Intensive Care Nurses’ and Physicians’ Experiences With Demandsfor Treatment: Some Implications for Clinical Practice

Stephen Workman, Patricia McKeever, William Harvey, and Peter A. Singer

Objective: This study was conducted to develop an

empiric description of intensive care unit (ICU) physi-

cians’ and nurses’ (participants) experiences provid-

ing life-sustaining treatments at the insistence of fam-

ily members, treatments that they believed should

have been withheld or withdrawn. From this descrip-

tion, steps to minimize or prevent their sources of

distress in such situations are suggested.

Design: Semistructured, open-ended interviews. Par-

ticipants were asked to describe cases in which treat-

ment had been provided primarily in response to de-

mands from family members.

Participants: Six physicians and 6 nurses from 6 uni-

versity-affiliated ICUs in Canada. All were members of

a task force developing a multicenter policy to ad-

dress demands for treatment, and physician members

were heads of their ICUs.

Outcome Measures: Systematic analysis of interview

transcripts and synthesis of findings.

Results: Participants recalled 28 cases in which treat-

ment had been provided at the insistence of family

members. Many cases described were very distress-

ing for both medical staff and family members. Con-

sistently problematic areas included: (1) suffering of

dying patients, (2) the marked distress of family mem-

bers, and (3) a breakdown in the relationship between

care providers and families.

Conclusions: Conflict with family members about de-

cisions to limit life-sustaining treatment can be very

stressful for health care providers. Three important

areas that give rise to distress were identified in this

study. These key sources of distress should be looked

for. They could be addressed by: (1) identifying to

family members the importance of minimizing suffer-

ing and ongoing bodily injury of patients at risk for

dying, (2) by doing so addressing directly the distress

of family members by the provision of emotional sup-

port, and when appropriate directed toward helping

them accept that the patient is dying, and (3) pursuing

efforts to maintain or create a good relationship with

family members despite disagreement about the ap-

propriateness of continuing life-sustaining treatment.

Copyright 2003 Elsevier, Inc. All rights reserved.

IN 1996, A MULTIDISCIPLINARY task force,representing 6 university-affiliated intensive

care units (ICUs) in a large Canadian urban center,was formed to address ethical issues in intensivecare. By consensus, a mandate was reached tocreate policy to address demands for treatmentbelieved by clinicians to be futile or inappropriate.The problem of demands was perceived by thisgroup to be the most important ethical problemfaced at that time. We took this opportunity toexplore the experience of providing treatment ow-ing to demands from family members, hoping touncover in part the reasons why this experiencewas seen as so problematic.

Intensive care staff frequently decide that life-sustaining treatment should be withdrawn or with-held (limited).1-7 When patients or proxies persis-tently disagree and demand treatment, “futility”disputes are created.8 Such conflicts can be pro-tracted1 and distressing for both ICU staff andfamily members.9

Initial attempts to limit the requirement to pro-vide treatment described specific situations, suchas providing intensive care to patients in persistentvegetative states,10-12 as futile. Treatment that hasless than a 1% chance of success, or that will neverend dependence on intensive care, has been labeledas futile and, it was suggested, need not be pro-

vided.13,14 Others have argued that the standard ofcare delineates the requirement to provide treat-ment.15,16 More recently, policies have providedprocedural mechanisms to resolve17,18 or preventsuch disputes. Although the futility movement hasnow waned, the requirement for physicians to talkto family members about limiting treatment re-mains as a fundamental requirement.19

Factors that influence ICU physicians’ andnurses’ decisions to withdraw or withhold treat-ment have been described20 as has the process bywhich withdrawal is negotiated with families.21

From the Division of General Internal Medicine, DalhousieUniversity; Faculty of Nursing, Home Care Evaluation andResearch Centre, Program in Bioethics, and the Joint Centrefor Bioethics; University of Toronto; Department of Philoso-phy, MHSC; University Health Network, Canada.

This research was performed while Dr. Workman was a MScstudent in the Institute of Medical Science at the University ofToronto.

Supported in part by a fellowship from the University ofToronto Department of Medicine (S.W.), and by a CanadianInstitutes of Health Research Investigator award (P.A.S.).

Address reprint requests to Stephen Workman, MD, MSc,FRCP(C), Division of General Internal Medicine, DalhousieUniversity, Room 437, Bethune Building, QEII HSC Halifax,Nova Scotia, Canada B3H 2Y9. E-mail: sworkman@is.dal.ca.

