End-of-life care in the pediatric intensive care

unit: research review and recommendations

Jeffrey P. Burns, MD, MPHa,*,Cynda Hylton Rushton, DNSc, RN, FAANb

aMedical-Surgical Intensive Care Unit, Childrens Hospital Boston, 300 Longwood Avenue,

Boston, MA 02115, USAbPhoebe Berman Bioethics Institute, Harriet Lane Compassionate Care Program School of Nursing,

525 N. Wolfe Street, Box 420, Baltimore, MD 21205-2110, USA

In the United States, most children die in hospitals, typically in a critical care

setting [1,2] About 55,000 children die each year in America, more than half

of them in the first year of life [1]. Beyond infancy, unintentional injuries,

congenital anomalies, malignant neoplasms, and intentional injuries are the

leading causes of death. Although this distribution of the underlying causes for

pediatric deaths has remained relatively stable over the last 20 years, what is

changing, are the events leading to the death of a child. The success of critical

care medicine in the last 25 years has meant that the progression of many

disorders can be forestalled but not permanently reversed. In many instances

families and clinicians must weigh the benefits and burdens of ongoing life-

sustaining treatments. Indeed, the most recent observational studies of death in

the pediatric intensive care unit (PICU) reveals that 40% to 60% of all deaths

occur following a decision to limit life-sustaining therapies [3,4]. End-of-life care

is receiving increased scrutiny from all aspects of our society and profession [5].

The recent Institute of Medicine report calls for sweeping improvements to

address current gaps in care [1]. To address these gaps, several national initiatives

are addressing various aspects of pediatric palliative care. The Initiative for

Pediatric Palliative Care (IPPC) has conducted research, provided technical

assistance to childrens hospitals developing new quality improvement initiatives,

Crit Care Clin 20 (2004) 467485and completed a comprehensive, interdisciplinary curriculum [6].

The National Hospice and Palliative Care Organization has also developed

educational materials targeted to hospices [7]. In addition, a curriculum specifi-

0749-0704/04/$ see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.ccc.2004.03.004

* Corresponding author.

E-mail address: jeffrey.burns@tch.harvard.edu (J.P. Burns).

agreed or strongly agreed that withholding and withdrawing are ethically the

same (P < 0.001). Physicians were more likely than nurses to report that familiesare well informed about the advantages and limitations of further therapy (99%

versus 89%, P < 0.003); that ethical issues are discussed well within the team

(92% versus 59%, P < 0.0003), and that ethical issues are discussed well with

the family (91% versus 79%, P < 0.0002). On multivariable analyses, this study

found that fewer of years of practice in pediatric critical care was the only

clinician characteristic associated with attitudes on end-of-life care dissimilar to

the consensus positions reached by national medical and nursing organizations on

these issues. The investigators found no association between clinician character-

istics such as their political or religious affiliation, practice related variables such

as the size of their intensive care unit (ICU) or the presence of residents and

fellows, and particular attitudes about end-of-life care.

One notable finding of this national survey was that the attitudes and reportedcally targeted to pediatric nurses has been developed as part of the End-of-Life

Nursing Education Consortium project [8] and Childrens Hospice International

is working with the Center for Medicaid and Medicare Services to encourage

more coordination of care for children and families across the continuum of care.

Taken together, these efforts reflect burgeoning clinical and behavioral research

leading to organizational and educational initiatives to improve the quality of care

in this area [9,10]. Indeed, increased accountability for the quality of end-of-life

care provided demands that actual practice in this important area of pediatric

critical care medicine receive more rigorous examination than in years past [11].

Our goal here is to review the empirical data that exists on end-of-life care in the

PICU and to address some of the practical aspects of providing this care.

The literature on end-of-life care in the pediatric intensive care unit

Attitudes of pediatric critical care physicians and nurses on end-of-life care

Several studies have examined attitudes of critical care clinicians and provide

insights into the current quality of end-of-life care in pediatric intensive care

units. One of the largest studies on end-of-life care in the Pediatric intensive care

unit was a three-part investigation recently reported as a series of special articles

in the journal Critical Care Medicine [3,12,13]. The first part was a survey of the

attitudes and practices of pediatric critical care clinicians on end-of-life care from

a random sample of clinicians at 31 pediatric hospitals in the United States [12].

The survey was completed by 110 of 130 (85%) pediatric critical care attending

physicians and 92 of 130 (71%) pediatric critical care nurses. As seen in Tables 1

and 2, no physician or nurse agreed with the statement that withholding and

withdrawing life support is unethical. More physicians (78%), than nurses (57%)

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485468practices of these pediatric critical care practitioners were in alignment with the

existing theory on how clinicians should behave. Two thirds of pediatric critical

care physicians and nurses reported attitudes about end-of-life practice in strong

Table 1

Clinicians attitudes toward end-of-life care

Physicians, % Nurses, % P-value

Percentage of respondents who agreed or strongly agreed with the statement:

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 469agreement with the ethical and legal consensus that has evolved in these areas.

Further, over 90% of physicians and nurses cited patient-centered variables as the

most important determinants of decision making about life-sustaining treatments,

and the majority rated as less important nonpatient-centered variables such as

financial costs to society or ICU bed availability. Both physicians and nurses

were far more likely to rate as important in decision-making quality-of-life

considerations from the perspective of the patient and family rather than the

(On a scale where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree)

Withholding or withdrawing

life support is unethical.

0 (1.15 0.39) 0 (1.24 0.43) 0.09

Withholding is more

ethical than withdrawing.

5 (1.67 0.89) 18 (2.45 0.99) 0.0001

Withdrawing is more ethical

than withholding.

7 (1.77 0.99) 9 (2.51 0.87) 0.0001

Withholding and withdrawing

are ethically the same.

78 (4.16 1.06) 57 (3.43 1.13) 0.0001

Percentage of respondents, in answering the question How important are the following factors in

influencing your decision on the extent of life support therapy to provide a patient? who rated the

following variables as important or very important (On a scale where 1 = not important, 2 = less

important, 3 = neutral, 4 = important, 5 = very important)

Quality of life as viewed by the patient 99 (4.60 0.64) 99 (4.77 0.42) 0.05

Quality of life as viewed by the family 95 (4.39 0.72) 96 (4.49 0.69) 0.24

Patient unlikely to survive 94 (4.08 0.96) 91 (4.07 1.00) 0.98

Potential for neurologically intact survival 81 (4.12 0.85) 77 (4.33 0.79) 0.06

Quality of life with a chronic disorder 61 (3.50 0.99) 73 (3.85 1.09) 0.009

Fear of litigation or breaking the law 23 (2.46 1.12) 32 (2.91 1.10) 0.002

Financial costs to society 13 (2.19 1.00) 33 (2.82 1.18) 0.0002

ICU bed availability 3 (1.32 0.71) 4 (1.50 0.85) 0.05

Percentage of respondents, in answering the question If the circumstances for withdrawing life

support are indicated, do you think the following medications should be added or increased in the

patients regimen as life support is discontinued? who responded frequently or always to the

following variables.

