Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/229153728

ImpactofNavigatorsonCompletionofStepsintheKidneyTransplantProcess:ARandomized,ControlledTrial

ArticleinClinicalJournaloftheAmericanSocietyofNephrology·July2012

DOI:10.2215/CJN.11731111·Source:PubMed

CITATIONS

48

READS

43

13authors,including:

CatherineMSullivan

MetroHealthMedicalCenter

21PUBLICATIONS480CITATIONS

SEEPROFILE

JaneenLeon

CaseWesternReserveUniversity

34PUBLICATIONS835CITATIONS

SEEPROFILE

KennethBodziak

UniversityofMassachusettsMedicalSchool

22PUBLICATIONS806CITATIONS

SEEPROFILE

JeffreyMAlbert

CaseWesternReserveUniversity

84PUBLICATIONS2,225CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyCatherineMSullivanon08March2016.

Theuserhasrequestedenhancementofthedownloadedfile.

Article

Impact of Navigators on Completion of Steps in theKidney Transplant Process: A Randomized,Controlled Trial

Catherine Sullivan,*†‡ Janeen B. Leon,*†‡ Srilekha S. Sayre,*†‡ Marquisha Marbury,*†‡ Michael Ivers,*†‡ Julie A. Pencak,*†‡

Kenneth A. Bodziak,‡§ Donald E. Hricik,‡§ E. Janie Morrison,‡§ Jeffrey M. Albert,| Sankar D. Navaneethan,¶

Christina M. Delos Reyes,** and Ashwini R. Sehgal*†‡§§

SummaryBackground and objectives Many patients with ESRD, particularly minorities and women, face barriers incompleting the steps required to obtain a transplant. These eight sequential steps are as follows: medicalsuitability, interest in transplant, referral to a transplant center, first visit to center, transplant workup, successfulcandidate, waiting list or identify living donor, and receive transplant. This study sought to determine the effectof navigators on completion of steps.

Design, setting, participants, & measurements Cluster randomized, controlled trial at 23 Ohio hemodialysisfacilities. One hundred sixty-seven patients were recruited between January 2009 and August 2009 and werefollowed for up to 24 months or until study end in February 2011. Trained kidney transplant recipients metmonthly with intervention participants (n=92), determined their step in the transplant process, and providedtailored information and assistance in completing the step. Control participants (n=75) continued to receive usualcare. The primary outcome was the number of transplant process steps completed.

Results Starting step did not significantly differ between the two groups. By the end of the trial, interventionparticipants completedmore than twice as many steps as control participants (3.5 versus 1.6 steps; difference, 1.9steps; 95% confidence interval, 1.3–2.5 steps). The effect of the intervention on step completionwas similar acrossrace and sex subgroups.

Conclusions Use of trained transplant recipients as navigators resulted in increased completion of transplantprocess steps.

Clin J Am Soc Nephrol 7: 1639–1645, 2012. doi: 10.2215/CJN.11731111

IntroductionCompared with long-term dialysis treatment, kidneytransplantation generally offers a longer life span andbetter quality of life (1,2). Obtaining a kidney transplantrequires patients to complete a series of steps (Table 1):medical suitability, interest in transplant, referral to atransplant center, first visit to center, transplantworkup, successful candidate, waiting list or identifyliving donor, and receive transplant (3). Medical suit-ability refers to the absence of absolute contraindica-tions to transplantation (e.g., a systemic infection orrecent malignancy) (4). The transplant center workuptypically requires several visits and involves a medicalhistory and physical examination, psychosocial assess-ment, evaluation and treatment of medical conditions,and laboratory studies (5). Patients who complete thisworkup and are found to be successful transplant can-didates may be placed on a deceased-donor waiting listor may obtain a kidney transplant from a living donor.

Many patients, particularly minorities and women,face barriers in completing the steps required to obtain

a transplant (3,6–10). These barriers include inade-quate assessment of medical suitability, lack of infor-mation about transplantation, reliance on nephrologiststo make referrals to transplant centers, and difficultycompleting the transplant center workup (3,6–11). Wehypothesized that navigators might help patients com-plete transplant process steps in a more efficient andequitable manner. Navigators are individuals whoeducate patients and help them navigate through themedical system (12). We reasoned that kidney trans-plant recipients may be ideal navigators for otherpatients because of a shared experience with ESRD(13). Similar peer-mediated interventions have beensuccessful in educational settings (14).

