●

MeRtic0prpHcti©

IA

OUPsma

mvMtpetrppsaidvotfacs

A

Peer Mentoring: A Culturally Sensitive Approach to End-of-LifePlanning for Long-Term Dialysis Patients

Erica Perry, MSW, June Swartz, MA, Stephanie Brown, PhD, Dylan Smith, PhD,George Kelly, MSW, and Richard Swartz, MD

Background: This study is designed to explore the impact of peer mentoring on end-of-life decision making.ethods: A controlled randomized intervention study with 203 patients from 21 dialysis centers across Michiganxplored the impact of peer mentors, dialysis patients trained to help other patients, on end-of-life planning.esults: Communicating information on advance directives (ADs) through peer mentoring significantly influenced

he completion of ADs overall compared with distributing standard printed material or no specific designedntervention. However, the influence was most prominent among African Americans, not only increasing actualompletion of ADs (P < 0.001) and comfort discussing ADs (P < 0.01), but also improving subjective well-being (P <.05) and anxiety (P < 0.05) during the study period. These effects of peer mentoring did not appear among whiteatients, although printed material on ADs decreased reported suicidal ideation (P < 0.05). Conclusion: Theseesults suggest the importance of addressing specific cultural factors in the process of AD education. Commonractice assumes that printed materials are effective in educating patients about health care and decision making.owever, peer mentoring, a relationship-centered person-to-person approach, may be more effective in someultural groups because it partakes of oral, rather than written, traditions. Acknowledging cultural differences andailoring our approach could be powerful in enhancing trust and participation and decreasing potential disparitiesn health care outcomes. Am J Kidney Dis 46:111-119.

2005 by the National Kidney Foundation, Inc.

NDEX WORDS: Quality of life; end-stage renal disease (ESRD); health care disparity; peer mentors; race; African

merican; advance directives (ADs).

edaigcbrf

vSv

M

o

aI

ttM

KA

F MORE THAN 450,000 patients withend-stage renal disease (ESRD) in the

nited States, more than 79,000 died in 2004.1

atients beginning dialysis therapy are older andicker than ever before. Because of the highortality rate, death and end-of-life discussions

re critical issues for the dialysis community.The Patient Self-Determination Act of 19902

andates end-of-life education by offering ad-ance directives (ADs) to all patients admitted toedicare-participating hospitals. This legisla-

ion was intended to ensure the possibility ofatient input for medical decisions concerningnd-of-life care. In practice, patients being admit-ed to hospitals usually are offered printed mate-ials and given the opportunity to execute ADapers, often by personnel unfamiliar to theatient. The act did not address the outpatientetting or maintenance dialysis units. Severalrticles question the effectiveness of ADs, point-ng out that ADs often are not taken into accounturing acute hospital care,3-7 ADs may be tooague to be of use in specific clinical situationsr may lead caregivers to give up prema-urely,4,7,8 execution of ADs does not alwaysoster increased discussion between the patientnd family or caregivers,5,9,10 and ADs may haveultural biases that do not address the needs of

uch groups as African Americans.11,12 How-

merican Journal of Kidney Diseases, Vol 46, No 1 (July), 2005: p

ver, other research suggested that patients whoiscussed their end-of-life wishes with familynd staff had an increased likelihood of experienc-ng “a good death” irrespective of cultural back-round.10 In addition, AD completion is in-reased when a trusted health care staff memberroaches the subject, particularly when contact isecurring9,13 and takes place in such settings asamily meetings. In the final analysis, the process

From the National Kidney Foundation of Michigan; Uni-ersity of Michigan Division of Nephrology and Institute forocial Research; and Veterans Administration Health Ser-ices Research and Development Center, Ann Arbor, MI.

Received January 13, 2005; accepted in revised formarch 31, 2005.Originally published online as doi:10.1053/j.ajkd.2005.03.018

n May 23, 2005.The contents are solely the responsibility of the authors

nd do not represent the official views of the Nationalnstitute of Mental Health.

Supported in part by the Robert Wood Johnson Founda-ion; National Kidney Foundation of Michigan; and Na-ional Institute of Mental Health Career Grant no. K01-

H065423 (S.B.).Address reprint requests to Erica Perry, MSW, National

idney Foundation of Michigan, 1169 Oak Valley Dr, Annrbor, MI 48108. E-mail: eperry@nkfm.org© 2005 by the National Kidney Foundation, Inc.0272-6386/05/4601-0014$30.00/0

doi:10.1053/j.ajkd.2005.03.018

p 111-119 111

omin

awtpcwacsictsmeatm

umphlewwpIEfMabmhfaaopeti

i

mqtfiiewf

P

oaMsyrwIH(ai

S

dttFFgpmrowtqr

fuciaihaatp(at

PERRY ET AL112

f end-of-life discussion, rather than the docu-ent itself, may be most important in determin-

ng successful and reconciled outcome in termi-ally ill patients with ESRD.9,10,13