(On a scale where 1 = never, 2 = infrequently, 3 = sometimes, 4 = frequently, 5 = always)

Narcotics 74 (3.96 0.75) 86 (4.26 0.69) 0.005

Benzodiazepines 62 (3.70 0.72) 66 (3.84 0.94) 0.15

Barbiturates 24 (2.70 1.03) 27 (2.97 0.97) 0.07

Neuromuscular blocking agents

(ex. Pavulon)

2 (1.28 0.68) 2 (1.53 0.81) 0.10

Percentages represent respondents who answered with scores of 4 or 5, on a five-point scale shown.

Numbers in parentheses are group mean scores SD. P-values are for the comparison between

physician and nurse responses by Wilcoxon rank sums.

Data from Burns JP, Mitchell C, Griffith JL, Truog RD. End-of-life care in the pediatric intensive care

unit: attitudes and practices of pediatric critical care physicians and nurses. Crit Care Med 2001;29(3):

65864.

Table 2

Clinicians reports of actual end-of-life practices in their ICU

Physicians, % Nurses, % P-Value

In your experience, who usually initiates the discussion to limit life-sustaining treatment?

(On a scale where 1 = never, 2 = infrequently, 3 = sometimes, 4 = frequently, 5 = always)

Physicians 96 (4.05 0.40) 86 (4.00 0.53) 0.35

Nurses 28 (2.95 0.78) 55 (3.41 0.74) 0.0001

Family 4 (2.56 0.56) 4 (2.61 0.57) 0.56

Other (clergy, consultants) 1 (1.75 0.61) 1 (1.79 0.71) 0.86

Who else, if anyone, do you consult when making decisions about limitations of life support on

patients in your PICU?

(On a scale where 1 = never, 2 = infrequently, 3 = sometimes, 4 = frequently, 5 = always)

Hospital legal representatives 6 (1.89 0.80) 4 (2.08 0.77) 0.05

Hospital ethics committee 8 (2.28 0.85) 7 (2.46 0.77) 0.20

Other clinicians from within the unit 66 (3.78 0.91) 57 (3.54 1.02) 0.09

Clinicians outside the unit 8 (2.27 0.98) 9 (2.14 0.93) 0.44

At the time that withholding and withdrawing life support is discussed, how would you rate the

following areas as it is practiced in your unit?

(On a scale where 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree)

Families are well informed about prognosis 100 (4.86 0.34) 97 (4.67 0.61) 0.01

Families are well informed about the advantages

and limitations of further therapy

99 (4.77 0.48) 89 (4.47 0.80) 0.003

The ethical issues for each patient are

well discussed within the care team

92 (4.39 0.69) 59 (3.80 1.14) 0.0003

The ethical issues for each patient are well

discussed between the care team and family

91 (4.40 0.76) 79 (3.98 0.91) 0.0002

If the family insists on continuing life support, even if the care team concludes that further

therapy is futile, what is the usual response as it is practiced in your unit?

(On a scale where 1 = never, 2 = infrequently, 3 = sometimes, 4 = frequently, 5 = always)

We continue for a short time for them to reconsider,

but then we terminate life support unilaterally

1 (1.21 0.51) 3 (1.74 0.87) 0.01

We continue for a short time for them to reconsider,

but then we transfer care to another institution

1 (1.55 0.70) 3 (1.56 0.79) 0.79

We withdraw life support without the

agreement of the family, after consultation

with the hospital ethics or legal representatives

1 (1.21 0.58) 0 (1.37 0.55) 0.40

We continue to provide life support for as long as

requested, and usually, through ongoing

discussions, a consensus between the family

and care team is reached

93 (4.17 0.65) 88 (4.08 0.68) 0.27

When life support is withdrawn, how involved is the physician in the actual process at the bedside?

(On a scale where 1 = never, 2 = infrequently, 3 = sometimes, 4 = frequently, 5 = always)

100 (4.87 0.34) 84 (4.35 0.77) 0.0001

Percentages represent respondents who answered with scores of 4 or 5, on a five-point scale shown.

Numbers in parentheses are group mean scores SD. P-values are for the comparison between

physician and nurse responses by Wilcoxon rank sums.

Data from Burns JP, Mitchell C, Griffith JL, Truog RD. End-of-life care in the pediatric intensive care

unit: attitudes and practices of pediatric critical care physicians and nurses. Crit Care Med 2001;

29(3):65864.

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485470

statement quality of life with a chronic disorder. In accordance with published

consensus recommendations, quality-of-life determinations, especially when con-

sidered as a factor in the decision to forego life-sustaining treatment, should

only be judged from the patients perspective, and not by health care workers

personal assessments. Finally, over two thirds of physicians and nurses were

willing to add or increase analgesia and sedation to a patients regimen as life-

sustaining treatment is discontinued, a position also consistent with a series of

recent recommendations from national medical organizations [12]. A large

majority of physicians and nurses did not agree with the addition of neuromus-

cular blocking agents, medications devoid of any sedative or analgesic proper-

ties, as life-sustaining treatment is discontinued, a view again in keeping with

recent recommendations on this issue put forth by the Society of Critical Care

Medicine [14].

However, this national survey study also found that pediatric critical care

clinicians with fewer years of experience espoused attitudes and practices that are

divergent with consensus recommendations on these issues. For example, this

study found an association between fewer years of experience as practitioners of

pediatric critical care and a greater reluctance to withdraw life-sustaining

treatment and a greater reluctance to administer analgesics or sedatives as life-

sustaining treatment is withdrawn [12]. Previous studies have also found that less

experienced clinicians are less willing to forgo life-sustaining treatment. Cook

et al reported that, among ICU attending staff, house staff and nurses in 37

Canadian university-affiliated hospitals, the longer the duration since graduation,

the less likely clinicians were to offer aggressive care in response to 12 different

clinical scenarios [15].