Materials and MethodsParticipants and FacilitiesAll 23 hemodialysis facilities that belong to the

three largest chains in Cuyahoga County, Ohio,participated. A data manager used a random-number

*Center for ReducingHealth Disparities,‡Department ofMedicine,|Department ofEpidemiology andBiostatistics, and§§Department ofBiomedical Ethics,Case Western ReserveUniversity, Cleveland,Ohio; †Division ofNephrology,MetroHealth MedicalCenter, Cleveland,Ohio; §Division ofNephrology and**Department ofPsychiatry, UniversityHospitals CaseMedical Center,Cleveland, Ohio; and¶Department ofNephrology andHypertension,Cleveland Clinic,Cleveland, Ohio

Correspondence:Dr. Ashwini Sehgal,Division ofNephrology,MetroHealth MedicalCenter, 2500MetroHealth Drive,Cleveland, OH44109. Email: axs81@cwru.edu

www.cjasn.org Vol 7 October, 2012 Copyright © 2012 by the American Society of Nephrology 1639

generator to assign facilities to an intervention or controlgroup. To minimize the possibility of contamination by ne-phrologists who work at both intervention and controlfacilities, we first determined whether each nephrologist hadmorepatientsat intervention facilitiesorat control facilities.Wethen included in the study only the larger group of patients.Study coordinators abstracted medical records to iden-

tify community-dwelling patients age 18–70 years who hadno absolute contraindications to kidney transplantation(15). We excluded nursing home residents and patientsolder than 70 years because few transplantations are per-formed among such individuals (1). Absolute contraindi-cations to kidney transplantation at the two transplantcenters in our region include systemic infections, extremeobesity, and active or recent malignancy (4). As a result,we excluded patients with chronic systemic infections, abody mass index .40 kg/m2, or malignancies within thelast 2 years. We also excluded patients who had alreadymade a first visit to a transplant center or received a kid-ney transplant in the past; this was an exclusion criterionbecause such patients demonstrated an ability to completekey steps in the transplant process. Finally, we excludedpatients who had a communication barrier (e.g., those whowere mentally incompetent or did not speak English).Study coordinators described the study to eligible pa-

tients during a dialysis treatment and obtained writteninformed consent. Each participant was given $15 every 6months to thank him or her for participation. This studywas approved by the institutional review board of Metro-Health Medical Center, Cleveland, Ohio, and was regis-tered at ClinicalTrials.gov (identifier NCT00805038).

Baseline AssessmentUnblinded study coordinators abstracted medical re-

cords of intervention and control participants to obtaindemographic and medical characteristics. Coordinatorsalso abstracted medical records and interviewed partic-ipants to determine their baseline step in the transplantprocess. The baseline step was defined as the earliest stepthat was incomplete at the beginning of the study.

Intervention GroupWe hired and trained three study coordinators who were

kidney transplant recipients to act as navigators for in-

tervention group participants. Their training included instruc-tion on the kidney transplant process, human subjectsprotection, medical records abstraction, and motivationalinterviewing. A transplant navigator met monthly with eachintervention participant during a dialysis treatment, reviewedhis or her medical record, and determined the participant’scurrent step. On the basis of the current step, the navigatorcarried out the following tasks for the first seven steps:Step 1. Suitability for Referral to Transplant Center.