Our standard methods, such as written materi-ls or interactive Internet programs, do not al-ays address cultural differences, and a people-

o-people approach might be more effective. Arominent example in the United States is Afri-an Americans, who are 4 times more likely thanhites to develop ESRD, face a similar 15%

nnual mortality rate,1 and often are reluctant toomplete ADs.14 This reluctance comes as nourprise because there is well-recognized dispar-ty in both availability and access to healthare,15,16 and African Americans often are dis-rustful of the medical system16-19 because ofuch experiences as the Tuskegee Syphilis Experi-ent.20,21 Furthermore, there are cultural differ-

nces, such as family-oriented decision makingnd oral traditions,22,23 that we have failed toake into account when designing health care or,ore specifically, AD education.An important people-to-people method now

sed in the long-term dialysis setting is peerentoring, training of selected patients to sup-

ort and empower other patients. Peer mentoringas proved beneficial not only to patients onong-term dialysis therapy,24,25 but also in sev-ral other medical settings, including patientsith breast and prostate cancer, breastfeedingomen, women with postpartum depression, andatients with acquired immune deficiency.26-39

n Michigan, peer mentoring for patients withSRD is a statewide clinical service program

ostered by the National Kidney Foundation ofichigan. A 16-hour training program teaches

ctive listening and problem-solving skills com-ined with focused discussion on human develop-ent and mental health. Peer mentors have been

ighly effective in helping alleviate patient’sears about renal transplantation and helpingdolescents cope while on dialysis therapy andfter renal transplantation.24,25,40-46 Because theyften have faced multiple complications with theotential of dying during their own medicalxperiences, peer mentors are in a unique posi-ion to broach discussion of end-of-life and ADssues with other patients.

The present study is designed to explore the

mpact of peer mentoring on end-of-life decision i

aking. We propose to address the followinguestions: (1) Is peer mentoring more effectivehan other methods for increasing patient com-ort in discussing end-of-life issues and complet-ng ADs? (2) Are there demographic differencesn the effectiveness of peer mentoring? (3) Is ADducation through peer mentoring associated withorsened or improved psychosocial outcomes

or patients?

METHODS

articipantsTwo hundred eighty patients from 21 dialysis units (units

f varying size selected on the basis of social workervailability and willingness to participate) across the state ofichigan met initial eligibility criteria: patients needed to

peak English, be assessed as competent, be older than 18ears, and not yet have completed an AD. All patients wereecruited by unit social workers, and signed consent formsere approved, along with other study materials, by the

nvestigational Review Board of the University of Michiganealth System (Ann Arbor, MI). Subjects were largely

�95%) naive to peer-mentor intervention before this studynd were asked to participate in a study of both peerntervention and ADs.

tudy DesignExperimental Conditions. Patients were assigned by ran-

om lots to receive AD information in 1 of 3 ways: (1)hrough peer mentoring (group 1; peer intervention), (2)hrough printed material prepared by the National Kidneyoundation (“Advance Directives: A Guide for Patients andamilies,” question-and-answer format, targeted at 8th-rade reading level) distributed approximately at the mid-oint of the 2- to 4-month study period (group 2; printedaterial), or (3) through no additional means other than that

outinely provided by the dialysis unit (group 3; control). Nother study-related intervention was planned, but unit socialorkers performed their usual intake interview for all pa-

ients starting dialysis therapy and were free to answeruestions that came up during the study period regardless ofandomized study assignment.

Survey. Patients responded to a baseline survey andollow-up survey 2 to 4 months later. Surveys were distrib-ted and collected by dialysis unit social workers andontained psychosocial measures that included the follow-ng: depression, subjective well-being, death acceptance,nxiety, and suicidal thinking. Depression (6 questions assess-ng such symptoms as dysphoria, somatic symptoms, andopelessness), anxiety (2 items measuring such symptomss feeling trapped), and suicidal ideation (1 question) weressessed by using a modified version of the Hopkins Symp-om Checklist.47 The severity of each symptom during therevious 2-week period was rated on a scale varying from 1not at all) to 5 (extremely). Subjective well-being wasssessed based on patient ratings by using 5 statements fromhe Diener scale48 assessing current level of life satisfaction,

ncluding “the conditions of my life are excellent” and “I am

s1wtldmpsr

bcamIAp

idcafouseo

gbiet2pfschpp

P

ppaawpp2Wnaitom

Atsitmova

ap3v“oyt

S

alA

PEER MENTORING IN ESRD END-OF-LIFE PLANNING 113

atisfied with my life,” rated on a 5-point scale ranging from(strongly disagree) to 5 (strongly agree). Death acceptanceas measured by using 5 items that asked patients whether

hey believed death to be a natural part of the processes ofife and whether they were afraid of dying, using the sameisagree-agree scale.49 The entire survey took approxi-ately 30 minutes and usually was completed while the

atient was undergoing dialysis. Social workers read theurvey to a small number of patients who had visual oreading problems.