Unlike previous research that has demonstrated that physicians caring for adult

patients frequently make treatment decisions based on their own personal values

rather than patient-centered factors, this study of pediatric critical care practi-

tioners found no evidence that decision making is significantly influenced by the

clinicians religious beliefs, religiosity, specialty, age, or practice locale. Burns

et al speculated that although it is possible that the less-experienced physicians

and nurses are aware of recommendations and consensus statements regarding

life-sustaining treatments and simply disagree with them, it is more likely that

they are unfamiliar or uncomfortable with these issues [12]. In support of this,

other data from the medical literature demonstrates that both physicians and

nurses are often not well trained in end-of-life patient management. A study of

115 medical residents at three teaching hospitals found that one third of them

reported that they had never been supervised discussing do-not-resuscitate

discussions with patients [16]. These deficiencies have also been noted in the

education of nurses. The American Association of Colleges of Nursing consensus

group concluded that there were major deficiencies, or complete neglect, of end-

of-life content in nursing curricula [17]. More than half of 300 nurses at an acute

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 471care teaching hospital for adult patients reported that they did not have a good

understanding of advance directives [18]. In addition, there are special issues that

should receive particular attention. For example, a survey conducted by Solomon

et al found that 37% of pediatric specialists and 44% of nurses caring for

critically ill or dying children believed that Even if life supports such as

mechanical ventilation and dialysis are stopped, medically supplied food and

water should always be continued [9]. These beliefs are contrary to published

guidelines permitting the forgoing of artificial nutrition and hydration in pediatric

end-of-life care [19,20].

The data revealing inadequate training of clinicians in ethical concerns as-

sumes even greater importance because the pediatric critical care clinicians in

the national survey reported a relatively insulated process of decision making

[12]. Respondents stated that discussions to limit life-sustaining treatment are

typically initiated by the pediatric critical care clinicians. Further, in less than

10% of cases are clinicians outside of the ICU, hospital ethicists, or legal rep-

resentatives consulted for advice on these matters; yet a majority of respondents

stated that they do consult with other clinicians within the ICU for advice on

these issues. It could not be determined from this study whether the less-

experienced pediatric critical care clinicians seek advice for these pivotal de-

cisions from their more senior colleagues.

The national survey of pediatric critical care physicians and nurses also found

that significant perceptual differences exist between physicians and nurses on the

quality of end-of-life care as it is actually practiced [12]. Compared with

physicians, nurses were significantly less likely to report that ethical issues are

well discussed within the care team, between the care team and family, or that

families are well informed about the advantages and limitations of further

therapy. For example, 92% of physicians reported that ethical issues for each

patient were well discussed with the care team, while only 59% of nurses thought

so. In addition, nearly one out of five pediatric critical care nurses reported

that physicians are not always or frequently at the bedside as life-support

is withdrawn.

Other studies that have reported nurses feeling disenfranchised in the decision-

making process and dissatisfied with the level of support for families facing end-

of-life decisions for their childrendespite, or perhaps because of the fact that

nurses bear major responsibility for implementing end-of-life decisions, provid-

ing terminal care and helping families through their childs death. Using

qualitative methods, Oberle and Hughes identified and compared Canadian

doctors and nurses perceptions of ethical problems related to their work in

adult oncology [21]. They concluded that the most pressing ethical problem at the

end of life was witnessing and responding to the suffering of others, and that it

created significant moral burden for each group. These investigators postulated

that differences in responses of each discipline were a function of their pro-

fessional role rather than differences in ethical reasoning or moral orientation.

The primary differences were because doctors are responsible for making

decisions and nurses must live with and carry out the decisions. In response to

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485472moral quandaries and the associated distress, each group approached the issues

through a lens of competing values, hierarchical processes, scarce resources, and

communication patterns [21].

Observational studies of clinical practice regarding death in the pediatric

intensive care unit

A common concern with surveys of attitudes is that the responses may reflect

only what the clinicians believe, or want to believe, but not actually how they

behave. Yet, there is emerging data on how clinicians actually provide care in this

context. In the past 5 years several studies have examined the actual practices of

critical care physicians and nurses in providing end-of-life care in the pediatric

intensive care unit. For example, Part II of the Boston study was an observational

study of 53 consecutive patients who died following the withdrawal of life-

sustaining treatment in the PICU at three teaching hospitals [3]. Data on the

reasons why medications were given were obtained from a self-administered,

anonymous questionnaire completed by the critical care physician and nurse for

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 473each case. Data on what medications were given were obtained from a review of

the medical record.

As seen in Table 3, the investigators found that sedatives or analgesics were

administered to 47 (89%) patients who died following the withdrawal of life-

sustaining treatment. Patients who were comatose were less likely to receive

these medications. Physicians and nurses cited treatment of pain, anxiety, and air

hunger as the most common reasons, and hastening death as the least common

reason, for administration of these medications, as seen in Table 4. The mean

dose of sedatives and analgesics administered nearly doubled as life support was

withdrawn, and the degree of escalation in dose did not correlate with clinicians

views on hastening death. These investigators concluded that the clinicians in this

study frequently escalate the dose of sedatives or analgesics to dying patients as

life-sustaining treatment is withdrawn, citing patient-centered reasons as their

principle justification, and were satisfied with the care provided.

Hastening death was viewed by the pediatric critical clinicians in this study

as an unintended consequence of appropriate care. This line of reasoning is fre-

Table 3

Mean dose of benzodiazepines and opiates administered to patients around the withdrawal of life-

sustaining treatment

24 hours before

withdrawal

4 hours before

withdrawal

4 hours after

withdrawal

24 hours after

withdrawal

Benzodiazepine dose:

mg/kg/ha0.54 0.59 0.98 1.4

(diazepam equivalents) n = 44 n = 44 n = 44 n = 6

Opiate dose: mg/kg/ha 0.54 0.55 1.00 1.80

(morphine equivalents) n = 41 n = 44 n = 46 n = 4

a Comparison of doses 4 hours prior to withdrawal of life-sustaining treatment with doses 4 hours

after withdrawal of life-sustaining treatment, P < 0.0001 by Wilcoxon matched-pairs signed-ranks test.Data from Burns JP, Mitchell C, Outwater KM, Geller M, Griffith JL, Todres ID, et al. End-of-life care

in the pediatric intensive cave unit after the forgoing of life-sustaining treatment. Crit Care Med

2000;28(8):30606.