Patients with absolute contraindications to kidney trans-plantation listed in their medical records were not eligibleto participate. However, some patients were told byproviders that they were unsuitable for transplantationbut had no absolute contraindications listed in theirrecords. In these cases, study staff contacted the partic-ipant’s nephrologist for clarification. Participants with ab-solute contraindications remained at this step. Participantswithout absolute contraindications moved forward to thenext step.Step 2. Interest in Transplantation. Participants were

educated about the advantages and disadvantages oftransplantation, the steps in the transplant process, andwhat to expect after transplantation. Navigators alsoshared their personal experiences with dialysis and trans-plantation.Step 3. Referral Call to Transplant Center. Participants

were given the phone numbers of local transplant centers, alist of the information that they might be asked to provide,and questions to ask.Step 4. First Visit to Transplant Center. Participants

were given directions to the transplant center, a list ofthings to take, and questions to ask. In addition, navigatorsexplored transportation options for getting to the trans-plant center and reminded participants of upcomingappointments.Step 5. Transplant Center Workup. Navigators ex-

plained to participants what to expect in a transplantcenter workup. Navigators also monitored completion ofspecific aspects of the workup, communicated with trans-plant center staff about outstanding tasks, and encouragedparticipants to complete the workup in a timely fashion.Step 6. Successful Transplant Candidate. Navigators

served as an ongoing source of support and information.Navigators also educated participants about how to dis-cuss living donation with potential donors.

Table 1. Steps involved in obtaining a deceased-donor or living-donor kidney transplant

Step Step Completion

1. Suitability for referral totransplant center

Dialysis facility records or nephrologist indicates that patient has no absolutecontraindications to kidney transplantation.

2. Interest in transplantation Patient expresses an interest in consideringadeceased- or living-donor transplant.3. Referral call to transplant center Transplant center records indicate that referralwasmadebypatient, nephrologist,

or dialysis facility.4. First visit to transplant center Transplant center records indicate thatpatientmadean initial visit to transplant center.5. Transplant center workup Transplant center records indicate that patient completed workup.6. Successful transplant candidate Transplant center records indicate thatpatient is a successful transplant candidate.7. On waiting list or evaluatepotential living donor

Transplant center records indicate that patient is on a deceased-donorwaiting listor a potential living donor is being evaluated.

8. Receive transplant Transplant center records indicate that patient received a deceased- or living-donor kidney transplant.

1640 Clinical Journal of the American Society of Nephrology

Step 7. On Waiting List or Evaluate Potential LivingDonor. Navigators served as an ongoing source of supportand information. Navigators monitored participants’ statuson the waiting list and results of living-donor evaluation.

Control GroupControl participants continued to receive care from their

nephrologists and dialysis facilities. A study coordinator whowas not a transplant recipient assessed control participants every3 months. This assessment included medical record abstractionand a brief interview to determine participants’ current step.

Follow-up ProceduresParticipants were recruited between January 2009 and

August 2009 and were followed up for 24 months or untilthey died, moved, withdrew, or reached the study end inFebruary 2011. Study coordinators abstracted medicalrecords and interviewed participants to determine theirfinal step. Because of the sequential nature of the transplantprocess, completion of a particular step implied that priorsteps had also been completed (see Table 1). To parallel thecategorization of baseline step, the final step was definedas the farthest step that was incomplete at the end of thestudy. If intervention participants were unable to completespecific steps by the end of the study, navigators notedwhether this was due to a medical limitation, financialconcerns, patient reluctance, or ending the trial early.

OutcomesThe primary outcome measure was the number of steps

completed, defined as the difference between final andbaseline steps. Secondary outcomes were impediments tostep completion among intervention participants.

Statistical AnalysesWe used mixed-effects models for continuous variables

and generalized estimating equation (GEE) models forcategorical variables to compare the baseline characteristicsof intervention and control participants while accountingfor clustering within facilities (16). Intervention effectswere assessed using an F-test (in mixed-effects models)or generalized score test (in GEE analyses). We usedGEEs to assess the mean baseline and final number ofsteps for each group, the difference (total number of stepscompleted) for each group, and the difference in total stepsbetween the two groups (assuming a Poisson distributionand using a linear link). The identity link for the latteranalysis was used to directly assess mean differences.The intervention effect was estimated for the overall sam-ple and within each level for sex, race, and baseline step.GEE models for within-group analyses included only theintercept, while models for between-group analyses in-cluded the intercept and a group indicator. We used anexchangeable working covariance matrix for all GEE mod-els (except for the analysis by baseline step, for which anindependence covariance matrix was used to avoid occa-sional nonconvergence), and we obtained empirical (ro-bust) variance estimates to account for clustering withinfacilities.We carried out multiple imputation to account for