The follow-up survey was completed by 83% of originalaseline subjects. Researchers were available during theourse of the study to address questions or problems thatrose among peer mentors or social workers. All surveyaterials were collected and processed anonymously at the

nstitute for Social Research at the University of Michigan.t completion of the study, social workers, peers, andatients were paid a modest amount for their participation.Outcomes. Social workers at individual units provided

nformation about whether the patient had completed an ADuring the study period, whether the patient wanted toomplete an AD if they had not done so, and how comfort-ble the patient appeared to be in discussing ADs (on a scalerom 1 [not comfortable] to 5 [very comfortable]). These ADutcomes were based on direct assessment by the individualnit social workers at the end of each subject’s course on thetudy (2 to 4 months). Social worker roles in the study werexplained at a special workshop, but additional verificationf outcomes was not undertaken after the study.In 5 instances, patients assigned to the peer-intervention

roup were moved to a control group by the social workerecause of dialysis-schedule conflicts that prevented peerntervention. In addition, the peer mentor at 1 unit became illarly in the study and could not carry out the peer interven-ion; therefore, patients at this unit were assigned to the other

groups. Of 237 patients who agreed to participate, 203atients completed all phases of the study and providedollow-up data. Thirty-four patients could not complete thetudy because of complications or changes in clinical statusommon among patients with ESRD, including death orospital admission (25 patients), kidney transplantation (6atients), and transfer to a different dialysis facility (3atients).

eer-Support InterventionMany participating social workers and all 17 participating

eers attended a regional AD workshop. In the workshop,eer mentors were given a packet of materials that includedconsent form to participate in the study, the survey to be

dministered to study participants, a research timeline, aorksheet that outlined discussion topics during the studyeriod, and several literature publications to share witharticipants. They also viewed the “Five Wishes” video, a5-minute videotape on ADs developed through the Robertood Johnson Foundation in which Mother Theresa’s attor-

ey compares dying in a respectful way in India to dyinglone in American hospitals. This video presents the possibil-ty of creating a peaceful dying process through the use ofhe “Five Wishes” document and explains the legal aspectsf AD documents.50 In the course of the workshop, peer

entors discussed facts, myths, and barriers to completing s

Ds, and filled out their own ADs and shared their personalhoughts about the process during the intervention. Peershowed their AD knowledge through pretesting and posttest-ng and role-playing. ADs were open ended (ie, not justermination of treatment, but also directive of specific treat-ents), and peer mentors were trained to speak from their

wn experience, but not to judge or dictate a specific point ofiew. Bias was a potential issue, but training was aimed atddressing bias to the degree possible.

During a 2- to 4-month period of the study, patientsssigned to receive peer intervention were contacted byeers on 8 occasions that included 5 telephone contacts andface-to-face meetings. Patients and peers discussed the

alue of AD completion as a contribution to loved ones, eg,you would want to know what was important to your lovednes, just as they would want to know what was important toou.” Peer mentors asked patients to share any reason thathey might be concerned or worried about filling out an AD.

1. During the first face-to-face contact, the peer intro-duced the peer-mentoring program and made arequest of the patient to read a 1-page article, “TheSense of a Goose,” to be discussed at the subsequentcontact. The peer also brought up confidentialityissues and the study schedule.

2. During the second contact (accomplished by tele-phone), the peer asked the patient about the value ofthe article. The peer then spoke personally about howchronic illness initially affected his or her life andthat one can only do the best one can “given the handlife deals.”

3. The third contact discussed what has made lifeworthwhile since having a chronic illness. Thisshifted focus and goals to something outside ofkidney failure.

4. The fourth contact addressed what gave both thepatient and the peer strength during hard times,touching on subjects’ spiritual orientation and discuss-ing specific fears.

5. The fifth contact was in person. The peer shared hisAD and end-of-life considerations and initial barriersto doing an AD, role modeling how one can workthrough concerns. A blank Five Wishes AD documentwas offered to the patient.

6. During the sixth contact, the patient and peer dis-cussed the value of contributing to others, ways thepeer and patient contributed to others in life, and howan AD was a contribution to others.

7. The seventh contact was not structured, and thepatient was asked to choose the discussion topic,continuing to foster interdependence.

8. The last in-person contact shared the peer’s writtenobservations of the patient’s strengths, and the peeroffered to be available to the patient after the study.

tatistical AnalysisBecause the patient sample included responses of patients

t similar dialysis clinics, we computed the intraclass corre-ation for the clinic-level effect on our primary outcome ofD completion by using the recursive generalized least-

quares estimation for multilevel binomial models (MLwIN

sUc�alSt

dr3bl1dim2At

pppmteri(g(ooc(

fsApncisMasc1v

aLtr

GGG

cm

PPC

dArd

PERRY ET AL114

oftware, version 1.1; 2000; Institute of Education, London,K). Results of this procedure indicated there was no

linic-level effect on AD completion (intraclass correlation0.04; not significant), supporting treatment of individuals

s independent sources of data without significant clinicocation effects. Statistical analysis used logistic regression,tudent t-test, and analysis of variance (ANOVA) or mul-

iple ANOVA, as deemed appropriate.