Copyright 2003 Elsevier, Inc. All rights reserved.0883-9441/03/1801-0004$30.00/0doi:10.1053/jcrc.2003.YJCRC4

17Journal of Critical Care, Vol 18, No 1 (March), 2003: pp 17-24

The experiences of nurses22 and physicians23 with-holding and withdrawing life-sustaining treatmentalso have been examined. However, to our knowl-edge no study has attempted to determine why theprovision of futile treatment can be such a difficultproblem for medical personnel. Therefore, we ob-tained first-person narrative accounts of futiletreatment to better understand this experience fromthe perspective of medical staff. From the essentialfeatures described by participants in this study, wepropose several approaches to manage difficult as-pects of demands for treatment.

METHODS

DesignICU physicians and nurses from 6 ICUs were asked to relate

their experiences providing treatment that they felt should havebeen withdrawn, but that was provided at the insistence offamily members. The rationale for using narrative data to obtaina rich and complete description of a complex phenomenon hasbeen described previously.24 This study was approved by theUniversity of Toronto Committee on Use of Human Subjects.

SettingParticipants were recruited from a 5-hospital task force rep-

resenting 6 ICUs. The mandate of the task force was to createpolicy to address demands for treatment in ICUs (the policy wascompleted and may be viewed: http://www.utoronto.ca/jcb/Resources/ccm_policy.htm). A description of the developmentof the policy was published recently in January 2001.25 Theprimary researcher and interviewer (S.W.) was a member of thetask force and a general internist who was working at the timein an ICU not affiliated with the task force.

SampleThe sample consisted of one ICU director (physician) and

one nurse from each of the 6 ICUs. One of the units was apediatric ICU. All participants had ongoing research and policyexposure to the topic as well as first-hand experience. Partici-pants had between 5 and 25 years of experience in an ICU. Thephysicians were men, the nurses were women.

Data CollectionParticipants were asked to describe cases in which they had

provided life-sustaining treatment at the insistence of familymember(s), treatment that they felt should not have been pro-vided. Interviews were conducted by the primary researcher inprivate at the ICU or in physician’s offices. Participants recalledfrom 1 to 5 cases each for 28 cases in total. Four cases wereongoing at the time of the interview. Interviews ranged from 60to 90 minutes duration and were transcribed verbatim.

Data AnalysisData were analyzed in 5 steps. First, each interview was read

through to become familiar with the text. Second, each inter-view was coded with identification of thoughts, feelings, and

actions, in keeping with a previously described method forexamining accounts of moral experience.26 Third, commonthemes were developed for each interview by comparing thecompiled thoughts, feelings, and actions. Next, comparisonbetween interviews allowed the development of the commonthemes presented in this study. Finally, the coded transcripts ofphysicians were compared with those of nurses, although thesmall sample size limits the use of this comparison.

The validity of the findings was enhanced in 4 ways. Firstphysician and nurse participants were recruited from 6 ICUs,allowing multiple perspectives of the phenomenon to be sam-pled. Second, the themes and the raw data that gave rise to theirdevelopment were presented to the 3 co-authors (P.A.S., P.M.,W.H.) for confirmation. However, the data were not analyzedindependently by each. Third, participants were provided withan opportunity to review and comment on the results section ofthis article. All read them and none felt that their experienceswere being misrepresented. Finally, although the subject of theinterviews was predetermined, to minimize interviewer bias,participants were allowed to describe at length the experiencesthey found most relevant.

RESULTS

Participants described difficult cases involvingdying patients whose distraught relatives de-manded treatment be continued or initiated. Phy-sicians and nurses voiced similar accounts and hadsimilar concerns, consistent with a prior study thatexamined the ethical concerns of intensive carestaff.27 Representative and succinct quotes, chosenby the principal author, are included for each of the3 main sources of concern described by partici-pants: (1) the suffering of dying patients, (2) dis-tressed family members, and (3) a breakdown inthe relationship with family members. Quotes fromnurses are followed by “RN,” quotes from physi-cians by “MD.” The number of nurses and physi-cians who made statements supporting each of thethemes also is provided.