Table 4

Clinicians rationale for managing patient discomfort as life-sustaining treatment was withdrawn

Among patients who did receive palliative medications (n = 47)

Decrease

pain

Decrease

anxiety

Decrease

air hunger

Comfort

family

Hasten

death

Ordered by physicians 89% (42) 89% (42) 85% (40) 66% (31) 4% (2)

Administered by nurses 87% (41) 81% (38) 79% (37) 77% (36) 6% (3)

Physician/nurse paira,b 83% (39) 77% (36) 74% (35) 49% (23) 2% (1)

Among patients who did not receive palliative medications (n = 6)

Patient was in

vegetative state

Patient appeared

comfortable

Pain control was

unethical in this case

Physician/nurse paira 100% (6) 100% (6) 0% (0)

a Percentages of physiciannurse matched pairs that agreed/strongly agreed with reasons for

administering medications at time of treatment withdrawal. Numbers are in parentheses.b Among matched pairs not in agreement, differences between physicians and nurses not sig-

nificant (P > 0.05) by McNemar chi square.

Data from Burns JP, Mitchell C, Outwater KM, Geller M, Griffith JL, Todres ID, et al. End-of-life care

in the pediatric intensive cave unit after the forgoing of life-sustaining treatment. Crit Care Med 2000;

28(8):30606.

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485474quently attributed to the ethical principle of double effect, which distinguishes

between prohibited death causing and permitted death-hastening analgesic

administration to terminally ill patients. According to this principle, clinicians are

permitted to give sedatives and analgesics to dying patients, despite the foreseen

risk of possibly hastening death as a result of the medications, provided the

intention is to relieve the patients pain and suffering and not to cause the

patients death [22]. Wilson et al found that 75% of adult patients from whom life

support was withheld or withdrawn at two San Francisco teaching hospitals

received nearly a threefold increase in sedatives, analgesics, or both during this

process [23]. Clinicians in the Wilson series justified this practice principally to

decrease pain, anxiety, and air hunger. Similarly, the only patients in the Wilson

series who did not receive sedatives or analgesics during life-support withdrawal

were those judged to be deeply comatose. Wilson et al also found a high rate

of agreement between physicians and nurses on satisfaction with the care

provided, and the tendency of clinicians to appeal to double effect reasoning to

justify their practice.

Part II of the Boston Study, an observational study of death in the PICU, also

found that in 13% of cases there was dissatisfaction with the care provided, and in

each instance it was the nurse who reported a concern that the amount of sedation

and analgesia provided was inadequate [3]. In a similar study at the Cleveland

Clinic, Daly found that 10% of 39 nurses caring for adult patients following the

withdrawal of mechanical ventilation were also dissatisfied with the level of

patient comfort achieved [24]. These data suggest that in a significant minority

of cases involving dying patients in the ICU following the forgoing of life-

sustaining treatment, nurses have very different perceptions of the quality of

comfort care provided. Whether these data reflects variability in clinicians

perception of suffering or systematic problems related to untreated discomfort

remains unknown.

Parental perspectives on end-of-life care in the pediatric intensive care unit

Bereaved parents are in a unique position to comment on current practice in

end-of-life care. Indeed, gaining an understanding of the perspectives of the

family on the dying process is an essential step in understanding the quality of

care provided. Part III of the Boston study sought to identify priorities for quality

end-of-life care from the parents perspective by surveying the bereaved parents

of the children who died in Part II of the study [13]. In 90% of cases, physicians

first raised withdrawal of life support, although nearly half of parents had con-

sidered it independently. Among decision-making factors, as seen in Table 5,

parents rated the quality of life, likelihood of improvement, and pain as most

important. Twenty percent of parents disagreed that their children were comfort-

able in their final days. Fifty-five percent of parents felt that they had little to no

control during their childs final days, and nearly a quarter reported that, if able,

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 475they would have made decisions differently. There were significant differences

(P < 0.001) between the involvement of family, friends, and staff members at the

Table 5

Factors important to parental end-of-life decision making

Factors N

% Parents who rate

as very important Mean (SD)

My childs quality of life 55 81.8% 4.75 (0.67)

My childs chance of getting better 55 78.2% 4.75 (0.52)

My childs pain or discomfort 55 76.4% 4.64 (0.78)

My child was unlikely to

survive hospitalization

53 69.8% 4.49 (0.95)

What I believe my child would

have wanted

52 53.8% 4.08 (1.22)

Information that the hospital staff provided 55 45.5% 4.22 (0.83)

My religious/spiritual beliefs 55 30.9% 3.00 (1.71)

The way my child looked or behaved 52 30.8% 3.50 (1.36)

Advice the hospital staff provided 56 28.6% 3.82 (0.99)

The attitudes of hospital staff 54 22.2% 3.07 (1.46)

Advice family and friends provided 53 5.7% 2.66 (1.25)

The financial costs 55 0% 1.22 (0.69)

Factors were rated by parents on a 15 scale (1 = not important, 5 = very important). Significantdifferences (P < 0.001) were found on one-way weighted least squares analysis of repeated categorical

data across all five response choices for parents who responded to all items (n = 46).

Data from Meyer EC, Burns JP, Griffith JL, Truog RD. Parental perspectives on end-of-life care in

the pediatric intensive care unit. Crit Care Med 2002;30(1):22631.

time of death, and greater agreement (P < 0.01) about the decision to withdraw

support between parents and staff members than with other family members.

Nearly one fifth of parents reported that their childs pain management was not

adequate, and nearly one quarter would have made decisions differently if they

were able to do so. One of the most striking findings of this study of bereaved

parents was that clinicians, usually previously unknown to the family, were

quickly drawn into the familys inner circle of support because of their expertise,

availability, and familiarity with the hospital culture, as well as the emotional

needs of the family. As evidence of this, nurses were considered more involved

during the actual dying process than other family members or friends.

The perceived inadequacies of pain management provided to children dying

in the PICU have also been documented in others. Wolfe et al interviewed

103 parents of children who had died of cancer between 1990 and 1997 at an

academic teaching hospital in Boston, and found that almost all children with

cancer who died in the hospital of treatment-related complications died in the

intensive care unit, after ventilatory support was withdrawn [25]. These children

dying of cancer had more symptoms, less successful control of pain and dyspnea,

and a poorer quality of life than those who died of progressive disease. Wolfe et al

also found significant discordance between the reports of parents and physicians

regarding the childrens symptoms, with some suggestive evidence that this may

result in part from a lack of recognition of the problem by the medical team.

Bereaved parents in the Wolfe study who reported that the physician was not

actively involved in care at the end of life were more likely to report that their

child suffered a great deal from pain. Still other studies have found that there

remains significant concern about untreated suffering in dying newborns despite

advances in the field of pain management and pediatric anesthesia [26,27].

Improving end-of-life care in the pediatric intensive care unit: whats to be

learned from the literature?