missing data. Specifically, we imputed the final step for

12 intervention and 9 control patients who moved orwithdrew (detailed in Figure 1) and for 51 intervention and35 control patients who did not complete 24 months beforethe study end in February 2011. To account for informa-tion on the steps achieved at the time of dropout, eachimputed value was obtained, in the manner of empiricalBayes/shrinkage estimation, as an optimal (minimum var-iance) linear combination of the predicted value based onthe baseline covariates (i.e., from a regression model fit tothe nondropout sample) and a predicted value based onthe individual responses (baseline step and final step atdropout) under a linear model. The covariate-predictedcomponent of the imputations was obtained by predictedmean matching, using five as the number of closest obser-vations for the draws and including all baseline character-istics of participants as covariates (listed in Table 1) (17). Inaddition, imputed total steps were constrained so that thefinal number of steps was no more than eight (Table 1).Twenty completed datasets were obtained by imputingtotal steps in this manner. We did not impute endpointsfor patients who died but used the number of steps theycompleted before death as their endpoints. The multipleimputed (completed) datasets were each analyzed usingthe relevant GEE model, and t tests accounting for thewithin- and between-dataset variability were used to as-sess intervention effects (18).As secondary analyses, GEE/multiple imputation meth-

ods were used within prespecified subgroups defined bykey baseline covariates (sex, race, and baseline step). Wealso verified the analyses of the primary outcome with aGEE model assuming that the number of completed steps(out of the potential number of steps given the baselinenumber) follows a binomial distribution and using a logitlink (with the model otherwise the same as before). Thisanalysis gave similar results, which are not presented. Allanalyses were performed using SAS software, version 9.2(SAS Institute, Cary, NC).On the basis of prior work, we anticipated that the SD of

the number of steps completed would be approximately 2.0steps. To detect a clinically important effect size of 1.0 step(or standardized effect size of 0.5) would require 126 totalparticipants with a two-tailed a level of 0.05 and 80%power (19). However, these estimates must be increasedto account for the possible nonindependence of patientsclustered within facilities. We found a small amount ofclustering of outcomes by facility in our previous work(intraclass cluster coefficient, r , 0.025). This gives an in-flation factor of 1 + (m21) r, where m is the average num-ber of patients at each facility (in this study, 126 patientsdivided by 23 facilities). This gives an inflation factor of1.11 and a total sample size requirement of 140 patients forour primary outcome.

ResultsParticipant CharacteristicsFigure 1 shows the flow of participants through the trial.

One hundred sixty-seven patients began the trial: 92 in-tervention participants and 75 control participants. We ex-cluded 377 patients in intervention facilities because theirnephrologists had more patients in control facilities; wetherefore assigned these patients to the control group.

Clin J Am Soc Nephrol 7: 1639–1645, October, 2012 Impact of Navigators, Sullivan et al. 1641

Similarly, we excluded 432 patients in control facilities be-cause their nephrologists had more patients in interventionfacilities. A total of 1617 patients did not meet eligibility cri-teria. One hundred sixty-nine eligible patients declined toparticipate. Compared with the 167 participants, these 169nonparticipants were somewhat older (58 versus 55 years;P=0.01) and had been undergoing hemodialysis longer (4.0versus 3.0 years; P=0.0001) but did not differ in other demo-graphic characteristics. Intervention and control participantshad generally similar baseline characteristics (Table 2).

Completion of Steps in Transplant ProcessIntervention patients completed more than twice as

many steps as control patients (3.5 versus 1.6 steps; dif-ference, 1.9 steps; 95% confidence interval, 1.3–2.5 steps;

Table 3). The effect of the intervention on step completionwas similar across race and sex subgroups.By the end of the trial, 17 (18%) intervention participants

and 6 (8%) control participants were on a deceased-donortransplant waiting list (P=0.07). In addition, potential liv-ing donors were identified for 3 (3%) intervention partic-ipants and 0 (0%) control participants (P=0.06). However,no deceased- or living-donor transplants occurred by theend of the trial. No adverse events or adverse effects wereassociated with the intervention.

