PIHJ

Bciojapotiaolsupdto

U

ma6

4

alliative Care in End-Stage Renal Disease:llness Trajectories, Communication, andospice Use

ean L. HolleyPalliative care is comprehensive, interdisciplinary care focusing on pain and symptom management,

advance-care planning and communication, psychosocial and spiritual support, and, in end-stage

renal disease (ESRD), the ethical issues in dialysis decision making. End-of-life care is one aspect of

palliative care and incorporates all of the previously mentioned components as well as hospice and

bereavement care. ESRD patients and their families are appropriate candidates for palliative care

because of their high symptom burden, shortened survival, and significant comorbidity. The usual

pattern of illness trajectory in ESRD is a progressive decline punctuated by episodes of acute

deterioration prompted by sentinel events like limb amputation or myocardial infarction. Such events

provide opportunities for advance-care planning and communication between providers and patients

and families. Although communication is an integral component of palliative care, little is understood

about effective provider-patient communication, especially in estimating and discussing prognosis.

Palliative care has much to offer toward improving the quality of dialysis patients’ lives as well as

planning for and improving the quality of their deaths. The palliative care issues of illness trajectory,

communication, and hospice use among ESRD patients will be reviewed.

© 2007 by the National Kidney Foundation, Inc.

Index Words: Hemodialysis; End-of-life care; Hospice; Chronic kidney disease

ooiptccSEfotdtppn

caatcmleowb

eginning in 1968 with Glaser and Strauss’sdescription,1 patterns of functional de-

line and illness trajectories are increasinglynforming and shaping the study and deliveryf palliative care.2-4 The 3 typical illness tra-

ectories that have been described are (1)brupt, surprise deaths without noticeablereceding decline in functional status (usuallybserved in those dying of cancer); (2) long-erm and progressive disability punctuated byntermittent episodes of serious decline oftenssociated with hospitalization and a failuref the patient to return to his or her previous

evel of function (common in those with organystem impairment like congestive heart fail-re or end-stage renal disease [ESRD]); and (3)rolonged dwindling as occurs in those withementia.1-4 Although trajectories of func-

ional decline can be quite variable, theseverall patterns are commonly observed and

From the Department of Medicine, University of Illinois,rbana-Champaign, and Carle Clinic, Urbana, IL.

Address correspondence to Jean L. Holley, MD, Depart-ent of Medicine, University of Illinois, Urbana-Champaign,

nd Carle Clinic, 602 W. University Avenue, Urbana, IL1801. E-mail: jean.holley@carle.com

© 2007 by the National Kidney Foundation, Inc.1548-5595/07/1404-0012$32.00/0

cdoi:10.1053/j.ackd.2007.07.002

Advances in Chronic Kidney Disease, Vol 102

ffer opportunities for palliative care. As withther examples of organ system failure, the

llness trajectory typical of ESRD is a slow,rogressive decline in functional status punc-

uated by acute or sentinel events. Acute myo-ardial infarctions or limb amputations areommon sentinel events in dialysis patients.entinel events also frequently impair theSRD patient’s return to his/her pre-event

unctional status, thereby contributing to anverall functional decline as well as increasinghe risk of mortality and withdrawal fromialysis.5-7 Recognizing the pattern of illness

rajectory and the occurrence of sentinel eventsrovides opportunities to address goals of care,atients’ symptoms, and advance-care plan-ing, 3 primary components of palliative care.

Components of palliative care in ESRD in-lude identification and management of painnd other symptoms, advance-care planningnd communication, psychosocial and spiri-ual support (including grief and bereavementare), and ethical issues in dialysis decisionaking.8,9 Inherent in these categories of pal-

iative care is end-of-life care, recognizing thatnd-of-life care is only 1 component or aspectf palliative care. In fact, palliative care beginsith the diagnosis of chronic kidney disease

ecause at that point in time care shifts from a

urative to a supportive focus. Supportive or

4, No 4 (October), 2007: pp 402-408

pwhciiptgpiepiifpccaltctcct

Ca

TsWtRcdtromirsipmtrts

aTfspteodphtwwspaptiEtm

taaavem1anwCdUtpnpcncetCvob

403Palliative Care in ESRD

alliative care becomes paramount and theell-being or quality of life of the patient and

is/her family becomes the focus of care. Be-ause the symptom burden of dialysis patientss high10,11 and the dialysis population is ag-ng and has a shortened life expectancy,12

