Advance care planning with patients who have end-stage kidneydisease a systematic realist review

O'Halloran, P., Noble, H., Norwood, K., Maxwell, P., Shields, J., Fogarty, D., Murtagh, F., Morton, R., & Brazil, K.(2018). Advance care planning with patients who have end-stage kidney disease a systematic realist review.Journal of Pain and Symptom Management, 56(5), 795. https://doi.org/10.1016/j.jpainsymman.2018.07.008

Published in:Journal of Pain and Symptom Management

Document Version:Publisher's PDF, also known as Version of record

Queen's University Belfast - Research Portal:Link to publication record in Queen's University Belfast Research Portal

Publisher rightsCopyright 2018 the authors.This is an open access article published under a Creative Commons Attribution-NonCommercial-NoDerivs License(https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits distribution and reproduction for non-commercial purposes, provided theauthor and source are cited.

General rightsCopyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or othercopyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associatedwith these rights.

Take down policyThe Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made toensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in theResearch Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.

Download date:25. Feb. 2022

Vol. 56 No. 5 November 2018 Journal of Pain and Symptom Management 795

Review Article

Advance Care Planning With Patients Who Have

End-Stage Kidney Disease: A Systematic Realist Review

Peter O’Halloran, PhD, RN, Helen Noble, PhD, RN, Kelly Norwood, PhD, Peter Maxwell, MD, PhD,Joanne Shields, MB BCh, MPhil, MRCP, Damian Fogarty, MB BCh BAO, MRCP, MD,Fliss Murtagh, PhD, FRCP, MRCGP, Rachael Morton, PhD, MSc(Clin Epi), and Kevin Brazil, PhDSchool of Nursing and Midwifery (P.O’H., H.N., K.B.), Queen’s University Belfast, Medical Biology Centre, Belfast; School of Psychology

(K.N.), Ulster University, Coleraine Campus, Coleraine; School of Medicine, Dentistry and Biomedical Sciences (P.M.), Queen’s University

Belfast, Health Sciences Building, Belfast; Regional Nephrology Unit (P.M., J.S., D.F.), Belfast City Hospital, Belfast; Wolfson Palliative

Care Research Centre (F.M.), Hull York Medical School, Allam Medical Building, University of Hull, Hull, United Kingdom; and Sydney

Medical School (R.M.), University of Sydney, NSW, Australia

Abstract

Context. Patients with end-stage kidney disease have a high mortality rate and disease burden. Despite this, many do not

speak with health care professionals about end-of-life issues. Advance care planning is recommended in this context but is

complex and challenging. We carried out a realist review to identify factors affecting its implementation.

Objectives. The objectives of this study are 1) to identify implementation theories; 2) to identify factors that help or hinder

implementation; and 3) to develop theory on how the intervention may work.

Methods. We carried out a systematic realist review, searching seven electronic databases: Medline, Embase, CINAHL,

PsycINFO, Cochrane Library, Google Scholar, and ScienceDirect.

Results. Sixty-two papers were included in the review.

Conclusion. We identified two intervention stagesd1) training for health care professionals that addresses concerns,

optimizes skills, and clarifies processes and 2) use of documentation and processes that are simple, individually tailored,

culturally appropriate, and involve surrogates. These processes work as patients develop trust in professionals, participate in

discussions, and clarify values and beliefs about their condition. This leads to greater congruence between patients and

surrogates; increased quality of communication between patients and professionals; and increased completion of advance

directives. Advance care planning is hindered by lack of training; administrative complexities; pressures of routine care;

patients overestimating life expectancy; and when patients, family, and/or clinical staff are reluctant to initiate discussions. It is

more likely to succeed where organizations treat it as core business; when the process is culturally appropriate and takes

account of patient perceptions; and when patients are willing to consider death and dying with suitably trained staff. J Pain

Symptom Manage 2018;56:795e807. � 2018 The Authors. Published by Elsevier Inc. on behalf of American Academy of Hospice and

Palliative Medicine. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Key Words

advance care planning, advance directives, kidney failure, chronic, palliative care, realist review, renal dialysis

IntroductionPatients with end-stage kidney disease have a high

disease burden and a mortality rate that exceeds age-

Address correspondence to: Peter O’Halloran, PhD, RN, Schoolof Nursing and Midwifery, Queen’s University Belfast, Medi-cal Biology Centre, 97 Lisburn Road, Belfast BT9 7BL,United Kingdom. E-mail: p.ohalloran@qub.ac.uk

� 2018 The Authors. Published by Elsevier Inc. on behalf of AmericanAcademy of Hospice and Palliative Medicine. This is an open accessarticle under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

matched individuals with cancer and cardiovasculardisease.1,2 Renal replacement therapy is life-long treat-ment to replace the normal function of the kidneyswith either chronic dialysis (peritoneal dialysis or

Accepted for publication: July 5, 2018.