RESULTS

Subject characteristics in the experimental con-ition groups are listed in Table 1. Overall ageanged from 19 to 83 years, 49% were women,8% were African American, and educationalackground varied from high school diploma oress in 30% to a college or graduate degree in8%. There were no significant differences inistributions of these characteristics or underly-ng causes of chronic renal failure among experi-ental groups. Peer mentors ranged in age from

9 to 81 years and included 11 whites and 6frican Americans. Study results did not vary on

he basis of peer-mentor racial identity or age.As listed in Table 2, a greater number of

Table 1. Characteristics of Patients AcrossExperimental Conditions

Women(%)

AfricanAmerican

(%)Mean Age

(y)

roup 1: peer intervention 54 39 44 (20-83)roup 2: printed materials 46 39 44 (23-74)roup 3: control 46 37 45 (20-80)

NOTE. There were no significant differences in theseharacteristics among conditions. Values expressed asean (range) or percent.

Table 2. Logistic Regression

Experimental Condition Completed AD

eer intervention (group 1) 22/63 (35%)rinted materials (group 2) 7/59 (12%)ontrol (group 3) 8/81 (10%)

NOTE. See text for details.*Peer support (group 1) is the reference group for the odd†A small number of patients did not have follow-up data

id not complete an AD, 37 of 41 patients had follow-up daD. In group 2, a total of 52 patients did not complete an AD

eported a desire to complete an AD. In group 3, a total of 7ata, and 26 of 49 patients reported a desire to complete an‡Significant at P � 0.01, see text.

§Significant at P � 0.05, see text.

atients receiving peer mentoring (group 1) com-leted an AD during the course of the studyeriod compared with patients receiving written-aterial (group 2) or those in the control condi-

ion (group 3). Using 2 logistic regression mod-ls, AD outcomes were each (in separate models)egressed on dummy codes that compared peerntervention (group 1) with both written materialgroup 2) and controls (group 3). Compared withroup 1 (peer intervention), patients in group 2printed materials) or group 3 (controls) werenly one fourth as likely to complete an AD andnly one third as likely to report a desire toomplete an AD, even if they had not yet done soP � 0.05 or P � 0.01; Table 2).

Survey questions regarding the level of com-ort discussing ADs with staff (reported on acale from 1 to 5) were analyzed by using 1-wayNOVA. Results indicated that patients in theeer-intervention condition (group 1) had a sig-ificantly greater level of comfort with AD dis-ussion (mean score, 4.21 � 1.18) than patientsn groups 2 (mean score, 3.63 � 1.23) or 3 (meancore, 3.63 � 1.27; F2,196 � 4.67; P � 0.05).ean differences between actual response scores

lso were compared by using Student t-test andhowed that the intervention group was signifi-antly different from the other 2 conditions (groupversus group 3, t140 � 2.75; P � 0.01; group 1ersus group 2, t118 � 2.61; P � 0.05).

Table 3 lists frequencies of AD completion asfunction of race and experimental condition.ogistic regression analysis in which AD comple-

ion was regressed on experimental condition,ace, and the interaction of experimental condi-

ls Predicting AD Outcomes

Ratio*Desire to Complete AD of

Those Not Doing So† Odds Ratio

25/37 (68%)5‡ 20/49 (41%) 0.33§0‡ 26/69 (38%) 0.29‡

.e excluded. For example, in group 1, a total of 41 patients

25 of these 37 patients reported a desire to complete an52 patients had follow-up data, and 20 of these 49 patientsts did not complete an AD, 69 of 73 patients had follow-up

Mode

Odds

0.20.2

s ratioand arta, and, 49 of

3 patienAD.

to1Awcrpicstei

(Aofci�ieamo21

ue

cicAtm(dcpw0c

iotAoeAtttnt

sbfpigbdgaawPtoite

ol

P

P

C

iqlte“Tii

PEER MENTORING IN ESRD END-OF-LIFE PLANNING 115

ion and race indicated a significant interactionf intervention condition and race (odds ratio,5.9; P � 0.01). As listed in Table 3, Africanmericans receiving peer intervention (group 1)ere most likely to complete an AD during the

ourse of the study, whereas African Americanseceiving printed material (group 2) did not com-lete an AD during this period. Conversely, peerntervention did not appear to impact on ADompletion among white participants. Results ofubsequent logistic regression models indicatehat other demographics, such as sex, age, orducational level, did not similarly modify thentervention effect on AD completion.