Dying Patients (4 MDs, 6 RNs)

The suffering of dying patients, as indicated bypain behavior and/or ongoing injury such as isch-emic limb necrosis, was very distressing for nursesand physicians. Although symptom control wasprovided in response to pain behavior, ongoingbodily injury also was problematic. Two nursesnoted that if family members were unaware of theextent or severity of injuries, revealing the body ofthe patient could result in their accepting the ne-cessity to withdraw treatment.

“The patient was slowly losing blood pressureand was developing necrosis, gangrene of the ex-tremities, upper and lower. . .The death took at

18 WORKMAN ET AL

least a good 3 weeks. And I was thinking I couldnever do this to my parents, to any of my familymembers, to any human being. . .Nurses would notwant to care for this patient, they would leave,literally, crying about how bad this was.” (MD)

“Even the one day there was myself and one ofthe other nurses, she—we turned him and I said,‘You know, maybe if we show this daughter hisback, you know, maybe she doesn’t realize whatwe’re talking about here; maybe she’s never seenthat his back is necrotic and he’s rotting in the bedhere.” (RN)

Distressed Family Members andPrognostic Uncertainty (6 MDs, 6 RNs)

The distress of the family members of dyingpatients was witnessed by both nurses and physi-cians and was, in fact, a significant cause of dis-tress for the nurses and physicians in this study.Some family members sometimes struggled verypainfully with the decision to limit treatment.

“One family member made the decision to with-draw treatment, and the other just could not acceptit. I have this life-long sense that we had failedthem, that we may have irreversibly damaged therelationship between them.” (RN)

“I think it’s one of the hardest things I have everseen in my entire life. It [prolonged treatment]causes a lot of pain for the families. You knowwhere they are coming from, [in wanting treatmentcontinued] you feel very strongly for them. Itwould be so nice to stop because we know how itis going to end. [With the death of the patient].”(RN)

“I can’t imagine the anguish of a parent, or howthey would come to utter the words of, youknow. . .’I agree to what you are saying. (aboutwithdrawing treatment).”

A Breakdown in the Relationship WithFamily Members (4 MDs, 5 RNs)

Loss of trust in physicians and a very difficultrelationship with family members was described.

“The family actually threatened the physicianwho was caring for this patient. They said that theywould charge him with murder should the patientdeteriorate. And that if we didn’t resuscitate thepatient, they would charge us with murder. Veryunpleasant, incredible tension of course. . .Finally Isaid the ‘D’ word [dying]. . .And it was like, ‘Oh?No one told us that before.’” (MD)

“Sometimes talking with families about treat-ment withdrawal, I ask myself ‘Am I a monster?’Because the questioning and the attitude of somefamilies [about withdrawing treatment] reallymakes me wonder if I am perceived as a monster,as if I am trying to kill their loved one.” (MD)

“It was very accusatory. They [the family mem-bers] were afraid if you asked them to leave theroom that you would try to bump the patient off.That’s how paranoid and accusatory this familywas.” (RN)

DISCUSSION

The physician and nurse participants in thisstudy described similar sources of concern anddistress arising when family members would notallow life-sustaining treatments to be limited. Wesuggest that in situations of conflict, or potentialconflict, each of these areas be considered by med-ical personnel. Each of the 3 areas of concern isdiscussed later, and some practical suggestions toaddress them are provided.

Patient Suffering

Unrelieved suffering of patients was very dis-tressing for medical personnel. Identifying andcommunicating the strongly felt obligation to pre-vent suffering and respect the bodily integrity ofpatients could create a shared mandate to pursuepatient well-being as an important goal separatefrom curing the patient. Because, “In intensive caresettings, goals seldom follow a strict dichotomybetween cure and palliation,”28 an early emphasison comfort is appropriate. Participants in this studydescribed a strongly felt obligation to minimize orprevent ongoing injury to the bodies of patientsthought to be dying. This obligation also could bediscussed with family members, perhaps as part ofcomfort care.