0885-3924/$ - see front matterhttps://doi.org/10.1016/j.jpainsymman.2018.07.008

796 Vol. 56 No. 5 November 2018O’Halloran et al.

hemodialysis) or renal transplantation. However, old-er patients with end-stage kidney disease and addi-tional comorbidities may not achieve improvedfunctional status or live longer on renal replacementtherapy. They may be unsuitable for renal transplanta-tion and have complex technical challenges associatedwith initiating dialysis. Furthermore, some older pa-tients may face the decision of whether or not toeven start renal replacement therapy.2e4 Those whoaccept renal replacement therapy often experience arange of distressing psychological and physical symp-toms.5,6 Despite this, many do not have conversationswith loved ones or clinicians about end-of-life issuessuch as preferences in relation to admission to inten-sive care, referral to palliative care services, withdrawalfrom renal replacement therapy, resuscitation, orwhere they might choose to die.7,8 In this context,advance care planning may be a useful interventionto promote shared decision-making among patients,their families, and health care professionals.

Advance care planning is a process aimed at elicit-ing preferences of patients in relation to future care,including end-of-life preferences.9 It can help patientsidentify their future health priorities; make anadvance decision to refuse a specific treatment (anadvance directive); or to appoint someone with lastingpower of attorney.10 Advance care planning is associ-ated with considerable benefits in the general popula-tion including improved quality of life, decreasedanxiety and depression among family members,reduced hospitalizations, increased uptake of hospiceand palliative care services, and care that concurs withpatient preferences.11e15

Advance care planning is recommended as goodpractice for patients with end-stage kidneydisease,16e18 although recent systematic reviews havedemonstrated the limited evidence base for the effec-tiveness of advance care planning in this popula-tion.19,20 Furthermore, advance care planning is acomplex process that can provoke cultural and per-sonal sensitivities related to dying11; and uptake isinfluenced by a range of personal and organizationalfactors.21 A review of barriers to advance care planningin the general population22 found that implementa-tion was more likely to succeed when advance careplanning was embedded in an evaluative researchinfrastructure, supported by organizational guidance,and part of clinical governance requirements. It wasmade more difficult when staff lacked time andtraining, by staff fears around upsetting patients andtheir relatives, and due to the difficulty of pickingthe right time for the discussion. Adhering to patients’preferences could be jeopardized by the unpredict-ability of clinical and organizational factors, andcommunication difficulties between health care orga-nizations. Given that advance care planning is

recommended as best practice in end-stage kidney dis-ease and yet is a complex and challenging process,23

we carried out a realist review of the literature to iden-tify factors that may help or hinder the implementa-tion of advance care planning in this population.

Realist ReviewA realist review aims to critically examine the inter-

action between context, mechanism, and outcome(characterized as context-mechanism-outcome config-urations) in a sample of studies of an intervention,24

seeking to establish ‘‘what works, for whom, and inwhat circumstances.’’25

Underlying realist reviews are the assumptions ofrealist evaluation.25 This proposes that an interventionworks by changing the context for the people actingwithin it. Context is defined as ‘‘the spatial orgeographical or institutional location into which pro-grams are embedded’’ and ‘‘the prior set of socialrules, norms, values and interrelationships gatheredin these places which sets limits on the efficacy of pro-gram mechanisms’’ (25, P. 70). These ‘‘mechanisms’’are the reasoning, beliefs, feelings, motivations, andchoices of individuals and groups, which lead to pat-terns of behavior that we recognize as outcomes.26,27

When an intervention is introduced, it changes thecontext (by providing further reasoning, opportu-nities, permissions, legitimations, authorizations, andlimitations), so presenting people with a different setof circumstances in which to exercise agency, leadingto different outcomes. Of course, this is an interactiveprocess and people will both reproduce and transformtheir new context, potentially producing a range ofoutcomes, some of them unintended by those devisingthe program.This understanding is neatly summarized by Cheyne

et al.:‘‘A programme comprises multiple elements or

components which introduce ideas and/or opportu-nities for change into existing social systems; the pro-cess of how people interpret and act upon theseopportunities/ideas are known as the programme’smechanisms.’’ (28, P1112)

ObjectivesThe objectives of the review were as follows:

1. To identify the dominant program theories inrelation to how advance care planning can besuccessfully implemented.

2. To identify factors that may help or hinder theimplementation of advance care planning,with reference to the Consolidated FrameworkFor Implementation Research,29 a pragmaticstructure for identifying potential influences

Table 1Search Strategy for MedlinedModified for Searches in

Vol. 56 No. 5 November 2018 797Advance Care Planning in Kidney Failure

on implementation, focused on the followingconstructs:

CINAHL, Embase, PsycINFO, Cochrane Library, GoogleScholar, and ScienceDirect

1) exp Advance Care Planning/2) (advance care adj (plan or plans or planning)).tw.3) (advance adj (directive* or decision*)).tw.4) living will*.tw.

� The characteristics of the intervention� The outer and inner settings forimplementation

� The characteristics of the individuals involved� The implementation process

5) Right to Die/6) right to die.tw.7) Patient Advocacy/8) ((patient or patients) adj5 (advocat* or advocacy)).tw.

3. To develop theory based on context-mechanism-outcome configurations to helpexplain how the intervention may work.

9) power of attorney.tw.10) ((end of life or end-of-life) adj5 (care or discuss* or decision* or

plan or plans or planning or preference*)).tw.11) Terminal Care/12) terminal care.tw.13) Treatment Refusal/14) exp Withholding Treatment/15) (treatment adj5 (refus* or withhold* or withdraw*)).tw.16) 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or

14 or 1517) Renal Replacement Therapy/18) renal replacement therapy.tw.19) exp Renal Dialysis/20) dialysis.tw.21) (hemodialysis or haemodialysis).tw.22) (endstage kidney or endstage renal or end stage kidney or end

stage renal).tw.23) (ESKD or ESKF or ESRD or ESRF).tw.24) 17 or 18 or 19 or 20 or 21 or 22 or 2325) 16 and 24

MethodsWe followed the RAMESES publication standards

for realist syntheses in conducting and reporting ourreview.24

Scoping the LiteratureBefore systematically searching the literature, we

consulted with patient advocacy groups, health careprofessionals, and fellow researchers and carried outa rapid survey of systematic reviews, guidance, and pol-icy documents relevant to advance care planning toidentify key issues in this area of practice.