Given these racial differences, we used a 2race) � 3 (intervention condition) multipleNOVA to examine effects of the interventionn time-dependent changes (from baseline toollow-up survey) in mean scores for psychoso-ial outcomes. There is a marginal multivariatenteraction of race and condition overall (F10,320

1.843; P � 0.053). Follow-up univariate testsndicated a significant racial difference in theffect of intervention on change in suicidal ide-tion (F2,163 � 3.22; P � 0.05), as well asarginal race by experimental-condition effects

n changes in subjective well-being (F2,163 �.34; P � 0.10) and death acceptance (F2,163 �.79; P � 0.10).

As shown in Figure 1, additional analysissing a series of 1-way ANOVAs to examine the

Table 3. Frequency of AD Completion as a Functionof Race and Experimental Condition

Experimental Condition Race AD No AD

eer intervention Black 10 12White 12 26

rinted material Black 0 20White 7 27

ontrol Black 2 29White 6 35

NOTE. See text for details. There was a significantnteraction of race and experimental condition on fre-uency of completing AD (odds ratio, 15.9; P � 0.01 using

ogistic regression). Patients for whom racial designation inhe baseline survey was other than black or white, notntered, or ambiguous (�1 race chosen) were treated asmissing data” and were not included in the analysis inable 3. There were 17 such patients (9%): 3 patients (5%)

n group 1, 5 patients (8%) in group 2, and 9 patients (11%)n group 3.

ffect of the intervention on each psychosocial w

hange score within each race indicated that peerntervention was associated with both a signifi-ant increase in subjective well-being in Africanmericans compared with the other experimen-

al conditions (F2,75 � 3.12; P � 0.05) and aarginal tendency toward decreased anxiety

F2,75 � 2.56; P � 0.084) and slightly greatereath acceptance that was not statistically signifi-ant (F2,68 � 1.89; P � 0.16). Analysis of whiteatients showed a decrease in suicidal ideationith use of printed material (F2,103 � 4.79; P �.05), but no other notable changes in psychoso-ial variables over time.

DISCUSSION

The present study shows that peer interventions associated with a significant effect on ADutcomes, including completion of ADs, desireo complete ADs, and level of comfort discussingDs. The major effect of intervention on ADutcomes was accounted for by a racial differ-nce in response to peer mentoring. Specifically,frican Americans were more likely than whites

o complete ADs as a function of peer interven-ion. Printed materials may have deterred comple-ion of ADs among African Americans, but hado specific effect on whites compared with con-rols.

Because the present findings suggest that peerupport may resonate for African Americans inroaching such sensitive subjects as ADs, we areorced to question the predominant use of pre-ared written (or electronic) materials. Somenvestigators maintained that African Americansenerally do not complete ADs,14,51 but it maye the process that fails to make end-of-lifeiscussion meaningful. As Holley et al9 sug-ested, printed material does not appear to changettitudes or foster completion of ADs without thedded input of trusted health care personnel withhom patients have some ongoing relationship.resent results suggest that this is particularly

rue for African Americans, and peer support isne way to provide the necessary additionalnput. In the final analysis, the process, and nothe AD document itself, may be the pivotallement in meaningful end-of-life discussion.

The impact of our experimental interventionsn psychosocial variables differed along racialines. Over time, peer support was associated

ith improved subjective well-being and lower

aPrilvwnttllAcgmw

dse

iotoewoRiebm

pdcg 6 for wo

PERRY ET AL116

nxiety in African Americans, but not whites.revious reports suggested a strong and oftenacially distinct influence of social support,52

nterpersonal relationships,53 and religious be-iefs or spirituality53,54 on such psychosocialariables as depression, anxiety, or subjectiveell-being. For example, in studies with a promi-ent majority of African-American dialysis pa-ients, Kimmel52,55 reported that patient percep-ions of increased social support correlated withess depression and more global satisfaction withife. Unruh et al56 suggested that African-merican patients with ESRD manifested signifi-

antly better psychological well-being than otherroups. Lopes57 and others53 also found thatinorities reported a better quality of life than

Fig 1. (A-E) Changes in mean survey scores frosychosocial variable shown individually in each grapescribed in text. Surveys that were incompletely filleompleted surveys for each group were as follows: groroup 2 (printed material), 18 of 23 for blacks and 33 of 3f 49 for whites.

hites. It comes as no surprise then to find a

ifferences in how patients respond to AD discus-ion, differences that originate in patient culturalxperiences.

