the SUP-
PORT study, although these interventions each substan-
tially increased documentation of existing advance di-
rectives [68]. The authors suggested that increasing the
frequency of advance directives would be unlikely to
be of substantial help in improving the care of seriously
ill patients [66]. Part of the problem with advance direc-
tives for seriously ill patients was that these often did
not guide medical decision making apart from naming
a healthcare proxy or documenting general preferences
in a standard living will format [70].
In the future, practice should be improved by better
communication and more comprehensive advance care
planning [68], but for this to work physicians must take
these plans into consideration. Fins suggests that ad-
vance care planning will become effective only when
medical personnel and the public become more com-
fortable discussing end-of-life care and when physicians
understand the cultural determinants that have made
American medicine so hesitant to accept human fini-
tude [71]. Proactive ethics consultations for high-risk
patient populations offer a promising approach to im-
proving decision making and communication, and re-
ducing length of ICU stay for dying patients [42].
Summary
Physicians are increasingly involved in how their criti-
cally ill patients die [72]. The more this happens, the
more physicians will have to understand not only how
their own backgrounds and biases influence their medi-
cal management, but also the cultural and religious
backgrounds of the patient and surrogate [72, 73]. The
medical profession must realise that, despite tremen-
dous advances in medical knowledge and technology,
not everyone can be saved all the time, even in the area
of intensive care. Physicians must understand that ªdo-
 1419

ing everythingº that is best for the patient may not mean
starting epinephrine or performing CPR, but rather may
imply moving from a process of curing to caring with
palliative care [10]. This process should be initiated by
discussions with the patient or surrogate, and should in-
clude knowledge of the patients' wishes as demonstrat-
ed by advance directives and durable power of attorney.
The patient's code status and the intention of forgoing
life-sustaining treatment should be discussed with other
members of staff together with the patient and/or family
in a compassionate and humane manner. The wishes of
the patient and family should be taken into consider-
ation and the physician must try to make an impartial
decision by doing what is medically and ethically correct
and best for this specific patient. Hopefully, in this way,
a more ethical and compassionate approach to end-of-
life decisions in the ICU will be obtained.

References
1. Randall Curtis J, Park DR, Krone MR,
Pearlman RA (1995) Use of the medi-
cal futility rationale in do-not-attempt-
resuscitation orders. JAMA 273:
124±128
2. Luce JM (1990) Ethical principles in
critical care. JAMA 263: 696±700
3. Kouwenhoven WB, Jude JR, Knicker-
bocker GG (1960) Closed chest cardiac
massage. JAMA 173: 1064±1067
4. Blackhall LJ (1987) Must we always use
CPR? N Engl J Med 317: 1281±1285
5. Youngner SJ (1990) Futility in context.
JAMA 264: 1295±1296
6. Bedell SE, Delbanco TL, Cook EF, Ep-
stein FH (1983) Survival after cardio-
pulmonary resuscitation in the hospital.
N Engl J Med 309: 569±576
7. Heffner JE, Barbieri C, Casey K (1996)
Procedure-specific do-not-resuscitate
orders. Effect on communication of
treatment limitations. Arch Intern Med
156: 793±797
8. Snider GL (1991) The do-not-resusci-
tate order. Ethical and legal imperative
or medical decision? Am Rev Respir
Dis 143: 665±674
9. Luce JM (1995) Physicians do not have
a responsiblility to provide futile or un-
reasonable care if a patient or family in-
sists. Crit Care Med 23: 760±766
10. Sprung CL, Eidelman LA (1996)
Worldwide similarities and differences
in the forgoing of life-sustaining treat-
ments. Intensive Care Med 22:
1003±1005
11. Vincent JL (1990) European attitudes
towards ethical problems in intensive
care medicine: results of an ethical
questionnaire. Intensive Care Med 16:
256±264
12. Prendergast TJ, Luce JM (1997) In-
creasing incidence of withholding and
withdrawal of life support from the crit-
ically ill. Am J Respir Crit Care Med
155: 15±20
13. Prendergast TJ, Claessons MT, Luce
JM (1998) A national survey of end-of-
life care for critically ill patients. Am J
Respir Crit Care Med 158: 1163±1167
14. Evans AL, Brody BA (1985) The do not
resuscitate order in teaching hospitals.
JAMA 253: 2236±2239
15. Zimmerman JE, Knaus WA, Sharpe
SM, Anderson AS, Draper EA, Wagner
DP (1986) The use and implications of
do not resuscitate orders in intensive
care units. JAMA 255: 351±356
16. Bedell SE, Pelle D, Mager PL, Cleary
PD (1986) Do not resuscitate orders
for critically ill patients in the hospital.