alliative care is appropriate for ESRD pa-ients and their families. A clinical practiceuideline, Shared Decision-Making in the Appro-riate Initiation of and Withdrawal from Dialysis,7

s available to aid nephrologists in reconcilingthical issues that arise in the care of dialysisatients. Although nephrologists are becom-

ng more comfortable addressing such ethicalssues,13 nephrology training programs oftenail to adequately train fellows in end-of-life andalliative care.14 Moreover, although palliativeare is recognized as an aspect of general ESRDare,8,15 the high untreated symptom burdenmong dialysis patients suggests adequate pal-iative care is not being provided.10,11 Articles inhis issue address various aspects of palliativeare for ESRD patients throughout their illnessrajectory. The remainder of this article will fo-us on the topics of communication, advance-are planning, and hospice use in ESRD pa-ients.

ommunicating With ESRD Patientsnd Their Families: Prognosis

he clinical practice guideline, Shared Deci-ion-Making in the Appropriate Initiation and

ithdrawal of Dialysis, includes recommenda-ions based on expert consensus opinion.7

ecommendation number 2 states that physi-ians should fully inform patients about theiriagnosis, prognosis, and all treatment op-

ions.7 The rationale for this recommendationests on ethical principles of patient auton-my. The evidence supporting this recom-endation is observational (level B) and

ncludes data from studies on advance di-ectives in dialysis patients. These studieshow that patients often discuss end-of-lifessues with their families but rarely with theirhysicians and want information about theiredical condition but seldom complete writ-

en advance directives.16-18 The guideline’secommendation number 3 states that a pa-ient’s life expectancy and quality of life

hould be discussed when dialysis is initiated m

nd that some estimate of survival be provided.7

his recommendation is made in order toacilitate informed decision making abouttarting dialysis. Although the ethical princi-les underlying recommendations 2 and 3 of

he guideline are clear, there is little clinicalvidence to either support or refute these rec-mmendations. Studies on provider-patientiscussions at the end of life in a variety ofatient populations have shown that patientsave very individualized desires for informa-

ion and that provider assumptions abouthat patients and families do and do notant to hear are often wrong.19 Observational

tudies of patients with chronic obstructiveulmonary disease, congestive heart failure,nd cancer failed to identify predictors foratients’ preferences for information.19,20 Cul-

ural and ethnic differences in desires fornformation have been documented in non-SRD groups, but, again, few specific iden-

ifiers to guide the amount and type of infor-ation desired were found.19,20

There is little information about ESRD pa-ients’ and families’ wishes for informationbout prognosis. When asked, patients gener-lly want information about their health statend want to engage in discussions about ad-ance-care planning,15-17 but few studies havexamined patients’ wishes for specific infor-ation about estimated survival. A single,

-time survey of 100 Canadian patients duringn initial visit to a nephrology clinic for kid-ey disease reported that 97% of patientsanted explicit information on prognosis.21

onversely, a survey of 152 more ethnicallyiverse chronic hemodialysis patients in thenited States found that although 77% of pa-

ients wanted to “discuss their health com-letely with their doctors including badews,” only 51% wanted to know their actualrognosis.22 Only 13% of these patients re-alled having discussed prognosis with theirephrologists before starting dialysis.22 Noorrelation with patients’ reported perceivedxpected survival (93% expected to live longerhan 2 years) and estimated survival based onharleson Comorbidiy Index (56% 2-year sur-ival) was found.22 Although full disclosuref prognostic information to ESRD patientsased primarily on ethical principles is pro-

oted by some,7,23 additional study of this

if

ppwMsttssapapptcbgmidhirtcppsdtw

CaD

Atctahcsfsta

gswefccbceeqbpvopdiuatsmforvtamcdduanshtdomv

ahhaaP

404 Jean L. Holley

ssue is needed to better define patients’ desireor this information.