The literature suggests that practitioners of pediatric critical care need to invest

more effort into ensuring optimal palliative care for patients and their families in

the PICU environment. Clinicians need to anticipate and treat bothersome

symptoms of dying children more effectively; recognize that family support

and contact between the dying child and family facilitate decision making and

acceptance of death; and facilitate the coordination of care and the development

of alternative care teams to optimize end-of-life care. Yet, much remains

unknown as well. What are the different trajectories of death in the PICU and

how does the quality of care vary across these different dying trajectories? What

are the barriers and facilitators for communication among the patient, families,

and PICU staff? How can excellent palliative care and communication skills be

identified, measured, and taught? Finally, future efforts must be based on

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485476previously learned lessons. In particular, this includes not only optimizing pal-

liative care for patients and their families, but developing ongoing educational

initiatives and bereavement programs for PICU practitioners themselves.

Practical aspects of end-of-life care

How much is too much analgesia for a child dying in the pediatric intensive

care unit?

Can one externally validate the right amount of sedation and analgesia that

should be given to dying child? Only the patient can truly validate the adequacy

of the management of terminal symptoms, and in the PICU this is typically not

possible. In the PICU context, then, the right amount of symptom management

will always be a surrogate assessment. Yet, the constellation of patient signs,

symptoms, or pain scores in an individual case that would meet a universally

accepted threshold for treatment, let alone the extent of treatment to be given

once that threshold has been reached, is not apparent.

Recently, the Society of Critical Care Medicine put forth a comprehensive set

of guideline on this issue entitled, Recommendations for end-of-life care in the

intensive care unit: The Ethics Committee of the Society of Critical Care

Medicine [14]. They write, These agents should be titrated to effect, and the

dose should not be limited solely on the basis of recommended or suggested

maximal doses. In most cases, patients who do not respond to a given dose of an

opioid or benzodiazepine will respond if the dose is increasedthere is no

theoretical or practical maximal dose. Other experts have expressed the views

of many experts on pain and symptom management: The optimal dose of

morphine for relief of pain or dyspnea is determined by increasing the dose until

the patient responds. Patients who have not previously received opioids should

initially be given low doses, which should be rapidly increased until symptoms

are relieved. For patients with particularly severe or acute symptoms, rapid

titration requires that an experienced clinician be at the bedside [28].

The amount of pain relief to provide to a dying patient has also been addressed

by the United States Supreme Court [29]. In the 1997 case of Vacco v Quill,

writing for the majority Chief Justice Rehnquist stated, It is widely recognized

that the provision of pain medication is ethically and professionally acceptable

even when the treatment may hasten the patients death if the medication is

intended to alleviate pain and severe discomfort, not to cause death. Although

this opinion does not necessarily legally sanction any dose of sedative or

analgesic administered to dying patient, it does point to a widely shared ethical

and legal consensus that symptom management should not be limited by some

arbitrary level. But what is the ethical justification that supports a practitioner of

pediatric critical care in administering sedatives and analgesics in the PICU even

when it is foreseen that it will hasten the patients death?

Ethical justification for relieving pain and suffering

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 477The ethical justification most often cited for the administration of sedatives

and analgesics to a dying patient, even when it is foreseen that this will hasten the

patients death, is the Doctrine of Double Effect. The doctrine states that an action

that has two effects, one of which is inherently good and the other of which is

inherently bad, can be justified if certain conditions are met [22].

Many critics argue that double effect reasoning is neither a necessary or

sufficient determinant of morally permissible behavior by clinicians in providing

palliative care [30,31]. First, critics assert that the fundamental ethical justifica-

tion for care is based on the patients informed consent, not the intentions of

clinicians. Therefore, the Doctrine of Double Effect is not morally relevant or

logically valid, for it relies on an overly simplistic notion of intent that is im-

possible to externally verify. Moreover, it may have the paradoxic effect of

constraining some clinicians from providing adequate medication for relief of

suffering because of their fear of violating the principles absolute prohibition

against intentionally causing death. A second line of criticism argues that those

who appeal to the moral cover of double effect reasoning are really engaging in

disingenuous hairsplitting, a form of rationalization for a practice that is really

surreptitious euthanasia.

Yet, some pediatric ethicists and practitioners of pediatric critical care have

argued that double effect reasoning provides a guiding construct for morally

permissible clinical management of children dying in the PICU. For example,

Burns et al have written, Most terminally ill patients on life-support, whether

pediatric or adult, lack decisional capacity; therefore, voluntary consent is not an

option. Further, a rapid decline in patient comfort is common when life-support is

withdrawn from the critically ill. The sedation and analgesia adequate for a

patient receiving mechanical ventilation is usually inadequate to treat the air

hunger experienced by imminently dying patients without severe neurologic

injury as controlled ventilation is removed. Double effect reasoning provides a

defensible rationale for escalating doses among practitioners who support neither

euthanasia at one extreme nor the practice of allowing patients to die with

untreated suffering on the other [12].

Despite the emerging theoretical understandings of the ethical justification of

relieving pain and suffering, pediatric critical care professional may need

additional support to implement decision they perceive may violate important

ethical boundaries. These concerns must be anticipated and addressed through

ongoing dialog, ethics consultation, institutional policies, and clinical protocols.

For example, Kelly has advocated that moral decisions are shaped by conditions

in the workplace [32]. This suggests that institutions where end of life care is

provided have a responsibility to create support systems to address the needs of

the caregivers who participate in and witness the suffering of patients.

Neuromuscular blockade and end-of-life care

The presence of neuromuscular blockade at the time of life-sustaining

treatment withdrawal, which has been reported in both adult and pediatric pa-

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485478tients, raises issues distinct to the practice of critical care medicine. The relatively

high prevalence of neuromuscular blockade as part of standard practice for the

treatment of children with poor lung compliance requiring mechanical ventila-

tion, and the subsequent decision-making dilemmas produced when life-sus-

taining treatment is withdrawn from these patients, has made this a problem of

special concern to practitioners of pediatric critical care [3,34]. Whether this

practice occurs with the intention of trying to assure the appearance of patient

comfort for the benefit of the family, or because clinicians preoccupied with all of

the concerns around the withdrawal of life-sustaining treatment fail to consider

neuromuscular blockade, is not known. Neuromuscular blocking agents, used

to reduce ventilator-patient asynchrony and minimize oxygen consumption by

eliminating patient movement, have no sedative or analgesic properties. As such,

most commentary on this issue has concluded that the initiation of these agents

as the ventilator is being withdrawn is morally indefensible [3,14,28,34].