Reasons for Failing to Complete StepsEven though our intervention was an overall success,

many intervention participants failed to become successfultransplant candidates. Table 4 lists the specific reasons that

Figure 1. | Flow of participants through the trial.

1642 Clinical Journal of the American Society of Nephrology

intervention participants were unable to complete theirfinal step. For example, 24 intervention participants didnot make a referral call to a transplant center by the endof the trial. Sixteen participants had medical limitations,such as acute or chronic conditions, that they wanted toaddress before calling. Three participants had concernsabout the cost of transplantation or immunosuppressivemedications. Ten participants were reluctant to call be-cause of fears regarding surgery and rejection or becausethey felt fine on dialysis. Twelve participants died or leftthe dialysis facility before the trial ended.

Table 2. Baseline characteristics of intervention and controlgroups

CharacteristicIntervention

Group(n=92)

ControlGroup(n=75)

PValue

Age18–44 yr 14 (15) 9 (12)45–54 yr 31 (34) 18 (24) 0.1055–64 yr 39 (42) 30 (40)65–70 yr 8 (9) 18 (24)

Male 47 (51) 47 (63) 0.07Raceblack 65 (71) 49 (65)white 17 (18) 21 (28) 0.50other 10 (11) 5 (7)

Cause of renalfailurehypertension 37 (40) 36 (48)diabetes 35 (38) 23 (31) 0.82GN 9 (10) 8 (11)other 11 (12) 8 (11)

Time receivingdialysis

.

,18 mo 45 (49) 24 (32)18–36 mo 18 (20) 26 (35) 0.06.36 mo 29 (32) 25 (33)

Educationless than highschool

11 (12) 13 (17)

high schoolgraduate

37 (40) 34 (45) 0.12

some college 31 (34) 21 (28)collegegraduate

13 (14) 7 (9)

Mean comorbidconditions 6

SD (n)a

1.661.2 1.861.3 0.17

Baseline stepsuitability 39 (42) 35 (47)interest 3 (3) 8 (11) 0.63referral call 46(50) 26 (35)first visit 4 (4) 6 (8)

Data are expressed as number (percentage) of participants un-less otherwise indicated.aOn the basis of the presence of the following 10 disease cate-gories: coronary artery disease, congestive heart failure, pe-ripheral vascular disease, cerebrovascular disease, depressionor psychosis, previous solid tumor or hematologic malignancy,connective tissue disease, asthma or chronic obstructive pul-monary disease, diabetes mellitus, and drug or alcohol abuse.

Tab

le3.

Number

ofstep

sco

mpletedam

ong92interven

tionan

d75co

ntrolparticipan

ts

Variable

Interven

tion

Con

trol

Differenc

ein

Differenc

es

Patien

ts(n)

MeanBaseline

Step

6SE

MMeanFina

lStep

6SE

MDifferenc

e(95%

CI)

Patien

ts(n)

MeanBaseline

Step

6SE

MMeanFina

lStep

6SE

MDifferenc

e(95%

CI)

Differenc

e(95%

CI)

PValue

Allpa

rticipan

ts92

2.16

0.2

5.66

0.1

3.5(3.1–4.0)

752.06

0.2

3.66

0.3

1.6(1.2–2.0)

1.9(1.3–2.5)

,0.00

1Sex male

472.06

0.3

5.66

0.1

3.8(3.2–4.5)

472.16

0.2

3.76

0.3

1.6(1.1–2.3)

2.1(1.2–2.9)

,0.00

1female

452.26

0.2

5.66

0.2

3.4(3.1–3.8)

282.06

0.2

3.56

0.5

1.5(1.0–2.3)

1.9(1.1–2.7)

,0.00

1Race

black

652.16

0.2

5.66

0.1

3.5(3.1–3.9)

492.26

0.2

3.86

0.4

1.6(1.2–2.3)

1.9(1.1–2.6)

,0.00

1white

172.46

0.3

5.86

0.4

3.4(2.5–4.6)

211.86

0.3

3.86

0.4

1.8(1.4–2.3)

1.7(0.6–2.8)

0.00

3othe

r10

1.96

0.1

5.16

0.4

3.5(2.4–5.2)