In addition to lack of information aboutatients’ wishes for discussions centering onrognosis, there are barriers to nephrologists’illingness to engage in such discussions.ultiple factors influencing dialysis patients’

urvival have been identified and can be usedo predict expected survival.7 Functional sta-us, age, serum albumin, comorbid conditionsuch as diabetes mellitus, and the influence ofentinel events like myocardial infarction ormputation influence dialysis patients’ ex-ected survival.7 Although the presence orbsence of such factors is useful in estimatingrognosis for a population of ESRD patients,redicting the survival of an individual pa-

ient is impossible, regardless of risk factors,linical indices, or scales. In large part, it maye the nephrologist’s inability to predict aiven patient’s expected clinical course thatakes him/her reluctant to provide prognostic

nformation. Although advance-care planningiscussions can proceed without eliminatingope,24 the fear of providing incomplete or

ncorrect prognostic information as well aseluctance to extinguish a patient’s hope seemo reinforce nephrologists’ avoidance of dis-ussions of prognosis. The realization thatrognostic information may be of value toatients and their families should act as atimulus for additional investigation. Evi-ence supporting the guideline recommenda-

ions promoting discussion of prognosisould be welcome.

ommunicating With ESRD Patientsnd Their Families: End-of-Lifeiscussions

s with communicating prognosis to patients,here are little data confirming that communi-ation between health care providers and pa-ients alters outcomes.25 However, nearly allvailable observational studies note thatealth care providers do not adequately dis-uss treatment options and quality of life is-ues and rarely respond to emotional cuesrom patients and families.18,19 The lack of atandardized method to evaluate the effec-iveness of communication between providers

nd patients hampers care as well as investi- s

ation into this area of clinical practice.26 De-pite the lack of evidence-based guidelines onhich a health care provider can build and

valuate communication with patients andamilies, communication remains an integralomponent of palliative care.8,9 Basic skills inommunication can be learned and practicedy health care providers and general tenetsan be followed.26,27 For example, using open-nded questions is necessary to promote andncourage discussions about goals of care,uality of life, and advance-care planning (Ta-le 1). It is also now clear that health careroviders are simultaneously obliged to pro-ide suggestions or recommendations as partf these discussions.26,27 Weiner and Roth20

oint out that presenting value-laden medicalecisions without “right or wrong” answers

n a hypothetical, impersonal manner leads tonnecessary patient and family suffering andctually hinders medical decision making. Ifhe patient is asked to make a medical deci-ion without understanding available treat-ents and their efficacies, decisions are ef-

ectively divorced from potential treatmentutcomes.20 For example, as shown in Table 2,ecommendations may or may not be pro-ided when addressing do not resuscitate sta-us. Although each example concludes withn open-ended question, provider input isuch different (Table 2). When asked, most

hronic hemodialysis patients want to un-ergo cardiopulmonary resuscitation (CPR)espite the poor outcome of CPR in this pop-lation.28 It may well be that such decisionsre made in the absence of informationeeded to make this medical decision. Forhared decision making to be successful,ealth care providers must supply their pa-

ients with the information they need to makeecisions. Outcome data as well as personalpinion based on clinical experience and judg-ent are pieces of relevant information pro-

iders should feel obligated to discuss.In addition to lack of time, lack of training,

nd discomfort with end-of-life discussions,29

ealth care providers exhibit behaviors thatinder end-of-life discussions, including initi-ting discussions about goals of care beforessessing the patient and family’s readiness.20

atient decision making evolves through

tages of precontemplation (they are not yet

r(vtnmtiaica

nsepmwa

CPA

Mafgtnvicsvt

T

A rican S

T

A

405Palliative Care in ESRD

eady to make a decision), contemplationthey need relevant information from the pro-ider to make the decision), and determina-ion/action (they develop a plan).30 Providerseed to recognize this process in decisionaking to engage patients and families effec-

ively and in a timely manner. Until clinicalnvestigation identifies useful guidelines andlgorithms for communicating and engagingn shared decision making, end-of-life dis-ussions will likely continue to be focusedround goals of treatment, advance-care plan-

able 1. Examples of Open-Ended Questions to P

Addressing goals of care

What are your most important hopes?What concerns you most about your illness?What is your quality of life like now?

How do you think about balancing quality of lIs it more important for you to live as long as po

suffering but for a shorter time?What are your biggest fears?Given the severity of your illness, what is most imWhat do you understand about your illness?How much do you want to know?

Advance-care plann

ValuesWhat makes life most worth living for you?

Are there circumstances in which you wouldHave you seen or been with someone who had

What have been the worst and the best thingDirectives

If you are unable to speak for yourself in the fuand values? (proxy directive)

Have you given any thought to what kinds of tunable to speak for yourself in the future? (

Have you considered circumstances in which yWhere would you like to be and who would yo

dapted from the Renal Physician Association and Ame

able 2. Discussing CPR and Do Not Resuscitate

If you were to die suddenly, that is, youstopped breathing or your heart stopped, wecould try to revive you by using CPR. Areyou familiar with CPR? Have you thoughtabout whether you would want it?