Is it acceptable to administer neuromuscular blockade with the intention of

comforting the family in a context where the child clearly cannot survive

(extreme levels of mechanical ventilation or ECMO, for example)? Some have

argued that the desire to comfort the patients family is an important considera-

tion, and given the certainty of the patients death following the withdrawal of

life-sustaining treatment in some situations regardless of muscle relaxation, these

clinicians have argued that initiating neuromuscular blockade at the time of

withdrawal is acceptable [33]. However, others believe that the patients well-

being always takes precedence over family interests [3,34]. Neuromuscular

blockade potentially masks symptoms of patient suffering, and therefore inter-

feres with the clinicians primary obligation of ensuring that a dying patient does

not experience untreated suffering. Such an action also does not allow for the

chance that the patient may survive without mechanical ventilation when there is

some degree of prognostic uncertainty.

What should be done when the patient is experiencing the effects of residual

neuromuscular blockade and the family decides to withdrawal life-sustaining

treatment such as mechanical ventilation? The Ethics Committee of the Society of

Critical Care Medicine has taken the position in published recommendations that

efforts should be made to allow for the restoration of neuromuscular function

before withdrawal of mechanical ventilation from patients who have previously

been receiving therapeutic neuromuscular blockade [14]. A similar position has

been advocated by other experts in the field [28].

Clearly, more education about the nuances of this dimension of clinical

practice at the end of life is needed. Forums such as ethics rounds, team rounds,

or interdisciplinary conferences should include exploration of the boundaries of

ethically justified practice and clinical protocols that include procedural safe-

guards for examining the issues before implementation. These include clear lines

of decision-making authority and avenues for raising clinical concerns.

Family presence at resuscitation attempts

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 479The official position of the Guidelines 2000 for Cardiopulmonary Resuscita-

tion and Emergency Cardiovascular Care is that family members should be

given (the option of being present at resuscitation attempts), but they will require

losses; and the death of ones self.

In the same study, Papadatou proposed a model for the common griefreactions health care professionals experience before and after a patients death

[36]. These included crying, thinking about the patient, temporary withdrawal

from activities and relationships, sorrow, anger, guilt, and despair. The length of

response varied from several hours to several days or even months in more

extreme cases. The timing of grief also varied; some health care professionals

experienced immediate grief responses while others postponed or repressed their

grief with reactions occurring at unexpected times.

In a qualitative study of nurses caring for chronically ill children who died,

Davies et al reported that nurses experienced two types of distress when they

recognized that a childs death was inevitable [37]. The first, grief distress,support and specific attention during the resuscitation. This statement stems

from surveys that have found that most people would like to be present during the

attempted resuscitation of a loved one, especially when the resuscitation attempt

involves a child. If family members are present, this is best done with a clinician

who is free to meet the unexpected needs of the family, a need that may not be

possible for the limited resources of the team to fill [35]. Despite increased

attention to this issue, there may be emotional and philosophic barriers to im-

plementing such policies. Concerted efforts to explore institutional and profes-

sional barriers to offering families this option will be necessary.

Grief and bereavement of health care professionals

Pediatric health care professionals face unique challenges when caring for

dying children and their families. In a recent report, When Children Die, the

Institute of Medicine highlighted the need to understand and develop strategies

for supporting these caregivers so that they can provide quality care [1].

Although research in this area has been limited to date, several qualitative

studies have helped to describe the way health care professionals experience and

manage grief. Their conclusions suggest that further work is needed to develop a

new model for grief among health care professionals and new support strategies

for them.

Papadatou identified special considerations for health care professionals facing

the death of pediatric patients [36]. These included the level of investment in the

relationship with the patient and family, expectations of the health care profes-

sionals identity and roles, and personal/social constructs. According to Papada-

tou, each death involved loss on one or more levels: loss of relationship with the

patient; loss due to identifying with the pain of the patients family; loss of unmet

goals and expectations of the professional self-image; loss of beliefs and

assumptions about self, life, and death; past unresolved or future anticipated

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485480involved the conflict between experiencing of emotions such as sadness and

openly expressing them, because they felt the latter contradicted norms of

professional behavior. The second, moral distress, occurred when nurses felt

obligated to implement a treatment plan in conflict with their values about pro-

moting a dignified and peaceful death.

Davies et al identified strategies used by nurses to manage distress [37]. These

ranged from strategies that facilitated adjustment including open expression of

grief, creating meaning out of the childs life and death, and peer support and

validation to constrained responses that included diminished engagement with

others, regret, diminished energy, and withdrawal.

A small qualitative study of PICU nurses described the effects of cumulative

exposures to death. Individualized grief response was found to be an ongoing

experiential learning process. Similar themes in how grief was managed emerged:

self-nurturance, termination of the relationship activities, engaging in control-

taking activities, and self reflection [40].

In another study, Papadatou et al compared the grief responses and experi-

ences of Greek physicians and nurses who provided care to children dying of

cancer [39]. Both groups experienced high stress levels; for both, common grief

reactions involved crying, sadness, withdrawal, and recurring thoughts of the

dying process and the childs death. However, there were also differences. The

physicians experienced withdrawal, guilt, and a search for explanations that

would reduce their distress more often than nurses who experienced anger but did

not withdraw. In addition, physicians experienced grief privately and rarely

sought the support of colleagues, while nurses reported connecting with col-

leagues as a way to manage their grief [39].

A study of interventions designed to manage grief gathered quantitative data

from health care professionals at the Johns Hopkins Childrens Center [38]. In

findings that mirrored the qualitative studies cited above, health care profes-

sionals who participated in routine debriefing sessions after a childs death cited

professional distress as the most frequent reason for initiating a session (86%).

They identified the loss of a long-term relationship with the patient as the most

difficult aspect of the case (81%). Other concerns ranked lower: conflict with the

family (25%); sudden/unexpected death, inability to relieve pain, and provision

of aggressive care (2123%).

Because the traditional model of grief may not fit the experience of health care

professionals, attempts have been made to develop newer models of health care

professionals grief. One type of model describes a process of oscillation between

focusing on the loss and moving away from the loss in a dual process model of

coping [36,41] This model of approaching/avoiding the grief response allows the

individual nature of the griever and the social impact of working in health care to

affect the process. The oscillation between the dual processes and resulting fluc-

tuation allow health care professionals to accomplish several tasks: manage their

emotional responses to the death, restore/maintain their integrity, find meaning in

the death, and transcend the suffering to reinvest in life [36,42]. This fluctuation in

process is thought to be a normal and adaptive response that allows health care

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 481professionals to experience grief without becoming overwhelmed by it [36].