Search Methods for the ReviewP. O’H and K. N. searched the following databases

(from 1960 to 2016) using the relevant search termsor MESH/Thesaurus/Keyword headings for eachdatabase: Medline, Embase, CINAHL, PsycINFO, Co-chrane Library, Google Scholar, and ScienceDirect.The Medline search terms, which were based on thoseproposed in a recent Cochrane review protocol,30 areprovided in Table 1.

Selection and Appraisal of DocumentsDocuments were selected based on their relevance

for theory building rather than their quality or meth-odological approach, which allowed us to include arange of literature. We included documents if they ad-dressed advance care planning and end-stage kidneydisease. We excluded documents that were not in En-glish as we had no access to translation services.Finally, we excluded documents published before2000 as we found no randomized controlled trials(RCTs) published before this date and the organiza-tional and clinical context was likely to be significantlydifferent for papers published earlier.

Realist reviews include a broad spectrum of studies,but the quality of studies is used to moderate findings.The methodological quality of empirical studies wasassessed using the appropriate appraisal tool fromthe Critical Appraisal Skills Programme,31 and studieswere classified as strong, moderate, or weak in termsof rigor.

Data ExtractionData were harvested from papers selected for full-

text review by five reviewers (P. O’H., K. N., H. N., J.B., and J. C.) using a standardized data extractionform developed for a previous realist review,32 basedon the RAMESES publication standards.24 This formincluded sections related to the realist assessment,requiring the reviewer to seek information on thetheoretical background and characteristics of theintervention, how it was thought to work, outer and in-ner settings for implementation, characteristics of theindividuals involved, the implementation process, andthe characteristics of the context thought to influenceoutcomes.

Analysis and SynthesisCandidate theories were identified through a close

reading of texts by two reviewers (P. O’H. and K.N.). Explicit theories were noted, and where absent,implicit theories deduced from the elements of the in-terventions. Data synthesis involved the two reviewersindependently assessing each paper, identifying com-mon components from the data extraction forms,and reflecting on candidate theories before comingtogether to discuss findings and achieve a consensusregarding the explanatory power of each. All authorsthen assessed the candidate theories and providedfeedback. We privileged empirical papers in our final

798 Vol. 56 No. 5 November 2018O’Halloran et al.

analysis as these represent studies closest to particularpractice contexts.

ResultsStudy Selection

Our initial search of databases produced 5687 doc-uments (Figure 1). Removing duplicates left 3562 doc-uments. Title review looking for documents relevantto advance care planning and end-stage kidney diseaseby two reviewers (K. N. and P. O’H.) excluded 3251 pa-pers, with abstract review reducing this to 91 papers.The full text of these papers was reviewed by the tworeviewers and a decision taken at this stage by the re-view authors to exclude 20 papers published before2000 in the interests of organizational and clinicalrelevance to contemporary practice. This left 62 docu-ments, of which 16 were opinion pieces, eight were

5687 records iden�fied through database searching.

CINAHL 56EMBASE 1796MEDLINE 1314PsycINFO 82CENTRAL 1284Science Direct 56Science Direct – recommended ar�cles 903Google Scholar 196

3562 records screened by �tle

311 records screened by abstract

62 records included in the reviewGuidelines/policies: 8Opinion pieces: 16Empirical papers: 38

91 full text records assessed for eligibility

Fig. 1. Flowchart illustrati

guidelines or policies, and 38 were empirical papersthat formed the basis of this review.

Study CharacteristicsStudy Designs. The 38 empirical studies were ofvarious designs (Supplemental Table 1). Three wereliterature reviews19,20,33; five were RCTs34e38; andanother39 a secondary analysis of an RCT.38 Threewere nonequivalent group designs40e42; one single-group test-retest study43; nine were qualitativestudies44e52; and 16 observational studies.53e68

Methodological Quality. Thirteen studies were classi-fied as ‘‘strong’’19,20,46,47,49e51,53,55,60,63,65, includingone RCT38; and 10 as ‘‘weak’’33,34,40,41,43,48,54,59,62,67;with the remaining 14 classified as ‘‘moderate.’’ Onepaper42 reported a protocol, so a judgment was notmade.

220 records excluded as irrelevant

2125 records excluded as duplicates

3251 records excluded as irrelevant

9 records excluded as irrelevant

Papers pre-2000 excluded: 20

ng the search process.