51.86

0.4

2.56

0.6

0.6(0.1–2.7)

2.8(1.2–4.4)

,0.00

1Baselinestep

suitab

ility

391.06

0.0

5.26

0.2

4.3(4.0–4.7)

351.06

0.0

2.56

0.4

1.4(0.8–2.3)

2.9(2.1–3.7)

,0.00

1interest

32.06

0.0

4.66

0.7

2.6(1.5–4.4)

82.06

0.0

3.16

0.4

1.2(0.6–2.2)

1.4(-0.2–

3.0)

0.09

referral

call

463.06

0.0

5.86

0.2

2.8(2.4–3.3)

263.06

0.0

4.96

0.3

1.9(1.3–2.7)

1.0(0.2–1.8)

0.02

firstv

isit

44.06

0.0

6.96

0.3

3.1(2.5–4.0)

64.06

0.0

5.96

0.6

2.0(1.3–3.1)

1.2(0.02–

2.3)

0.05

CI,confi

den

ceinterval.

Clin J Am Soc Nephrol 7: 1639–1645, October, 2012 Impact of Navigators, Sullivan et al. 1643

DiscussionWe found that using trained transplant recipients as

navigators resulted in increased completion of steps in thekidney transplant process. Moreover, the effect of theintervention on step completion was similar across raceand sex subgroups. Although our trial was not designed todetermine the effect of navigators on transplantation,completion of sequential steps in the transplant processis necessary for transplantation to occur. Moreover, wefound promising results related to the number of inter-vention participants who reached the penultimate step inthe transplant process (deceased-donor waiting list orevaluate potential living donor).Previous randomized, controlled trials of navigators

have generally focused on cancer screening or treatmentbut have not used patients as navigators (20–27). Our in-tervention has the advantage of being simple but also eas-ily tailored to the circumstances of individual patients. Byengaging the participation of almost all dialysis facilitiesin a large geographic area, we enhanced the generalizabil-ity of our findings. With the exception of race, participantcharacteristics were similar to those of dialysis patientsnationally (1). The large number of black participants re-flects the inner-city location of many of the participatingdialysis facilities.Our results have important implications for patients,

providers, and health policy makers. Patients with ESRDmay benefit from learningmore about kidney transplantationand obtaining help in completing steps in the transplantprocess. In the absence of formal navigators, patients mayneed to actively obtain this knowledge and assistance fromtheir physicians and dialysis facilities, local transplant centersand kidney transplant recipients, and family and friends.Providers should realize that usual care is insufficient toadequately educate patients and help them achieve access tokidney transplantation. In particular, we found that manypatients are inappropriately categorized as unsuitable forreferral. Even though we restricted our study sample topatients without absolute contraindications to transplanta-tion, 89 participants were categorized by their providers asmedically unsuitable for referral at baseline (Table 2). How-ever, our intervention was very successful among this sub-group, with intervention participants completing an averageof 4.2 steps (95% confidence interval, 3.9–4.5 steps; Table 2).Policy makers may consider funding navigators to work at

each of the more than 5000 dialysis facilities nationwide (1).This would not only help patients complete steps in thetransplant process but also create meaningful job opportuni-ties for the many kidney transplant recipients who are cur-rently unemployed (28).Several limitations must be considered in interpreting

our results. A much larger sample size and follow-upduration would be necessary to determine the effect ofnavigation on actual transplantation. As a result, futurenavigator programs should carefully evaluate the effect ofnavigation on transplantation. The transplant processexperiences of our 167 study participants may not be rep-resentative of the experiences of all patients at the 23participating facilities (see Figure 1). The results rely on alarge amount of imputed data. Because of funding limita-tions, the study end date could not be extended beyondFebruary 2011. As a result, participants who were re-cruited after March 2009 were unable to complete 24months of follow-up. Our navigators were paid the samesalaries as study coordinators and spent much of theirtime collecting research data. Such data collection wouldbe unnecessary if navigators were used in clinical settings.We were unable to determine whether cost would impedesetting up navigator programs. Increasing the numberof individuals receiving living donor transplants willincrease the total number of transplants performed.However, increasing the number of individuals on thedeceased-donor waiting list will not increase the totalnumber of transplants performed. As a result, continuedefforts to increase organ donation are necessary. Althoughour intervention was an overall success, many interventionpatients did not become transplant candidates (Table 4).Note that navigation is unlikely to influence medical lim-itations or patients who die or have moved. However, fur-ther refinements to our approach may increase its potencyto influence financial concerns and patient reluctance.In conclusion, we found that navigators increase com-