Another type of new model shifts from managing disruptive symptoms and

sharing feelings to focus on meaning reconstruction as a way to promote

Going forward several recent initiatives give promise for significant advancesin addressing issues specific to the care of dying children. IPPC, a project of

Educational Development Corporations Center for Applied Ethics and Profes-

sional Practice, is a collaboration with the National Association of Childrens

Hospitals and Related Institutions and eight childrens hospitals [6]. The goal of

the initiative is to enhance the capacity of childrens hospitals to provide family-

centered care in these most difficult circumstances. The IPPC curriculum is

organized around six modules, each including videos featuring interviews with

patients and families discussing their experiences. The six modules are: Engaging

with Children and Families; Relieving Pain and Other Symptoms; Sharing

Decision Making; Improving Communications and Strengthening Relationships;

Responding to Suffering and Bereavement; and Establishing Continuity of Care.

Another recent initiative at Childrens Hospital Boston seeks to rectify the fact

that most clinicians learn about end-of-life care by trial and error. The Program to

Enhance Relational and Communication Skills (PERCS) was developed to teach

communication and relational skills through simulated Difficult Conversations

at the End of Life between trainees and professional patient and parent ac-

tors [45]. Structured scenarios on communicating with angry families; handling

intrafamilial and family-staff conflict relative to the plan of care; discussing

difficult news with families with whom there is no established rapport; talkingrecovery [43]. Constructs such as burnout and compassion fatigue syndrome do

not go far enough in addressing the one core issues of griefthe process of

making meaning out of tragic circumstances. Efforts to implement this new

model acknowledge the need to go further and develop interventions for health

care professionals. Mount [44] suggested preventive and restorative strategies

such as cultivating self awareness, working though past unresolved losses, setting

limits, clarifying team roles and organizational patterns, team support meetings,

and self-care.

A recent study by Rushton et al assessed the effectiveness of palliative care

rounds, patient care conferences, and routine bereavement debriefings in

enhancing health care professionals confidence and competence in providing

palliative and end-of-life care and their ability to manage their grief [38]. Other

activities included meaning-making rituals, such as annual and individual

memorial services and candlelighting ceremonies. Process and impact evaluation

data suggested that greater participation in educational, care planning, and

bereavement activities increased health care professionals perceptions of their

knowledge and skills in palliative and end-of-life care and ability to manage

their grief.

Future directions

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485482with families in times of medical uncertainty; and understanding different cul-

tural views of end-of-life care have been created by a multidisciplinary team of

pediatric critical care physicians, nurses, and as well as psychologists with

this learning environment is that there is never only one right answer to a

communication interaction. The IPPC and PERCS curriculum initiatives reflect

ongoing efforts to creatively and effectively enhance quality end-of-life care in

ment rounds will foster open discussion with all members of the care team on

their perceptions of how well palliative care measures were provided, identify

opportunities for improvement, and allow a forum to express the emotional toll ofproviding such care.

References

[1] Institute of Medicine, Committee on Palliative and End-of-Life Care for Children and Their

Families. When children die: improving palliative and end-of-life care for children and their

families. Washington (DC): National Academy Press; 2002.

[2] McCallum DE, Byrne P, Bruera E. How children die in hospital. J Pain Symptom Manage 2000;

20:41723.

[3] Burns JP, Mitchell C, Outwater KM, Geller M, Griffith JL, Todres ID, et al. End-of-life care in

the pediatric intensive cave unit after the forgoing of life-sustaining treatment. Crit Care Medthe pediatric intensive care unit.

Summary

Although there has been a strong national movement to improve palliative

care, data from the literature reveals that significant perceptual differences exist

among parents, pediatric critical care physicians, and nurses on their assessment

of the quality of end-of-life care as it is actually practiced in the PICU. Whether

these data reflect inadequacies in care as it is actually delivered, or gaps in

communication about that care, or some component of the two, is not known.

Data from the literature do support several measures that can be widely

implemented immediately. For example, more intensive interdisciplinary collabo-

ration, through case review as well as through the identification and support of

positive role models already on staff, can help to promote an emphasis on

competent and compassionate end-of-life care, and assist the bereavement

process for clinicians. Creation of regular, unit-based multidisciplinary bereave-extensive experience in bereavement support. Physician and nurse trainees then

go through each scenario, with the role of patients, parents, and other family

members by actors from the Boston actors guild. The interaction in each scenario

takes place in a simulator suite where it is videotaped, and the scenario is

immediately debriefed, and a framework with specific communication strategies

and skills that can be applied to the actual clinical setting is provided by a team of

psychologists and pediatric critical care physicians. The conceptual framework of

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 4832000;28(8):30606.

[4] Garros D, Rosychuk RJ, Cox PN. Circumstances surrounding end of life in a pediatric intensive

care unit. Pediatrics 2003;112(5):e3719.

[5] Committee on Palliative and End-of-Life Care for Children and their Families. Field MJ, Behr-

man RE, editors. Pediatric palliative care. In: When children die: improving palliative and end-

of-life care for children and their families. Washington (DC): The National Academy Press;

2003. p. 1940.

[6] Solomon MZ, Browning D, Fleischman A, Levetown M, Riegelhaupt L, Rushton CH, et al. The

initiative for pediatric palliative care curriculum faculty manual. Newton (MA): Education

Development Center, Inc.; 2002. Available online at www.ippcweb.org).

[7] National Hospice and Palliative Care Organization. (http://www.nhpco.org/templates/1/

homepage.cfm).

[8] American Association of Colleges of Nursing (AACN) and the Los Angeles-based City of Hope

National Medical Center. The End-of-Life Nursing Education Consortium (ELNEC) project

www.aacn.nche.edu/elnec.

[9] Solomon MZ, Sellers DE, Heller KS, Dokken D, Levetown M, Rushton CH, et al. New research

in pediatric end-of-life care. Presented at The American Society of Bioethics and Humanities

Annual Meeting, Salt Lake City; 2000.

[10] Danis M, Federman D, Fins JJ, Fox E, Kastenbaum B, Lanken PN, et al. Incorporating palliative

care into critical care education: principles, challenges, and opportunities. Crit Care Med

1999;27(9):200513.

[11] Nelson JE, Danis M. End-of-life care in the intensive care unit: where are we now? Crit Care

Med 2001;29(2 Suppl):N29.

[12] Burns JP, Mitchell C, Griffith JL, Truog RD. End-of-life care in the pediatric intensive care

unit: attitudes and practices of pediatric critical care physicians and nurses. Crit Care Med

2001;29(3):65864.