Vol. 56 No. 5 November 2018 799Advance Care Planning in Kidney Failure

Main Objectives of the Studies. Intervention studiessought to evaluate the effect of advance care planningon use of services and on congruence of care at end-of-life with patients’ wishes,19,20,37 or on congruencebetween patients and surrogates.34,36,38,39 Otherstudies focused on interventions to improve uptakeof advance care plans,35,41 or to encourage profes-sionals to support patients in making advance careplans.40,43 Most of the observational and qualitativestudies concentrated on establishing patients’ atti-tudes (or those of professionals55 or family mem-bers49) toward death, dying, end-of-life care,discussing treatment preferences, withdrawal of treat-ment, and making advance care plans or advance di-rectives. Exceptions were those that looked forcongruence between advance care plans and clinicaloutcomes,53 and between surrogates’ and patients’views on treatment preferences44,61; and those thatsought to examine the prevalence and contents ofadvance directives60,68 to evaluate patient-physiciancommunication about end-of-life care57; to explorethe association between advance care planning andwithdrawal from dialysis, use of hospice, and locationof death59; and to describe how seriously ill patientscan be assisted to identify a surrogate.58 Two studieswere intended to inform interventions for stafftraining.48,51

Study Populations. With the exception of three studiesthat focused on health professionals40,55 and familymembers49, all studies were of patients with end-stage kidney disease, some including family mem-bers/surrogate decision-makers, and three includingpatients with other long-term conditions.34,37,41 A mi-nority specified further characteristics for patient in-clusion, such as aged over 50 years34; or thought tobe at risk of death in the next six42, 1238, or24 months37; or with significant comorbidity.43

Advance Care Planning Interventions. ThreeRCTs34,36,37 evaluated patient-centered advance careplanning: a one-hour interview delivered by a trainedfacilitator to patients and their surrogates to help pre-pare surrogates to make decisions that honor thepatient’s preferences for care, leading to documenta-tion of preferences. Song et al.38 took a similarapproach but in two sessions over two weeks. Simonet al.52 a qualitative study, reported an advance careplanning process comprising trained facilitators, aworkbook with suggestions for levels of intervention,and organizational processes that audit and operation-alize advance care planning in medical records andtreatment plans. Perry et al.35 used trained peer men-tors who had three face-to-face and five telephonecontacts with patients over a two- to four-monthperiod.

Outcomes of Advance Care Planning InterventionsThe RCT intervention studies showed statistically

significant improvements in congruence in statementof treatment preferences,34,36,38 surrogate decision-making confidence,38 patient’s perceived quality ofcommunication,34,36 and decisional conflict34; an in-crease in completion of advance directives; and com-fort in discussing them.35 Unexpected results weremore intervention patients (37.7%) choosing to with-draw from dialysis than controls (17%)dalthoughthis had not been discussed in advance care plan-ning37dand a shift among participants toward usingall possible life-sustaining treatments to prolong theirlives.36 The non-RCT intervention studies reportedthat electronic medical record reminders improvedadvance directive documentation41; that nurses self-reported increased efficacy in relation to upholdingpatients informed choices40; and that there was no ef-fect on number of advance directives.43 In their reviewof RCTs or quasi-RCTs, Lim et al.20 reported no effecton hospital admissions or use of treatments with life-prolonging or curative intent or on following patient’swishes at end of life.

Program Theories in Relation to How Advance CarePlanning Can Be Successfully ImplementedLuckett et al.19 reflect the broad concept of advance

care planning found in most studies when they charac-terize it as

‘‘ .. a process of reflection, and discussion betweena patient, their family and healthcare providers forthe purpose of clarifying values, treatment prefer-ences and goals of end-of-life care. Advance careplanning provides a formal means of ensuringhealthcare providers and family members are awareof patients’ wishes for care if they become unable tospeak for themselves. Advance care planning is apatient-centered initiative that promotes shareddecision-making, which may include the patientcompleting an advance directive that documentstheir wishes and/or the appointment of a substitutedecision-maker.’’ (P. 762)

The theory is that once the patient, their family orsurrogate, and the health professionals have a com-mon understanding of the patient’s preferences forend-of-life care, then they will be more likely to acttogether to enact those preferences, even if the pa-tient loses capacity. A subtheme is that the patientwill be less likely to choose (possibly futile) life-prolonging treatments (such as resuscitation and dial-ysis) at end of life and more likely to choose palliativecare services and hospice care, which will enhancetheir quality of life until death. These views on the

800 Vol. 56 No. 5 November 2018O’Halloran et al.

nature and outcomes of advance care planning are re-flected in many guidelines and opinion pieces.7,69e74

The empirical studies provide more detailed expla-nations as to how advance care plans may work.Several studies report that the advance care planningprocess helps to prepare both the patient and the sur-rogate to make better decisions by helping the patientto clarify values and develop skills related to initiatingdiscussions, and by actively involving the surrogate tohelp prepare for the emotional burden of decision-making.34,36e39 Systematically, eliciting patient prefer-ences for end-of-life care allows concordant care tobe provided.42 Other studies focused on the role ofprofessionals, theorizing that addressing concernsabout advance care planning, optimizing skills,and clarifying processes will legitimize advance careplanning and encourage professionals to initiatediscussions.40,48

Some studies based their approach on previouslypublished theories. Five papers reporting fourRCTs34,36e39 described attempts to help the patientexamine their own belief systems and integrate new in-formation, based on the representational approach topatient education. This seeks to illuminate a person’sillness representationdthe set of thoughts that a per-son has about a health problem in five dimensions:identity, cause, timeline, consequences, and cure/con-troldto provide an opportunity for the person toreflect and to allow the clinician to provide individual-ized, acceptable information.75e78 One study66 drewon components of the theory of planned behavior79

to argue that a person who holds strong beliefs aboutdeath and dying and has a positive evaluation ofadvance care planning will more likely have a positiveattitude toward making an advance care plan.