Given the severity of your illness, CPR would inall likelihood be ineffective. I wouldrecommend that you choose not to have it, butthat we continue all potentially effectivetreatments. What do you think?

cdapted from Quill.27

ing, do not resuscitate orders, and other life-ustaining treatments and largely remain in-ffective for many, if not most, patients androviders. Development of algorithms oranuals for communication interventions26

ould do much to advance clinical care in thisrea.

ommunicating With ESRDatients and Their Families:dvance-Care Planning

ost dialysis patients engage in some level ofdvance-care planning with their families, butew discuss these issues with their nephrolo-ists.16-18 Like other populations, dialysis pa-ients’ purpose in advance-care planning isot merely the completion of a written ad-ance directive but rather a broader aim that

ncludes preparing for incapacity, achievingontrol, strengthening interfamily relation-hips, and relieving family burdens.18 Ad-ance-care planning is best viewed as an in-egral aspect of a patient’s care plan, a plan of

e Discussions

prognosis is uncertain

h length of life?despite some suffering, or to live without

nt to you to achieve?

d end-of-life issues

ife not worth living?ticularly good (or difficult) death?t this illness for you?

who would be best able to represent your views

ent you would want or not want if you becomection directive or living will)uld want to stop dialysis?to be there when you die?

ociety of Nephrology,7 Weiner et al,26 and Quill.27

romot

when

ife witssible,

porta

ing an

find la par

s abou

ture,

reatminstruou wou like

are that is flexible enough to adjust to pa-

tstvlhtptfctvtsnhocorcctpifppdnasto

H

AwdAohd2dbucEu

dfnthawfuwuheTwicislmtPtidsuhpwtrftsvtailcTspaahna

406 Jean L. Holley

ients’ changing treatment goals and healthtates.31 If dialysis care providers understandhe process of communication underlying ad-ance-care planning, they can address end-of-

ife issues as they arise and as a patient’sealth status warrants. Like other popula-

ions, about a third of dialysis patients com-lete some form of written advance direc-

ive.15-18 Many factors contributed to theailure of advance directives in this country;ommunication between providers and pa-ients about end-of-life issues is poor, no ad-ance directive can address all possible con-ingencies, pertinent information about theuccess and benefits of interventions is oftenot discussed with patients and families, thereas been a focus on interventions rather thann health states during advance-care planningonversations, and such discussions tend toccur during emergent situations rather thanoutinely over time.15,18,27,31 By including dis-ussion of wishes for levels and intensity ofare over an ESRD patient’s life and illnessrajectory, advance-care planning can be ap-ropriately directed, timely, and incorporated

nto the overall care plan for that patient andamily. Simply answering “no” to the “sur-rise question” (Would I be surprised if thisatient died within the next year?) opens theoor to consideration of advance care plan-ing and palliative care. Although barriers todvance-care planning exist within the dialy-is community,17,18,29 they can be overcome ifhe importance of advance-care planning toverall patient care is acknowledged.

ospice and ESRD

s discussed by Cohen and colleagues else-here in this issue, about 20% of United Statesialysis patients stop dialysis before death.12

lthough for most of these patents death willccur within 8 to 14 days,32 few are enrolled inospice; only 42% of patients who stoppedialysis between January 2001 and December002 used hospice.33 Overall, only 13.5% ofialysis patients were enrolled in hospiceefore death.33 The reasons for hospice under-tilization by the dialysis population are in-ompletely understood. Patients with a non-SRD terminal diagnosis such as cancer can

sually be enrolled in hospice and continue a

ialysis. Because both ESRD and hospice areederal benefits, most hospice programs willot accept ESRD patients who want to con-

inue dialysis unless they have a non-ESRDospice-acceptable terminal diagnosis. Thebsence of a non-ESRD terminal diagnosisould make the hospice program responsible

or dialysis costs, a financially prohibitive sit-ation for the hospice program. Patients whoithdraw from dialysis should not presentndue financial or administrative burdens toospice programs and should therefore beligible for the Medicare hospice benefit.hus, it is unclear why nearly 60% of patientsho withdraw from dialysis are not enrolled