Distressed Family Members andPrognostic Uncertainty

Some family members were described to beextremely distressed. Explaining to family mem-bers that prolonged treatment and decisions towithdraw treatment can be very difficult could helpcreate a supportive relationship and allow issues ofemotional support to be more openly addressed.Some family members suffered severe emotionaldistress while considering decisions to limit treat-ment and heated conflict within families was de-

DEMANDS FOR TREATMENT 19

scribed, although death was felt to be inevitable byhealth care providers. Because futile treatment hasbeen defined as treatment that will “only prolongthe final stages of the dying process,”15 and be-cause treatment frequently is withdrawn becausethe patient is dying,29 it is puzzling to us that onlyone participant described explaining to familymembers that the patient was dying. Of coursesuch discussions may have occurred and not beenrelated by participants, but we thought it unusualthat such an important aspect of the decision-mak-ing process was not described more frequently.(Only one participant described a case in whichdemands for treatment arose when there was agree-ment between medical personnel and family mem-bers that the patient was dying, and the conflict wasneither heated nor prolonged.) For these reasons,we wonder if it is possible that failure to obtainprognostic consensus caused or exacerbated someconflicts. Because of the small sample size anddesign of this study, further research to determinehow often and effectively that a terminal prognosisis conveyed to family members before discussionsabout treatment withdrawal occur would be neces-sary to better understand these results.

Breakdown in Relationship With FamilyMembers

A severe breakdown in the relationship be-tween ICU staff and family members was veryupsetting. Descriptions of family members ac-cusing health care providers of trying to kill thepatient showed that some health care providerswere perceived to be acting in bad faith. Clari-fying and separating capabilities from intentionscould help mitigate such perceptions. (‘We want,as you do, to make your loved one better, butlack the ability to do so.’)The negotiation liter-ature shows that simply meeting demands fortreatment will not improve strained relation-ships, and that specific efforts to build or main-tain a good relationship with family membersmust be identified and pursued.30 If trust alreadyhas begun to be lost, simple statements andquestions such as, ‘I am concerned you don’ttrust us. Why? What can we do to help?’ couldbe useful to start the process of relationshiprebuilding. One physician suggested that findinga health care provider who is trusted also wasuseful.

Limitations

This article has 3 main limitations. First, it wasconducted with a small select sample of experi-enced nurses and physicians who perceived con-flict over limiting treatment to be an importantproblem. Their views and experiences may notrepresent the majority and the information theyprovided about the experience of family membersis clearly second hand. Second, participants werefrom a homogeneous environment, all were fromCanadian tertiary care academic ICUs in a citywith a very high (�90%) ICU occupancy rate.Finally, empiric studies examining the experienceof treatment withdrawal from the perspective offamily members are lacking and the experience offamily members was not examined directly. Onerecent study has shown that conflict between ICUstaff and family members is perceived commonlyby family members.31

CONCLUSIONS

The nurses and physicians who participated inthis study described conflict arising from conflictover decisions to limit treatment to be very dis-tressing, especially when the patient was perceivedto be suffering, or there was a breakdown in therelationship with distraught family members. Byhighlighting these 3 areas of concern, this empiricstudy allowed steps directed at conflict over limit-ing treatment to be addressed more fully and me-thodically. As a first step, in situations of conflictor potential conflict, physicians and members ofthe health care team should ask themselves 3 ques-tions:

1. Has the importance of ensuring patient com-fort and maintaining the patient’s bodily in-tegrity been examined and discussed withfamily members?

2. Are family members extremely distressed?Do they require more emotional support? Ifso, has prognostic consensus been reached, oris the possibility of treatment withdrawal dis-tressing because family members believe thatsurvival remains possible?

3. Has there been a breakdown in the relation-ship between family members and medicalpersonnel? What steps, besides providingtreatment, can be taken to address this break-down.

Addressing these questions before initiating or

20 WORKMAN ET AL

continuing discussion about limiting treat-ment could help health care providers to ad-dress, if not resolve, these sources of theirdistress. Reducing caregiver distress could

create an environment more conducive to fullyaddressing decisions to limit treatment and tothe provision of compassionate end-of-lifecare.

REFERENCES

1. Prendergast TJ, Luce JM: Increasing incidence of with-holding and withdrawal of life support from the critically ill.Am J Respir Crit Care Med 155:15-20, 1997

2. Keenan SP, Busche KD, Chen LM, et al: A retrospectivereview of a large cohort of patients undergoing the process ofwithholding or withdrawal of life support. Crit Care Med 25:324-331, 1997

3. Balfour-Lynn IM, Tasker RC: At the coalface—medicalethics in practice. Futility and death in paediatric medical in-tensive care. J Med Ethics 22:279-281, 1996

4. Keenan SP, Busche KD, Chen LM, et al: Withdrawal andwithholding of life support in the intensive care unit: A com-parison of teaching and community hospitals. The Southwest-ern Ontario Critical Care Research Network. Crit Care Med26:245-251, 1998