There are several explanations for our find-ngs, the most compelling of which centers onral traditions and transmission of informationhrough personal contact and story telling thatriginate in Africa.11,22,58 Griots were storytell-rs within certain West African cultures whoere responsible for passing down history fromne generation to another. In his forward tooots, Jay Haley acknowledges the profound

mpact of Griots on the transmission of knowl-dge. “When a Griot dies, it is as if a library hasurned to the ground.”23 In the United States,any slaves were punished for becoming liter-

eline survey to follow-up survey for each specificlacks versus whites, along with statistical results, aswere not used in the analysis. Numbers of availabletervention), 22 of 25 for blacks and 29 of 38 for whites;hites; and group 3 (control), 29 of 32 for blacks and 38

m bash for bd out

up 1 (in

te, and such groups as the Ku Klux Klan have

caAwd

tbemttmsAtAlaap

horfRwmerrtcb

ststscfpsmtttdH

tmidsmimpro

bAgmdeA

dwMCGJST

RoMA

5

l

sJ

uw

sa

aM

s

PEER MENTORING IN ESRD END-OF-LIFE PLANNING 117

ontinued to persecute African Americans whospire to higher education.17,21 As a result, somefrican Americans may well shun reading andriting activities on the basis of their experienceuring the last several centuries.Cultural mistrust of the health care system in

he United States among African Americans coulde another explanation for our findings. In suchxperiences as the 40-year Tuskegee Experi-ent,20,21,59 African-American men were misled

hrough the use of printed materials to believehey were receiving help, whereas in truth, theseen remained untreated for syphilis. Such epi-

odes have fostered profound mistrust amongfrican Americans, mistrust that includes writ-

en medical protocols.21,59 Furthermore, Africanmericans are disproportionately represented in

ower-level jobs that provide little or no insur-nce coverage,16,60,61 engendering disparity invailability and access to health care, which am-lifies mistrust.Baker62 suggests that African Americans often

ave a familial, religious, and communal framef reference that deters autonomous individualesponsiveness, and this characteristic is mani-est particularly in the consideration of ADs.elationships with family members and peersho have had similar life experiences may be aore effective way to engender trust and connect-

dness with African Americans. Because peerace was not the critical factor in our study, theelationship itself, and not racial similarity, washe likely stimulus for effective AD education,hange in behavior, and psychosocial well-eing.63

There are several weaknesses of the presenttudy. One weakness is the relatively short dura-ion of intervention, 2 to 4 months. Although thistudy realized some success with our peer men-oring, 2 months may be too short to developufficient rapport and relationship to foster ADompletion in many patients. Furthermore, therequent and unpredictable interruptions im-osed by chronic illness often were obstacles tomooth interactions for both patients and peerentors. Another weakness of the study might be

he focus on the AD document itself, rather thanhe process of end-of-life discussion. The impor-ance of the process and shortcomings of the ADocuments have been noted by others.5,9,12,13

owever, given the complexity of defining and 2

hen quantifying the “process,” the AD docu-ent and patients’ general feelings about discuss-

ng ADs served as surrogates for quantitativeetermination. Finally, there may be biases in thetudy design related to contact time between peerentors and patients that were not controlled for

n the other groups. Even so, the impact of peerentoring was strongest in African-American

atients, a finding that was not likely a soleeflection of such bias and that will have to standr fall on the merits of the results as reported.Despite these difficulties, changes observed in

ehavior and psychosocial well-being in African-merican patients are very encouraging and sug-est the need for more work in this area. Peerentoring in dialysis units appears to address

ifferences in cultural value systems, allay anxi-ty, and promote discussion and acceptance ofD planning and, to some degree, death itself.

ACKNOWLEDGEMENTThe authors thank the 17 peer mentors who worked

iligently to complete this study and the participating socialorkers (all MSW): Vicky Abbott, Anora Ackerson, PhD,elissa Adams, Mary Armstrong, Carrie Bernat, Evelynlark, Carrie DeGrande, Nancy Galbraith, Sue Glover, Lisareenblades, Cathy Harper, Sandy Lopez, Connie Micklin,

enny Niec, Barb Paison, Michael Phillips, Gina Sikon, Gerimith, Valerie Stephenson, Leslie Thompson, and Donnareppa.

REFERENCES1. US Renal Data System: USRDS 2004 Annual Data

eport. The National Institutes of Health, National Institutef Diabetes and Digestive and Kidney Diseases, Bethesda,D, 2004. Available at: http://www.usrds.org. Accessed:pril 22, 20052. Patient Self-Determination Act. US Congress, PL 101-

08, 19903. Fagerlin A, Schneider C: Enough: The failure of the

iving will. Hastings Cent Rep 34:30-42, 20044. Danis M, Southerland L, Garrett J, et al: A prospective

tudy of advance directives for life-sustaining care. N EnglMed 324:882-888, 19915. Emanuel E, Weinberg D, Gonin R, Hummel L, Eman-

el L: How well is the Patient Self-Determination Actorking? An early assessment. Am J Med 95:619-628, 19936. Morrison R, Olson E, Mertz K, Meier D: The inacces-

ibility of advance directives on transfer from ambulatory tocute care settings. JAMA 274:478-482, 1995