n hospice.33 Because the ESRD death notifi-ation form is incomplete in assessment ofnformation concerning dialysis withdrawal,34

ome of the patients in Murray and col-eagues’ study may have died very soon after

aking the decision to stop dialysis, obviatinghe need for or benefits of hospice referral.atients who died in hospice were more likely

o be older, white, able to walk or transferndependently on initiation of dialysis, with-raw for chronic failure to thrive, and live inpecific geographic areas,33 so local hospicese and patient and family preferences forospice are also undoubtedly involved in hos-ice underutilization. Even when dying afterithdrawing from dialysis, patients’ symp-

om burden is relatively high.35 Hospice isecognized for excellence in delivering com-ort care near the end of life so it is reasonableo assume that management of end-of-lifeymptoms would be improved if hospice pro-iders were involved in caring for ESRD pa-ients who are dying. Murray and coworkers33

lso found that median per-patient cost dur-ng the last week of life was significantlyower in the hospice patients, primarily be-ause of their lower rates of hospitalization.hus, hospice is financially advantageous toociety as well as clinically advantageous toatients and families. Because ESRD patientsre appropriate candidates for hospice care,dditional investigation into the barriers toospice enrollment in this population iseeded and opportunities to improve hospiceccess to ESRD patients should be encour-

ged.

S

Stangktrapcmacpcjnttcavfd

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

407Palliative Care in ESRD

ummary

upportive or palliative care begins early inhe course of chronic kidney disease and isppropriate for ESRD patients.7-9,36 Commu-ication with patients and families is an inte-ral component of palliative care, but little isnown about specific benefits of communica-ion and no communication manuals or algo-ithms are yet available to guide discussionsnd assess their effectiveness. Estimatingrognosis in ESRD is helpful for advanceare planning,7,23 but it is not clear howuch patients and families want to know

bout prognosis.19-22,27 Incorporating advanceare planning into overall patient care plans isrudent and allows for interventions and dis-ussions of interventions throughout the tra-ectory of a patient’s life with ESRD. Recog-izing the pattern of ESRD illness trajectory,

he importance of communicating with pa-ients and families, and addressing deficien-ies and barriers in palliative care deliveryllows nephrologists and other dialysis pro-iders to improve the overall care and, hope-ully, the quality of their patients’ lives andeaths.

eferences

1. Glaser B, Strauss AL: Time for Dying. Chicago, IL,Aldine Publishing Co, 1968

2. Lunney JR, Lynn J, Hogan C: Profiles of older Medi-care decedents. J Am Geriatr Soc 50:1108-1112, 2002

3. Lunney JR, Lynn J, Foley DJ, et al: Patterns of functionaldecline at the end of life. JAMA 289:2387-2392, 2003

4. Murray SA, Kendall M, Boyd K, et al: Illness trajectoriesand palliative care. Br Med J 330:1007-1011, 2005

5. Herzog CA, Ma JZ, Collins AJ: Poor long-term sur-vival after acute myocardial infarction among pa-tients on long-term dialysis. N Engl J Med 339:799-805, 1998

6. Eggers PW, Gohdes D, Pugh JA: Nontraumatic lowerextremity amputations in the Medicare ESRD popu-lation. Kidney Int 56:1524-1533, 1999

7. Renal Physician Association and American Society ofNephrology: Shared Decision-Making in the Appro-priate Initiation and Withdrawal of Dialysis: ClinicalPractice Guideline 2. Washington, DC, Renal Physi-cians Association, 2000

8. Moss AH, Holley JL, Davison SN, et al: Core curric-ulum in nephrology: Palliative care. Am J Kidney Dis43:172-185, 2004

9. Chambers EJ, Germain M, Brown E, eds. SupportiveCare for the Renal Patient. Oxford, England, Oxford

University Press, 2004 2

0. Weisbord SD, Fried LF, Arnold RM, et al: Prevalence,severity, and importance of physical and emotionalsymptoms in chronic hemodialysis patients. J Am SocNephrol 16:2487-2494, 2005

1. Davison SN, Jhangri GS, Johnson JA: Cross-sectionalvalidity of a modified Edmonton symptom assess-ment system in dialysis patients: A simple assessmentof symptom burden. Kidney Int 68:1621-1625, 2006

2. U.S. Renal Data System, National Institutes of Health,and National Institute of Diabetes and Digestive andKidney Diseases: U.S. Renal Data System. USRDS2005. 2005 Annual Report: Atlas of End-Stage RenalDisease in the United States. Bethesda, MD, 2005