pletion of steps in the transplant process among all patientsas well as among race and sex subgroups. Further work isneeded to determine the effect of navigators on actualtransplantation and on race and sex disparities. Werecommend that nephrologists, dialysis facilities, andregulatory agencies use the transplant process steps weidentified (Table 1) to develop clinical performance mea-sures to monitor and improve access to transplantation.

Table 4. Reasons that patients in intervention group did not become successful transplant candidates

Final Step Total Patients (n)

Patients with Specific Reasons (n)

MedicalLimitation

FinancialConcern

PatientReluctance

Died or LeftFacility

Interest 2 1 0 2 0Referral call 24 16 3 10 12First visit 15 11 1 6 5First workup 23 11 3 11 6Successful candidate 8 8 0 0 0

Some patients had more than one reason.

1644 Clinical Journal of the American Society of Nephrology

Using peer navigators may also be useful in other settingsand for other conditions as a way to both improve careand reduce disparities.

AcknowledgmentsWe are grateful to the patients and health providers who partic-

ipated in this project.Thisworkwas supported by grants DK51472002265 andRR024989

from the National Institutes of Health, Bethesda, Maryland. Thefunding organization had no role in the design and conduct of thestudy; collection,management, analysis,and interpretationof thedata;or preparation, review, or approval of the manuscript.

DisclosuresS.D.N. reports receiving grant support from Genzyme. D.E.H.

reports receiving payments for lectures from Novartis andGenentech.

References1. US Renal Data System: USRDS 2010 Annual Data Report, Be-

thesda, MD, National Institutes of Health, National Institute ofDiabetes and Digestive and Kidney Diseases, 2010

2. Wolfe RA, Ashby VB, Milford EL, Ojo AO, Ettenger RE, AgodoaLY, Held PJ, Port FK: Comparison of mortality in all patients ondialysis, patients on dialysis awaiting transplantation, and re-cipients of a first cadaveric transplant. N Engl J Med 341: 1725–1730, 1999

3. Alexander GC, Sehgal AR: Barriers to cadaveric renal trans-plantation among blacks, women, and the poor. JAMA 280:1148–1152, 1998

4. Kasiske BL, Cangro CB, Hariharan S, Hricik DE, Kerman RH, RothD, Rush DN, Vazquez MA, Weir MR; American Society of Trans-plantation: The evaluation of renal transplantation candidates:Clinical practice guidelines.AmJTransplant1[Suppl2]:3–95, 2001

5. Gallon LG, Leventhal JR, Kaufman DB: Pretransplant evaluation ofrenal transplant candidates. Semin Nephrol 22: 515–525, 2002

6. Epstein AM, Ayanian JZ, Keogh JH, Noonan SJ, Armistead N,Cleary PD, Weissman JS, David-Kasdan JA, Carlson D, Fuller J,Marsh D, Conti RM: Racial disparities in access to renaltransplantation—clinically appropriate or due to underuse oroveruse? N Engl J Med 343: 1537–1544, 2, 1537, 2000

7. Ayanian JZ, Cleary PD, Weissman JS, Epstein AM: The effect ofpatients’ preferences on racial differences in access to renaltransplantation. N Engl J Med 341: 1661–1669, 1999

8. Ladin K, Rodrigue JR, Hanto DW: Framing disparities along thecontinuum of care from chronic kidney disease to transplantation:barriers and interventions. Am J Transplant 9: 669–674, 2009

9. Navaneethan SD, Singh S: A systematic review of barriers inaccess to renal transplantation among African Americans in theUnited States. Clin Transplant 20: 769–775, 2006

10. Schold JD, Gregg JA, Harman JS, Hall AG, Patton PR, Meier-Kriesche HU: Barriers to evaluation and wait listing for kidneytransplantation. Clin J Am Soc Nephrol 6: 1760–1767, 2011