Calvin et al.45,54 developed a new ‘‘theory of per-sonal preservation.’’ Patients on dialysis are thoughtto experience three phases: a redefined sense of selfas they acknowledge a shortened lifespan; a responseto this redefinition that asserted their individuality interms of their resilience in the face of life’s challenges,and finding meaning and optimism in their suffering,often through prayer or trusting in God’s will; andfinally living in the tension of acting responsibly tocare for themselves and others, while managing therisks and uncertainty associated with treatment byfallible professionals. They did not see advance careplans as a means to maintain personal autonomy,preferring to leave such decisions to trusted familymembers, and to focus on negotiating present chal-lenges. This theory (which has elements in commonwith that of Goff et al.51) is intended to informcommunication around advance care planning. Simi-larly, Simon et al.52 propose a model in which preexist-ing ideas formed by ‘‘witnessing illness in self and

others’’ and ‘‘acknowledging mortality’’ lead to reflec-tion on states of dysfunction in which patients wouldnot want to live. These combine with the advancecare planning process to produce a degree of peaceof mind in that participants had planned to avoidburdensome states and reduced family stress.

Factors That May Help or Hinder theImplementation of Advance Care PlanningThe Characteristics of the InterventionOne review concluded that interventions do not

give comprehensive attention to patient, caregiver,health professional, and system-related factors.19 Giv-ing structured feedback to facilitators on their deliveryof advance care planning was thought to enhance theprocess.37,38 An electronic medical recordebasedreminder appeared to improve completion of advancedirectives by prompting discussions within the outpa-tient visit workflow.41 Tailoring advance directivedocumentation to patient needs and keeping docu-ments short was thought to encourage completion,63

whereas administrative and legal complexities discour-aged completion.58

The Inner and Outer Settings for ImplementationIn terms of the inner organizational settingdwhich

includes organizational structure, communications,culture, and capacity for change29dimplementationwas enhanced by educating staff on the benefits ofadvance care planning.19 Advance care planning wasmore likely to be used when staff trained to facilitateappropriate discussions were available,34,36-38,48 whenphysicians actively recommended hospice care,59 andwhere there were processes reflecting organizationalendorsement such as administrative requirement foradvance directives,53 and a policy of providing supportfor staff education, patient information, and relatedsafety and quality systems.40 Overall, ensuring staffhave time for advance care planning as ‘‘core busi-ness’’ supported implementation.19

On the other hand, the pressures of (and focus on)routine clinical work,33,46 a ‘‘conveyor-belt’’ culture inrelation to providing hemodialysis,50 lack of funds tosupport staff training,48 and poor continuity of medi-cal care,57 all militated against uptake of advancecare planning.Turning to the outer settingdwhich includes

external policies and incentives, patient needs and re-sources, and relationships with other (peer) organiza-tions29dwider cultural differences were reported toexert profound effects on advance care planning pro-cesses. In general, it appears that advance care plan-ning is seen as a more normal part of health care in

Vol. 56 No. 5 November 2018 801Advance Care Planning in Kidney Failure

the U.S.dall but one of the intervention studies40

were carried out there. However, within the U.S., pa-tients and families with lower socioeconomic status re-ported feeling disempowered in interactions with thedialysis team and therefore less likely to build trustand complete advance care plans.51 African Americanswere described as more responsive than whites to anintervention that was highly inclusive of family surro-gates, as this may align with their familial, religious,and communal frame of reference39 and to be opento peer interventions because of a preference fororal traditions35. Similarly, Asian Americans and Ha-waiians (from cultures where family involvement anddiscussion, and collective decision-making, are impor-tant) expressed a preference for discussing advancecare planning with family rather than health profes-sionals.66 African and Asian American patients werefound to be more likely to engage in end-of-life discus-sions, but African Americans were less likely to preferdialysis withdrawal at end of life and to complete a do-not-resuscitate directive.67 In one rural study, reluc-tance of many patients to travel the necessary distancefor tertiary care contributed to the low prevalence ofin-hospital death.59

A Saudi Arabian study64 reported that Islam is inprinciple compatible with advance care planning,but that this was rarely discussed and patients weregenerally against ‘‘do-not-resuscitate’’ decisionsdatti-tudes echoed in a study in the Republic of Ireland,where it was also reported that advance directiveshad no legal status, with very few people aware of theirexistence.56 Miura et al.61 stated that Japanese cultureincludes the belief that one’s wishes are intuitivelyknown to others, leading to an erroneous assumptionthat family members can correctly predict patient’spreferences for cardiopulmonary resuscitation orcontinuation of dialysis at end of life. Physicians maybe paternalistic and families may see withdrawal oftreatment as abandoning their duty to the patient,so that the physician defers to the family even againstthe patient’s advance directive, leading to inappropri-ately aggressive treatment at end of life. Similarly,Yee et al.55 perceive Singaporeans may not be readyfor advance care planning, as it can be confusedwith advance directives (and advance directives witheuthanasia) by both health care professionals and pa-tients. The family also plays an important role inhealth-related decisions and may sometimes disagreewith patients’ preferences. A Spanish study reportedthat people are unfamiliar with advance directivesand fear that physicians may withhold necessary treat-ments, leading to low uptake.68 A study in the UnitedKingdom argued that there is a need for a culture shiftfrom a ‘‘disease-focused’’ model toward a ‘‘holisticcare-based’’ approach, normalizing discussions aboutpreferences, priorities, and future care.50