3. Davison SN, Jhangri GS, Holley JL, et al: Nephrolo-gists’ comfort with end-of-life decision-making. ClinJ Am Soc Nephrol 1:1256-1262, 2006

4. Holley JL, Carmody SS, Moss AH, et al: The need forend-of-life training in nephrology: National surveyresults of nephrology fellows. Am J Kidney Dis 42:813-820, 2003

5. Holley JL: Palliative care in end-stage renal disease:Focus on advance care planning, hospice referral, andbereavement. Semin Dial 18:154-156, 2005

6. Hines SC, Glover JJ, Holley JL, et al: Dialysis patients’preferences for family-based advance care planning.Ann Intern Med 130:825-828, 1999

7. Holley JL, Hines SC, Glover JJ, et al: Failure of ad-vance care planning to elicit patients’ preferences forwithdrawal from dialysis. Am J Kidney Dis 33:688-693, 1999

8. Singer PA: Advance care planning in dialysis. Am JKidney Dis 33:980-991, 1999

9. Tulsky JA: Interventions to enhance communicationamong patients, providers, and families. J Palliat Med8(suppl 1):S95-S102, 2005

0. Weiner JS, Roth J: Avoiding iatrogenic harm to pa-tient and family while discussing goals of care nearthe end of life. J Palliat Med 9:451-461, 2006

1. Fine A, Fontaine B, Kraushar MM, et al: Nephrolo-gists should voluntarily divulge survival data to po-tential dialysis patients: A questionnaire study. PeritDial Int 25:269-273, 2005

2. Chennupati S, Abo-Kamil T, Cubitt A, et al: Actualversus perceived prognosis of hemodialysis patients.J Am Soc Nephrol 16:114A-115A, 2005

3. Michel DM, Moss AH: Communicating prognosis inthe dialysis consent process: A patient-centered,guideline approach. Adv Chronic Kidney Dis 12:196-201, 2005

4. Davison SN, Simpson C: Hope and advance careplanning in patients with end stage renal disease:Qualitative interview study. Br Med J 333:886, 2006

5. SUPPORT Principal Investigators: A controlled trialto improve care for seriously ill hospitalized patients.The Study to Understand Prognoses and Preferencesfor Outcomes and Risks of Treatments (SUPPORT).JAMA 274:1591-1598, 1995

6. Weiner JS, Arnold RM, Curtis JR, et al: Manualizedcommunication interventions to enhance palliativecare research and training: Rigorous, testable ap-proaches. J Palliat Med 9:371-381, 2006

7. Quill TE: Initiating end-of-life discussion with seri-

2

2

3

3

3

3

3

3

3

408 Jean L. Holley

ously ill patients: Addressing the “elephant in theroom”. JAMA 284:2502-2507, 2000

8. Hijazi F, Holley JL: Cardiopulmonary resuscitationand dialysis: Outcome and patients’ views. SeminDial 16:51-53, 2003

9. Perry E, Swartz R, Smith-Wheelock L, et al: Why is itdifficult for staff to discuss advance directives withchronic dialysis patients? J Am Soc Nephrol 7:2160-2168, 1996

0. Levinson W, Cohen MS, Brady D, et al: To change ornot to change: “Sounds like you have a dilemma”.Ann Intern Med 135:386-391, 2001

1. Lynn J, Goldstein NE: Advance care planning for fatalchronic illness: Avoiding commonplace errors and un-

warranted suffering. Ann Intern Med 138:812-818, 2003

2. Neu S, Kjellstrand CM: Stopping long-term dialysis:An empirical study of withdrawal of life-supportingtreatment. N Engl J Med 314:14-20, 1986

3. Murray AM, Arko C, Chen S-C, et al: Use of hospicein the United States dialysis population. Clin J AmSoc Nephrol 1:1248-1255, 2006

4. Holley JL: A single center review of the death notifi-cation form: Discontinuing dialysis prior to death isnot a surrogate for withdrawal from dialysis. Am JKidney Dis 40:525-530, 2002

5. Cohen LM, Germain M, Poppel DM, et al: Dialysisdiscontinuation and palliative care. Am J Kidney Dis36:140-144, 2000

6. Cohen LM, Moss AH, Weisbord SD, et al: Renal pallia-

tive care. J Palliat Med 9:977-992, 2006