11. CooreyGM, Paykin C, Singleton-Driscoll LC, Gaston RS: Barriersto preemptive kidney transplantation. Am J Nurs 109: 28–37,quiz 38, 2009

12. Freeman HP, Muth BJ, Kerner JF: Expanding access to cancerscreening and clinical follow-up among the medically un-derserved. Cancer Pract 3: 19–30, 1995

13. Ramirez AG, Turner BJ: The role of peer patients in chronic dis-ease management. Ann Intern Med 153: 544–545, 2010

14. Mohan M: Peer tutoring as a technique for teaching the un-motivated. Child Study J 1: 217–225, 1971

15. Shapiro R, Sarwal MM: Pediatric kidney transplantation. PediatrClin North Am 57: 393–400, 2010

16. Donner A, Klar N:Design andAnalysis of Cluster RandomizationTrials in Health Research, New York, Oxford University Press,2000

17. Heitjan F, Little RJ: Multiple imputation for the fatal accidentreporting system. Appl Stat 40: 13–29, 1991

18. RubinDB:Multiple Imputation for Nonresponse in Surveys, NewYork, John Wiley & Sons, 1987

19. Hulley SB, Cummings SR: Designing Clinical Research, Balti-more, Williams & Wilkins, 1988

20. Christie J, Itzkowitz S, Lihau-Nkanza I, Castillo A, Redd W,Jandorf L: A randomized controlled trial using patient navigationto increase colonoscopy screening among low-income minori-ties. J Natl Med Assoc 100: 278–284, 2008

21. Ell K, Vourlekis B, Xie B,Nedjat-Haiem FR, Lee PJ,Muderspach L,Russell C, Palinkas LA: Cancer treatment adherence amonglow-income women with breast or gynecologic cancer: Arandomized controlled trial of patient navigation. Cancer 115:4606–4615, 2009

22. Ferrante JM, Chen PH, Kim S: The effect of patient navigation ontime to diagnosis, anxiety, and satisfaction in urban minoritywomen with abnormal mammograms: A randomized controlledtrial. J Urban Health 85: 114–124, 2008

23. Jandorf L, Gutierrez Y, Lopez J, Christie J, Itzkowitz SH: Use of apatient navigator to increase colorectal cancer screening in anurban neighborhood health clinic. J Urban Health 82: 216–224,2005

24. Maxwell AE, Jo AM, Crespi CM, Sudan M, Bastani R: Peer navi-gation improves diagnostic follow-up after breast cancerscreening among Korean American women: Results of a ran-domized trial. Cancer Causes Control 21: 1931–1940, 2010

25. Percac-Lima S, Grant RW, Green AR, Ashburner JM, Gamba G,Oo S, Richter JM, Atlas SJ: A culturally tailored navigator programfor colorectal cancer screening in a community health center: Arandomized, controlled trial. J Gen Intern Med 24: 211–217,2009

26. Lasser KE, Murillo J, Lisboa S, Casimir AN, Valley-Shah L,Emmons KM, Fletcher RH, Ayanian JZ: Colorectal cancerscreening among ethnically diverse, low-income patients:a randomized controlled trial. Arch Intern Med 171: 906–912,2011

27. Phillips CE, Rothstein JD, Beaver K, Sherman BJ, Freund KM,Battaglia TA: Patient navigation to increase mammographyscreening among inner city women. J Gen Intern Med 26: 123–129, 2011

28. van derMei SF, Krol B, van SonWJ, de Jong PE, Groothoff JW, vanden HeuvelWJ: Social participation and employment status afterkidney transplantation: A systematic review. Qual Life Res 15:979–994, 2006

Received: November 18, 2011 Accepted: June 14, 2012

Published online ahead of print. Publication date available at www.cjasn.org.

See related editorial, “Initiating and Completing the KidneyTransplant Evaluation Process: The Red Queen’s Race,” on pages1551–1552.

Clin J Am Soc Nephrol 7: 1639–1645, October, 2012 Impact of Navigators, Sullivan et al. 1645

View publication statsView publication stats