How are they used and what is their im-
pact? JAMA 256: 233±237
17. Schwartz DA, Reilly P (1986) The
choice not to be resuscitated. J Am Ge-
riatr Soc 34: 807±811
18. Webster GC, Mazer CD, Potvin CA,
Fisher A, Byrick RJ (1991) Evaluation
of a ªdo not resuscitateº policy in inten-
sive care. Can J Anaesth 38: 553±563
19. Asplund K, Britton M (1990) Do not re-
suscitate orders in Swedish medical
wards. J Intern Med 228: 139±145
20. Simpson SH (1994) A study into the
uses and effects of do not resuscitate or-
ders in the intensive care units of two
teaching hospitals. Intensive and critical
care nursing 10: 12±22
21. Nyman DJ, Sprung CL (1996) Manag-
ing ethics. In: Sibbald WJ, Massaro T
(eds) The business of critical care: a
textbook for clinicians who manage
special care units. Futura, Armonk,
N. Y., USA
22. Melltorp G, Nilstun T (1996) Decisions
to forego life-sustaining treatment and
the duty of documentation. Intensive
Care Med 22: 1015±1019
23. Williams F (1993) Medical futility in
context. Br J Hosp Med 50: 50±53
24. American Thoracic Society (1991)
Withholding and withdrawing life-sus-
taining therapy. Am Rev Respir Dis
144: 726±731
25. Luce JM (1997) Making decisions about
the forgoing of life-sustaining therapy.
Am J Respir Crit Care Med 156:
1715±1718
26. Scofield GR (1991) Is consent useful
when resuscitation isn't? Hastings Cen-
ter Report, Nov±Dec
27. Tomlinson T, Brody H (1990) Futility
and the ethics of resuscitation. JAMA
264: 1276±1280
28. The Ethics Committee of the Society of
Critical Care Medicine (1997) Consen-
sus statement of the Society of Critical
Care Medicine's Ethics Committee re-
garding futile and other possibly inad-
visable treatments. Crit Care Med 25:
887±891
29. Suter P, Armaganidis A, Beaufils F, et al
(1994) Predicting outcome in ICU pa-
tients. Intensive Care Med 20: 390±397
30. Youngner SJ (1988) Who defines futili-
ty? JAMA 260: 2094±2095
31. Weir RF, Gotlin L (1990) Decisions to
abate life-sustaining treatment for non-
autonomous patients. Ethical standards
and legal liability after Cruzan. JAMA
264: 1846±1853
32. President's commission for the study of
ethical problems in medicine and bio-
medical and behavioral research (1983)
Deciding to forgo life-sustaining treat-
ment: ethical, medical, and legal issues
in treatment decisions. Patients who
lack decision-making capacity. Wash-
ington, D. C., Government Printing Of-
fice, 121±170
33. Nyman DJ, Sprung CL (1991) Ensuring
informed consent: essentials and specif-
ic exceptions. J Crit Illness 6: 891±906
34. Brody H, Campbell ML, Faber-Langen-
doen K, Ogle KS (1997) Withdrawing
intensive life-sustaining treatment ±-
recommendations for compassionate
clinical management. N Engl J Med
336: 652±657
35. Eidelman LA, Jakobson DJ, Pizov R,
Geber D, Leibovitz L, Sprung CL
(1998) Forgoing life-sustaining treat-
ment in an Israeli ICU. Intensive Care
Med 24: 162±166
36. Wood GG, Martin E (1995) Withhold-
ing and withdrawing life-sustaining
therapy in a Canadian ICU. Can J Ana-
esth 42: 186±191
1420
37. O'Callahan JG, Fink C, Pitts LH, Luce
JM (1995) Withholding and withdraw-
ing of life support from patients with se-
vere head injury. Crit Care Med 23:
1567±1575
38. Turner JS, Michell WL, Morgan CJ,
Benatar SR (1996) Limitation of life
support: frequency and practice in a
London and a Cape Town intensive
care unit. Intensive Care Med 22:
1020±1025
39. Smedira NG, Evans BH, Grais LS, et al
(1990) Withholding and withdrawal of
life support from the critically ill.
N Engl J Med 322: 309±315
40. Kollef MH (1996) Private attending
physician status and the withdrawal of
life-sustaining interventions in a medi-
cal intensive care unit population. Crit
Care Med 24: 968±975
41. Silverman HJ (1996) How decisive are
physician values in end-of-life decision
making? Crit Care Med 24: 909±911
42. Dowdy MD, Robertson C, Bander JA
(1998) A study of proactive ethics con-
sultation for critically and terminally ill
patients with extended lengths of stay.