Characteristics of IndividualsA common finding was that patients and their fam-

ilies found it difficult to initiate conversations aboutfuture care with professionals, especially nephrolo-gists.19,45,49,51 Patients expected professionals to takethe initiative,19,33,46,47 but professionals were reluctantfor fear of upsetting patients and fam-ilies20,37,43,49,50,55,62 and in the case of nurses, upset-ting senior staff.19 One study argued that a focus ontechnology had taken nephrologists away from offer-ing emotional and spiritual support.65

Patients were thought to overestimate their life ex-pectancy19 and to distance themselves from thoughtsof death,45,52 thus reducing their sense of urgency tohave advance care planning discussions. Surrogateswere often unaware of patients’ views and helddifferent positions on care at end of life.19,44,55,61

A trusting relationship with the person proposingadvance care planning seemed to make acceptanceby the patient35,51,56,68 and family members49 morelikely, with lack of trust, making discussions moredifficult.

Theory SynthesisIn the following section, we seek to examine the

interaction between the intervention (I), context(C), mechanism (M), and outcome (O) in relationto advance care planning, with a view to offering a the-ory to explain and support implementation(Figure 2).Implementation of advance care planning entails

two major intervention stages (I1 and I2). First, thereis training for clinical staff that addresses concerns, op-timizes skills, and clarifies processes (I1). This helpsclinical staff to appreciate the usefulness of advancecare planning and to gain confidence to initiate dis-cussions with patients and their surrogates (M). Thesecond stage (I2) is when the advance care planningdocumentation and processes are implemented.These should be simple, individually tailored, cultur-ally appropriate, and actively involve surrogates.They are designed to explore the patient’s under-standing of their condition, to elicit what the patientvalues in relation to their health and relationships,and on this basis to document their preferences forcare and treatment in the event of incapacity.These processes have their effect as patients develop

trust in staff, clarify their values and beliefs about theircondition, and learn skills to participate in discussionswith clinical staff and others. Consequently, patients’preferences are made known to clinical staff and sur-rogates, and surrogates are prepared emotionally fordecision-making (M). These interventions, workingthrough the mechanisms described, lead to greater

Clinical staff recognize utility of ACPand gain confidence to start discussions with patients and surrogates. (M)

A simple, individually tailored, culturally appropriate approach to ACP; actively involving surrogate; designed to explore the patient’s understanding of their condition, and to elicit and document values and preferences for care and treatment in the event of incapacity. (I2)

Patient develops trust in staff, clarifiesvalues and beliefs about condition, learns skills to participate in discussions. Surrogate prepared emotionally for decision-making.Patient’s preferences made known to clinical staff and surrogates (M)

Greater congruence of patient/surrogate preferences; increased quality of communication, confidence of surrogate decision-making, and completion of ADs.Diverse impact on preferences for life-sustaining treatments. (O)

Training for clinical staff that addresses concerns about ACP, optimizes skills, and clarifies processes. (I1)

ACP is core business,endorsedor required by organization, and attracts systems support (C+)

Suitably trained staff available, and receive reminders and feedback on ACP process. (C+)

ACP is culturally appropriate and takes account of patient’s perception of their situation. (C+)

Patient is prepared to consider death and dying, and has positive view of ACP. (C+)

Patient overestimates life-expectancy; patient, family, clinical staff reluctant to initiate discussion (C-)

Routine pressures and focus on technology diverts time and resources from ACP. (C-)

ACP administration complex; lack of funded training; poor continuity of medical care (C-)

Key:

ACP = Advance care planningAD = Advance directiveI = InterventionsM = MechanismsC+ = Helpful contextC- = Hindering contextO = Outcomes

Fig. 2. Model for the implementation of advance care planning with patients who have end-stage kidney disease.

802 Vol. 56 No. 5 November 2018O’Halloran et al.

congruence between patient and surrogate on treat-ment preferences; increased quality of communica-tion between patients and clinical staff; greaterconfidence in surrogates about decision-making onpatients’ behalf; and an increased rate of completionof advance directives. Preferences for life-sustainingtreatments may increase or decrease under the influ-ence of the sociocultural context (O).

These outcomes are hindered by a number of orga-nizational issues: lack of funded training opportu-nities, administrative and legal complexities aroundadvance care planning, the prioritization and pres-sures of routine care, and a focus on technological(rather than personal and spiritual) aspects of care.

Barriers may also arise from interpersonal issues:when the patient overestimates life expectancy andtherefore does not appreciate the relevance of discus-sions about future care; and when the patient, family,and/or clinical staff are reluctant to initiate discus-sions about treatment decisions in the event of deteri-oration because of fears of causing distress (C).On the other hand, advance care planning is more

likely to succeed where it is acknowledged as core busi-ness and required or endorsed by the organizationand therefore attracts resources to support educationand quality assurance; when suitably trained staff areavailable and receive reminders and feedback on im-plementation; when the process and documentation

Vol. 56 No. 5 November 2018 803Advance Care Planning in Kidney Failure

are culturally appropriate and take account of the pa-tient’s perception of their situation; and when the pa-tient is prepared to consider death and dying and haspositive view of advance care planning.