7. Swartz R, Perry E: The frequency of withdrawal fromcute care is impacted by severe acute renal failure. J Palliated 7:676-682, 20048. McIntyre K: On advancing advance directives: Why

hould we believe the promise? Arch Intern Med 155:2271-

273, 1995

cd

wN

A1

r2

St

ig

beR

ECR

PA

cc

rcP

1eH

Tp

Y

Ppr

rR

ft

ss

e

pis9

g

a

aA

pC

rC

ht1

p

Ec

s

ec1

c

f

a

icc

tp

S4

Go8

ie

CPo

PERRY ET AL118

9. Holley J, Nespor S, Rault R: The effects of providinghronic hemodialysis patients written material on advanceirectives. Am J Kidney Dis 22:413-418, 199310. Swartz R, Perry E: Advance directives are associated

ith “good deaths” in chronic dialysis patients. J Am Socephrol 3:1623-1630, 199311. Tucker R: Patient Self-Determination Act: An African

merican perspective. Camb Q Healthc Ethics 3:417-419,99412. O’Brien L, Grisso J, Maislin F, et al: Nursing home

esidents’ preferences for life-sustaining treatments. JAMA74:1775-1779, 199513. Perry E, Buck C, Newsome J, Berger C, Messana J,

wartz R: Dialysis staff influence patients in formulatingheir advance directives. Am J Kidney Dis 25:262-268, 1995

14. Hauser J, Kleefield S, Brennan T, Fishbach R: Minor-ty populations and advance directives: Insights from a focusroup methodology. Camb Q Healthc Ethics 6:58-71, 199715. Li S, McAlpine D, Liu J, Li S, Collins A: Differences

etween African Americans and whites in the incidence ofnd-stage renal disease and associated risk factors. Adv Reneplace Ther 1:5-13, 200416. Brown E, Ojeda V, Wyn R, Levan R: Racial and

thnic Disparities in Access to Health Insurance and Healthare. Los Angeles, CA, UCLA Center for Health Policyesearch and the Henry J. Kaiser Family Foundation, 200017. Crawley L, Payne R, Bolden J, Payne T, Washington

, Williams S: Palliative and end-of-life care in the Africanmerican community. JAMA 284:2518-2521, 200018. Swanson G, Ward A: Recruiting minorities into clini-

al trials: Toward a participant-friendly system. J Natl Can-er Inst 87:1747-1759, 1995

19. Tucker C, Peterson S, Herman K, et al: Self-egulation predictors of medication adherence among ethni-ally different pediatric patients with renal transplants. Socediatric Psychiatry 26:455-464, 200120. Thomas S, Quinn S: The Tuskegee syphilis study,

932-1972: Implications for HIV education and AIDS riskducation programs in the black community. Am J Publicealth 81:1498-1504, 199121. Shavers V, Lynch C, Burmeister L: Knowledge of the

uskegee study and its impact on the willingness to partici-ate in medical research studies. JAMA 92:563-572, 200022. Hurston Z: Their Eyes Were Watching God. New

ork, NY, Lippincott, 193723. Haley A: Roots. New York, NY, Dell, 197624. Perry E, Swartz J, Kelly G, Brown S, Swartz R:

alliative care in chronic kidney disease: Peer mentoringrogram personalizes advance directives discussions. Neph-ol News Issues 17:28-31, 2003

25. Kapron K, Perry E, Bowman T, Swartz R: Peeresource consulting: Redesigning a new future. Adv Reneplace Ther 4:267-274, 199726. Merewood A, Philipp B: Peer counselors for breast-

eeding mothers in the hospital setting: Trials, training,ributes, and tribulations. J Hum Lactation 19:72-76, 2003

27. Peregrin T: Education, peer counseling, and paternalupport: Three ways to encourage a healthful breastfeedingchedule. J Am Diet Assoc 102:943-944, 2002

28. Molassiotis A, Callaghan P: A pilot study of the

ffects of cognitive-behavioral group therapy and peer sup- N

ort/counseling in decreasing psychological distress andmproving quality of life in Chinese patients withymptomatic HIV disease. AIDS Patient Care STDs 16:83-6, 200229. Evans J, Livneh H: Peer counseling: A training pro-

ram. J Rehabil 48:52-55, 198230. Luna G, Rotheram-Borus M: Youth living with HIV

s peer leaders. Am J Community Psychol 27:1-23, 199931. Magura S, Siddiqui Q, Shapiro J, et al: Outcomes of

n AIDS prevention program for methadone patients. Int Jddict 26:629-655, 199132. McLean DA: A model for HIV risk reduction and

revention among African American college students. J Amoll Health 42:220-223, 199433. Simonson S: Peer counseling in health care: A collabo-

ation of social work and voluntarism. Social Work Healthare 12:1-19, 198734. Sherman B, Sanders L, Yearde J: Role-modeling

ealthy behavior: Peer counseling for pregnant and postpar-um women in recovery. Womens Health Issues 8:230-238,99835. Coreil J, Behal R: Man to man prostate cancer sup-

ort groups. Cancer Pract 7:122-129, 199936. Ashbury F, Cameron C, Mercer S, Fitch M, Nielsen