Crit Care Med 26: 252±259
43. Christakis NA, Asch DA (1993) Biases
in how physicians choose to withdraw
life support. Lancet 342: 642±646
44. Task force on ethics of the Society of
Critical Care Medicine (1990) Consen-
sus report on the ethics of forgoing life-
sustaining treatments in the critically
ill. Crit Care Med 18: 1435±1439
45. Asch DA, Christakis NA (1996) Why
do physicians prefer to withdraw some
forms of life support over others? In-
trinsic attributes of life-sustaining treat-
ments are associated with physicians'
preferences. Med Care 34: 103±111
46. Sprung CL (1990) Changing attitudes
and practices in forgoing life-sustaining
treatments. JAMA 263: 2211±2215
47. Asch DA, Hansen-Flaschen J, Lanken
PN (1995) Decisions to limit or contin-
ue life-sustaining treatment by critical
care physicians in the United States:
conflicts between physicians' practices
and patients' wishes. Am J Respir Crit
Care Med 151: 288±292
48. Meier DE, Emmons CA, Wallenstein S,
Quill T, Morrison RS, Cassel CK (1998)
A national survey of physician-assisted
suicide and euthanasia in the United
States. N Engl J Med 338: 1193±1201
49. Sprung CL (1991) The Cruzan decision
and critical care physicians. Crit Care
Med 19: 598±599
50. Hofman JC, Wenger NS, Davis RB,
Teno J, Connors AF, Debiens N, Lynn
J, Phillips RS for the SUPPORT inves-
tigators (1997) Patient preferences for
communication with physicians about
end-of-life decisions. Ann Intern Med
127: 1±12
51. Sjokvist P, Sundin P-O, Berggren L
(1998) Limiting life support. Experienc-
es with a special protocol. Acta Anaes-
thesiol Scand 42: 232±237
52. Melltorp G, Nilstun T (1996) Age and
life-sustaining treatment. Attitudes of
intensive care unit professionals. Acta
Anaesthesiol Scand 40: 904±908
53. Raffin TA (1995) Withdrawing life sup-
port. How is the decision made?
JAMA 273: 738±739
54. Boyd K, Teres D, Rapoport J, Leme-
show S (1996) The relationship between
age and the use of DNR orders in criti-
cal care patients. Evidence for age dis-
crimination. Arch Intern Med 156:
1821±1826
55. Wachter EM, Luce JM, Hearst M, et al
(1989) Decisions about resuscitation:
inequities among patients with different
diseases but similar prognoses. Ann In-
tern Med 111: 525±532
56. Stewart K, Abel K, Rai GS (1990) Re-
suscitation decisions in a general hospi-
tal. BMJ 300: 785
57. Aarons EJ, Beeching NJ (1991) Survey
of ªdo not resuscitateº orders in a dis-
trict general hospital. BMJ 303:
1504±1506
58. Van Delden JJM, Van der Maas PJ, Pij-
nenborg L, Loomen CWN (1993) De-
ciding not to resuscitate in Dutch hospi-
tals. J Med Ethics 19: 200±205
59. Dautzenberg PLJ, Duursma SA, Be-
zemer PD, Van Engen C, Schonwetter
RS, Hooyer C (1993) Resuscitation de-
cisions on a Dutch geriatric ward. Q J
Med 86: 535±542
60. Vincent J-L (1999) Forgoing life sup-
port in western European intensive
care units: the results of an ethical ques-
tionnaire. Crit Care Med 27: 1626±1633
61. Vincent J-L (1998) Information in the
ICU: are we being honest with our pa-
tients? The results of a European ques-
tionnaire. Intensive Care Med 24:
1251±1256
62. Ip M, Gilligan T, Koenig B, Raffin TA
(1998) Ethical decision making in
Hong Kong. Crit Care Med 26: 447±451
63. Caresse JA, Rhodes LA (1995) Western
bioethics on the Navajo reservation.
Benefit or harm? JAMA 274: 826±829
64. Barry M (1988) Ethical considerations
of human investigation in developing
countries. The AIDS dilemma. N Engl
J Med 319: 1083±1085
65. Blackhall LJ, Murphy ST, Frank G, et al
(1995) Ethnicity and attitudes toward
patient autonomy. JAMA 274: 820±825
66. The SUPPORT Principal Investigators
(1995) A controlled trial to improve
care for seriously ill hospitalized pa-
tients. The study to understand progno-
ses and preferences for outcomes and
risks of treatments (SUPPORT).
JAMA 274: 1591±1598
67. Cook DJ, Guyatt GH, Jaeschde R, et al
(1995) Determinants in Canadian
health care workers of the decision to
withdraw life support from the critically
ill. JAMA 273: 703±708
68. Teno J, Lynn J, Wenger N, et al., for the
SUPPORT investigators (1997) Ad-
vance directives for seriously ill hospi-
talized patients: effectiveness with the
patient self-determination act and the
SUPPORT intervention. J Am Geriatr
Soc 45: 500±507
69. Teno J, Lynn J, Connors AF, et al., for
the SUPPORT investigators (1997)
The illusion of end-of-life resource sav-
ings with advance directives. J Am Ge-
riatr Soc 45: 513±518
70. Teno J, Licks S, Lynn J, et al., for the
SUPPORT investigators (1997) Do ad-
vance directives provide instructions
that direct care? J Am Geriatr Soc 45:
508±512
71. Fins JJ (1997) Advance directives and
SUPPORT. J Am Geriatr Soc 45:
519±520
72. Karlawish JH, Hall JB (1997) Managing
death and dying in the intensive care
unit. Am J Respir Crit Care Med 155:
1±2
73. Oppenheim A, Sprung CL (1998)
Cross-cultural ethical decision making
in critical care. Crit Care Med 26:
423±424