DiscussionMost of the included papers (25/38) reported

observational or qualitative designs. These provideuseful information on the perceptions, attitudes, andbehavior of patients, families, and professionals inrelation to death, dying, and advance care planning.However, there were only eight intervention studiesfound, of which five were RCTs, with 593 patientsand 390 relative participants. RCT outcomes werediverse, showing statistically significant improvementsin the quality of communication and confidence indecision-making, but no effect on hospital admissionsor use of treatments with life-prolonging or curativeintent, or on following patient’s wishes at end of life.Indeed, only one RCT37 attempted to measurewhether patients received care in accord with theirwishes, and low numbers precluded statistical analysis.No RCTs effectively investigated impact on use ofhealth services, hospital admissions, or treatment,which ruled out any evaluation of cost-effectiveness.Of the RCTs, we judged only one38 to be methodolog-ically strong; the others to be moderate or, in one case,weak.

The commonly held theory that advance care plan-ning will lead to a shared understanding of patientpreferences for end-of-life care, and subsequent enact-ment of those preferences, with less recourse to life-prolonging treatments and greater use of palliativecare services and hospice care, is only partially sup-ported by the evidence. One study37 reported thatmore intervention patients chose to withdraw fromdialysis than controls, despite withdrawal not beingdiscussed in advance care planning, but this resultshould be interpreted with caution as the paper doesnot specify numbers of participants for this subgroupanalysis. Song et al.36 report an increase in preferencefor life-sustaining treatment in their interventiongroup, attributing this to cultural and religious moresamong their African American sample. Again, a smallsample size (n ¼ 19) invites cautious interpretation. Itdoes seem that advance care planning can help clarifypreferences and improve communication, but otheroutcomes remain unsupported or equivocal. Thesefindings are consistent with two recent reviews,19,20

which concluded that the evidence base is weak inthis population.

Unpicking the program theories in relation to howadvance care planning could be successfully imple-mented, and identifying contexts that may help or

hinder, has allowed us to synthesize an integrated the-ory to explain and support implementation. Unsur-prisingly, the first stage is staff training. However, it isimportant to note the focus of the trainingdtoaddress staff concerns, clarify processes, and optimizeskillsdand its goaldto persuade staff of the usefulnessof advance care planning and to give them confidenceto start discussions with patients and their relatives.Training is less likely to be effective if it takes forgranted the usefulness of advance care planning, mar-ginalizes staff concerns, and does not engender confi-dence in the process.Similarly, when addressing the second stage of im-

plementation (the approach to the patient and rela-tives), the focus and goal of the approach are keyingredients for success. The advance care planningdocument and approach should be culturally appro-priate, actively involve the surrogate, explore the pa-tient’s understanding of their condition, and elicitand document their values and preferences for care.The goal is for the patient (and their surrogate) totrust staff sufficiently that they are able to share theirvalues and beliefs and gain confidence to enter intodiscussions with clinicians. The few detailed reportsof advance care planning interventions largelydescribed relatively short (e.g. one hour) guided inter-actions between the patient, their surrogate, and atrained facilitator, helping the patient and surrogateto clarify values and develop skills related to initiatingdiscussions, preparing for decision-making, and lead-ing to a document recording patient preferences.One hour is not a long time to spend making a planfor one’s care in the event of deterioration, but it issignificant in terms of professional time when thereare competing priorities. This introduces the impactof context. In organizational terms, the umbrellaconcept is that advance care planning should becore business. That is, it should attract support inthe form of policy, administrative processes, resources(funded training, staffing levels), quality assurance,and accountability. Advance care planning that is an‘‘add-on’’ is likely to be swamped by the pressures ofroutine care. Interestingly, only two studies directly ad-dressed organizational issues: one that reportedtraining nurses to offer advance care planning in thecontext of a supportive policy framework of educative,patient information, safety, and quality systems40; andthe other examining the effectiveness of an electronicmedical recordebased reminder in improvingadvance directive documentation.41 The vast majorityof studies focused on patient-staff interaction in rela-tive isolation from the organizational context.Surrounding and pervading organizations are the

cultural and personal values that heavily influence theresponses of both staff andpatients to discussions aboutcare at end of life. Advance care planning has largely

804 Vol. 56 No. 5 November 2018O’Halloran et al.

evolved within Western democracies and been devel-oped by professional elites. Consequently, theapproach typically embodies a cultural commitmentto individual autonomy, which may sit uneasily withthose with lower socioeconomic status, or from cul-tures, which privilege the family in decision-making.Similarly, professionals may assume advance care plan-ning will entail refusing or curtailing treatment, whenpatients are inclined to continue and maximize inter-ventions. Added to this is a reluctance on all sides toraise the issues of poor prognosis and death. For thesereasons, it is important that the advance care planningprocess and documentation are culturally appropriate.It should take into account patient and family percep-tions of their situation, including life expectancy andthe likely efficacy of life-sustaining interventions; reli-gious and cultural understandings of death, dying,and the place of medical care; together with prefer-ences in relation to involvement of the family indecision-making. Given that patients have difficulty inraising advance care planning with professionals, thereis a role for patient representative bodies and the char-itable sector in educating patients and families in rela-tion to advance care planning and advocating forhealth care organizations and professionals to takethe lead in introducing it to patients.