5. Prendergast TJ, Claessens MT, Luce JM: A national sur-vey of end-of-life care for critically ill patients. Am J RespirCrit Care Med 158:1163-1167, 1998

6. Smedira NG, Evans BH, Grais LS, et al: Withholding andwithdrawal of life support from the critically ill. N Engl J Med322:309-315, 1990

7. Wood GW, Martin E: Withholding and withdrawing life-sustaining therapy in a Canadian intensive care unit. Can JAnaesth 42:186-191, 1995

8. Prendergast TJ: Futility and the common cold. How re-quests for antibiotics can illuminate care at the end-of-life.Chest 107:836-844, 1995

9. Prendergast TJ: Resolving conflicts surrounding end-of-life care. New Horiz 5:62-71, 1997

10. American Thoracic Society Bioethics Task Force: With-holding and withdrawing life-sustaining therapy. Am Rev Re-spir Dis 144:726-731, 1991

11. Faber-Langendoen K, Caplan AL, McGlvae PB: Sur-vival of adult bone marrow transplant patients receiving me-chanical ventilation: A case for restricted use. Bone MarrowTransplant 12:501-507, 1993

12. Society of Critical Care Ethics Committee: Consensusstatement on the triage of critically ill patients. JAMA 271:1200-1204, 1994

13. Schneiderman LJ, Jecker NS, Jonsen AR: Medical futil-ity: Its meaning and ethical implications. Ann Intern Med112:949-954, 1990

14. Schneiderman LJ, Jecker NS, Jonsen AR: Medical futil-ity: Response to critiques. Ann Intern Med 125:669-674, 1996

15. Weijer C, Singer PA, Dickens B, et al: Bioethics forclinicians: 16. Dealing with demands for inappropriate treat-ment. CMAJ 159:817-821, 1998

16. Weijer C, Elliott C: Pulling the plug on futility [Edito-rial]. BMJ 310:683-684, 1995

17. Halevy A, Brody BA: A multi institution collaborativepolicy on medical futility. JAMA 276:571-574, 1996

18. Council on Ethical and Judicial Affairs, American Med-ical Association: Medical futility and end-of-life care. JAMA281:937-941, 1999

19. Helft PR, Siegler M, Lantos J: The rise and fall of thefutility movement. N Engl J Med 343:293-296, 2000

20. Cook DJ, Guyatt GH, Jaeschke R, et al: Determinants inCanadian health care workers of the decision to withdraw lifesupport from the critically ill. Canadian Critical Care TrialsGroup. JAMA 273:703-708, 1995

21. Miller DK, Coe RM, Hyers TM: Achieving consensus onwithdrawing or withholding care for critically ill patients. J GenIntern Med 5:475-480, 1992

22. Jones T, Fitzgerald M: Withdrawal of life-support treat-ment: The experience of critical care nurses. Aust Crit Care11:117-121, 1998

23. Asai A, Fukuhara S, Inoshita O, et al: Medical decisionsconcerning the end-of-life: A discussion with Japanese physi-cians. J Med Ethics 23:323-327, 1997

24. Greenhalgh T, Hurwitz B: Why study narrative? BMJ318:48-50, 1999

25. Singer PA, Barker G, Bowman KW, et al: Hospitalpolicy on appropriate use of life sustaining treatment. Crit CareMed 29:187-192, 2001

26. Tappan MB: Hermeneutics and moral development: In-terpreting narrative representations of moral experience. DevRev 10:239-265, 1990

27. Soderberg A, Norberg A: Intensive care: Situations ofethical difficulty. J Adv Nurs 12:2008-2014, 1993

28. Brody H, Campbell ML, Faber-Langendoen K, et al:Withdrawing intensive life-sustaining treatment—recommen-dations for compassionate clinical management. N Engl J Med336:652-657, 1997

29. Manara AR, Pittman JA, Braddon FE: Reasons for with-drawing treatment in patients receiving intensive care. Anaes-thesia 53:523-528, 1998

30. Fisher R Brown S: Getting together: Building relation-ships as we negotiate. Boston, Houghton Mifflin, 1988

31. Abbott K, Sago J, Breen C, et al: Families looking back:One year after discussion of withdrawal or withholding life-sustaining support. Crit Care Med 29:197-201, 2001

DEMANDS FOR TREATMENT 21