: One-on-one peer support and quality of life for breastancer patients. Patient Educ Counseling 35:89-100, 1998

37. Barba S: Peer mentorship: A crucial ingredient foruccess in nursing school. Imprint 50:27-28, 2003

38. Boudin K, Carrero L, Clark J, et al: ACE: A peerducation and counseling program meets the needs of incar-erated women with HIV/AIDS. J Assoc Nurses AIDS Care0:90-98, 199939. Broscious S, Saunders D: Peer coaching. Nurse Edu-

ator 26:212-214, 200140. Leshovitz M: Florida support group helps to ease

ears of transplantation. Nephrol News Issues 4:24-25, 199541. Gorynski L, Knight F: A peer support group for

dolescent dialysis patients. ANNA 19:262-264, 199242. Brinker K, Lichtenstein V: Value of a self-help group

n the psychosocial adjustment of end-stage renal diseaselients and their families. Am Assoc Nephrol Nurses Techni-ians J 8:23-27, 1981

43. Hastings C, Gray-Toft P: Development and evalua-ion of a support program for renal patients and theirartners. Dial Transplant 11:1101-1106, 198244. Faulk J: Peer-to-peer transplant mentor program: The

an Diego experience. Transplant Proc 31:75S, 1999 (supplA) (abstr)45. Tucker C, Chennault S, Green D, Ziller R, Findayson

: Assessment-based group counseling to address concernsf chronic hemodialysis patients. Patient Educ Counseling:51-61, 198646. Perry E, Swartz J: Peer mentoring program personal-

zes advance directives discussion with chronic kidney dis-ase patients. Kidney Kronicle 4:1-3, 2003

47. Derogatis L, Lipman R, Rickles K, Uhlenhuth E,ovi L: The Hopkins Symptom Checklist (HSCL), in Pichot(ed): Psychological Measurements in Psychopharmacol-

gy: Modern Problems in Pharmacopsychiatry. New York,

Y, Karger, 1974

W

CaC

ie

A9

K

vp1

AR

rl2

Sl2

gp1

A

r

CW

itR

d6

aB

PEER MENTORING IN ESRD END-OF-LIFE PLANNING 119

48. Diener E, Emmons R, Griffen S: The Satisfactionith Life Scale. J Pers Assess 49:71-75, 198549. University of Michigan Institute for Social Research:

hanging Lives of Older Couples: Surveys in Preparationnd in Testing. Available at: http://www.isr.umich.edu/LOC. Accessed: April 22, 200550. Robert Wood Johnson Foundation: Five Wishes. Ag-

ng With Dignity, 2001. Available at: http://www.isr.umich.du. Accessed: April 22, 2005

51. Kimmel P, Patel S, Peterson R: Depression in African-merican patients with kidney disease. J Natl Med Assoc4:S92-S102, 2002 (suppl 8)52. Kimmel P: Psychosocial factors in dialysis patients.

idney Int 59:1599-1613, 200153. Patel S, Shah V, Peterson R, Kimmel P: Psychosocial

ariables, quality of life, and religious beliefs in ESRDatients treated with hemodialysis. Am J Kidney Dis 40:1013-022, 200254. Leggat J, Swartz R, Port F: Withdrawal from dialysis:review with an emphasis on the black experience. Adv

en Replace Ther 4:92-97, 199755. Kimmel P: Psychosocial factors in adult end-stage

enal disease patients treated with hemodialysis: Corre-ates and outcomes. Am J Kidney Dis 35:S132-S140,

000 (suppl 1) 1

56. Unruh M, Miskulin D, Yan G, et al, and the HEMOtudy Group: Racial differences in health-related quality of

ife among hemodialysis patients. Kidney Int 65:1482-1491,00457. Lopes A: Relationships of race and ethnicity to pro-

ression of kidney dysfunction and clinical outcomes inatients with chronic kidney failure. Adv Ren Replace Ther1:14-23, 2004

58. Griots. Available at: http://www.si.umich.edu/chico.ccessed:59. Corbie-Smith G, Thomas S, St George D: Distrust,

ace, and research. Arch Intern Med 16:2458-2463, 200260. Smedley B, Stith A, Nelson A: Unequal treatment:

onfronting racial and ethnic disparities in health care.ashington, DC, National Academies, 200361. Spencer M, Kohn L, Woods J: Labeling vs early

dentification: The dilemma of mental health service utiliza-ion among low-income African American children. Afr Ames Perspect 8:1-15, 200262. Baker M: Cultural differences in the use of advance

irectives: A review of the literature. Af Am Res Perspect:35-40, 200063. Perry E, Gumban J, Kelly G: A multidisciplinary

pproach to end-of-life care, in Chambers J, Germain M,rown E (eds): Supportive Care for the Renal Patient, chap

4. New York, NY, Oxford University, 2004, pp 238-265