[13] Meyer EC, Burns JP, Griffith JL, Truog RD. Parental perspectives on end-of-life care in the

pediatric intensive care unit. Crit Care Med 2002;30(1):22631.

[14] Truog RD, Cist AF, Bracket SE, Burns JP, Curley MA, Danis M, et al. Recommendations for

end-of-life care in the intensive care unit: the Ethics Committee of the Society of Critical Care

Medicine. Crit Care Med 2001;29(12):233248.

[15] Cook DJ, Guyatt GH, Jaeschke R, et al. Determinants in Canadian health care workers of the

decision to withdraw life support from the critically ill. Canadian Critical Care Trials Group.

JAMA 1995;273:7038.

[16] Tulsky JA, Chesney MA, Lo B. See one, do one, teach one? House staff experience discussing

do-not-resuscitate orders. Arch Intern Med 1996;156:12859.

[17] AACN. Peaceful death: reccommended comptencies and curricular guidelines for end of life

nursing care. (1998) [online] available at http://www.aacn.nche.edu/Publications/deathfin.htm}.

[18] Crego PJ, Lipp EJ. Nurses knowledge of advance directives. Am J Crit Care 1998;7:21823.

[19] Nelson L, Rushton CH, Nelson R, Cranford R, Glover J. Foregoing medically provided hydra-

tion and nutrition in pediatric patients. J Law Med Ethics 1995;23(1):3346.

[20] American Academy of Pediatrics Committee on Bioethics. Guidelines on foregoing life-sustain-

ing medical treatment. Pediatrics 1994;93:5326.

[21] Oberle K, Hughes D. Doctors and nurses perceptions of ethical problems in end of life

decisions. J Adv Nurs 2001;33(6):70715.

[22] Beauchamp TL, Childress JF. Principles of biomedical ethics. New York: Oxford University

Press; 1994.

[23] Wilson WC, Smedira NG, Fink C, et al. Ordering and administration of sedatives and analgesics

during the withholding and withdrawal of life support from critically ill patients. JAMA 1992;

267:94953.

[24] Daly BJ, Thomas D, Dyer MA. Procedures used in withdrawal of mechanical ventilation. Am J

Crit Care 1996;5:3318.

[25] Wolfe J, Grier H, Klar N, Levin S, Ellenbogen J, Salem-Schatz S, et al. Symptoms and suffering

at the end of life in children with cancer. N Engl J Med 2000;342(5):32633.

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485484[26] Yellin PB, Levin BW, Krantz DK, Shinn M, Driscoll Jr JM, Fleischman AR. Neonatologists

decisions about withholding and withdrawing treatments of critically ill newborns. Pediatrics

1998;102:757.

[27] Abe N, Catlin A, Mihara D. End of Life in the NICU: a study of ventilator withdrawal. MCN Am

J Mater Child Care 2001;28(3):1416.

[28] Brody H, Campbell ML, Faber-Langendoen K, et al. Withdrawing intensive life-sustaining

treatment recommendations for compassionate clinical management. N Engl J Med 1997;

336:6527.

[29] Vacco v. Quill 117 S. Ct 2293 (1997).

[30] Quill TE, Dresser R, Brock DW. The rule of double effecta critique of its role in end-of-life

decision making. N Engl J Med 1997;337:176871.

[31] Brody H. Physician-assisted suicide in the courts: moral equivalence, double effect, and clinical

practice. Minn Law Rev 1998;82(4):93963.

[32] Kelly C. Investing or discounting self: are moral decisions shaped by conditions of the work-

J.P. Burns, C.H. Rushton / Crit Care Clin 20 (2004) 467485 485place? Adv Pract Q 1998;4(2):813.

[33] Perkin RM, Resnik DB. The agony of agonal respiration: is the last gasp necessary? J Med

Ethics 2002;28(3):1649.

[34] Truog RD, Burns JP, Mitchell C, Johnson J, Robinson W. Pharmacologic paralysis and with-

drawal of mechanical ventilation at the end of life. N Engl J Med 2000;342(7):50811.

[35] Guidelines 2000 for cardiopulmonary resuscitation and emergency cardiovascular care. Circu-

lation 2000;102(8 Suppl):I869.

[36] Papadatou D. A proposed model of health professionals grieving process. Omega 2000;41(1):

5977.

[37] Davies B, Clarke D, Connaughty S, Cook K, MacKenzie B, McCormick J, et al. Caring for

dying children: nurses experiences. Pediatr Nurs 1996;22(6):5007.

[38] Rushton CH, Hutton N, Reder E, Hall B, Sellers D, Commello K. An action plan to reduce

caregiver suffering in professionals who care for dying children (Abstract). The Initiative for

Pediatric Palliative Care National Symposium, New York; November 7, 2003.

[39] Papadatou D, Papazpglou I, Bellali T, Petraki D. Greek nurse and physician grief as a result of

caring for children dying of cancer. Pediatr Nurs 2002;28(4):34553.

[40] Rashotte J, Fothergill-Bourbonnais & Chamberlain. Pediatric intensive care nurses and their grief

experiences: a phenomenological study. Heart Lung 1997;September/October:37286.

[41] Strobe M, Schut H. The dual process model of coping with bereavement: rationale and descrip-

tion. Death Stud 1999;23(3):197225.

[42] Saunders J, Valente S. Nurses grief. Cancer Nurs 1994;174(4):31825.

[43] Neimeyer RA. Lessons of loss: a guide for coping. New York: McGraw-Hill; 1998.

[44] Mount BM. Dealing with our losses. J Clin Oncol 1986;4(7):112734.

[45] Meyer EM, Browning D, Pascucci RP, Comeau M, Stanley LJ, Burns JP, et al. Anatomy of a

learning team: principles and guidelines to enhance debriefing following communication simu-

lation. Presented at the 4th annual international meeting on medical simulation, Albuquerque/

Santa Fe, NM; January 17, 2004.

End-of-life care in the pediatric intensive care unit: research review and recommendationsThe literature on end-of-life care in the pediatric intensive care unitAttitudes of pediatric critical care physicians and nurses on end-of-life careObservational studies of clinical practice regarding death in the pediatric intensive care unitParental perspectives on end-of-life care in the pediatric intensive care unitImproving end-of-life care in the pediatric intensive care unit: what's to be learned from the literature?

Practical aspects of end-of-life careHow much is too much analgesia for a child dying in the pediatric intensive care unit?Ethical justification for relieving pain and sufferingNeuromuscular blockade and end-of-life careFamily presence at resuscitation attempts

Grief and bereavement of health care professionalsFuture directionsSummaryReferences