Strengths and LimitationsIntervention studies were relatively few and often

small, which limited our ability to comment on clinicalor cost-effectiveness of advance care planning in thispopulation. Other studies had diverse aims and designs,making synthesis a challenge.However, this diversity didserve to provide data on organizational and cultural as-pects. Having said that, most studies did not set out toexplain how advance care planning might work or toidentify helping or hindering factors, which meant theauthors had to glean some of these insights from thenarrative and discussion sections of the papers.

TableRecommendations for the Development and Implementatio

Kidney D

� Patient representative bodies and the charitable sector should educatadvocate for health care organizations and professionals to take the le

� Organizations should make necessary preparations for implementingpolicies and administrative and quality assurance processes, while procare planning effectively.

� Staff trainers should seek to address concerns, optimize skills, and clacare planning and to give them confidence to start discussions with p

� Professionals should develop an approach to advance care planning tinvolves surrogates, and explores the patient’s understanding of their

� When implementing advance care planning, professionals should aimdevelop skills that will facilitate their participation in decision-makingfaithfully for the patient, should that become necessary.

� Researchers should develop robust clinical trials to investigate the impmaking and emotional burden; enactment of patient preferences forpalliative care services and hospice care. The impacts of organizationafeasibility of advance care planning should also be investigated.

Conclusions and RecommendationsRobust clinical trials are needed in this population

to investigate the impact of advance care planningon enactment of patient preferences for end-of-lifecare; recourse to life-prolonging treatments; and useof palliative care services and hospice care.Nevertheless, it is evident that for patients with end-

stage kidney disease, advance care planning is muchmore than an administrative or merely intellectualtask. On the contrary, it is an intensely human processin which a vulnerable patient is invited to considertheir own deterioration and death, and to make plansfor navigating various threatening possibilities.Despite this, there is already some evidence of positiveoutcomes for patients and their families in the form ofcongruence in statement of treatment preferences,surrogate decision-making confidence, improvementsin perceived quality of communication and decisionalconflict; an increase in completion of advance direc-tives; and comfort in discussing them. Consequently,professionals have a responsibility to offer advancecare planning to patients and to support them duringthe process. Given the significance and complexity ofthis task, professionals need organizational support inthe form of policy endorsement; quality assurance;administrative systems; training that addresses con-cerns, optimizes skills, and clarifies processes; andstaffing levels that allow time for proper attention toimplementation. This will allow them to develop anapproach to advance care planning that is designedto explore the patient’s understanding of their condi-tion, and to elicit and document values and prefer-ences for care and treatment in the event ofincapacity, but is also simple, individually tailored,culturally appropriate, and actively involvessurrogates.Bearing in mind the profound impact of cultural

and organizational context, we recommend that pa-tient groups, professionals, and their organizations

2n of Advance Care Planning for Patients With End-Stageisease

e patients and families in relation to advance care planning andad in introducing it to patients.advance care planning as core business by developing appropriateviding resources for training and for sufficient staff to offer advance

rify processes, aiming to persuade staff of the usefulness of advanceatients and their relatives.hat is simple, individually tailored, culturally appropriate, activelycondition.to build trust with patients and their families; help the patient; and build confidence in surrogates for their role in speaking

act of advance care planning on patient and surrogate decision-end-of-life care; recourse to life-prolonging treatments; and use ofl context and cultural and personal values on the acceptability and

Vol. 56 No. 5 November 2018 805Advance Care Planning in Kidney Failure

take a diagnostic approach to their particular situationwhen implementing advance care planning. In otherwords, they should consider the beliefs and expecta-tions of the surrounding culture, their own resources,and how they can best organize advance care planningtraining and implementation to trigger the mecha-nisms and achieve the outcomes identified in our im-plementation model. Specific recommendations forstakeholders are offered in Table 2.

Disclosures and AcknowledgmentsThis work was supported by The Dunhill Medical

Trust [grant number R428/0715]. Dr. Jackie Boylanand Dr. Judith Cole assisted in data extraction forthis review.

The authors have nothing to disclose.

References

1. Davison SN, Jhangri GS, Johnson JA. Cross-sectional val-idity of a modified Edmonton symptom assessment system indialysis patients: a simple assessment of symptom burden.Kidney Int 2006;69:1621e1625.

2. USRDS. 2015 USRDS Annual Data Report Volume 1:CKD in the United States. United States Ren. Data Syst.2016.

3. Brown MA, Collett GK, Josland EA, Foote C, Li Q,Brennan FP. CKD in Elderly patients managed without dial-ysis: Survival, Symptoms, and quality of life. Clin J Am SocNephrol 2015;10:260e268.

4. Verberne WR, Geers ABMT, Jellema WT, Vincent HH,van Delden JJM, Bos WJW. Comparative Survival among old-er adults with advanced kidney disease managed Conserva-tively versus with dialysis. Clin J Am Soc Nephrol 2016;11:633e640.

5. Douglas C, Murtagh F, Chambers E, Howse M,Ellershaw J. Symptom management for the adult patientdying with advanced chronic kidney disease: a review ofthe literature and development of evidence-based guidelinesby a United Kingdom Expert Consensus Group. Palliat Med2009;23:103e110.

6. Davison SN, Jhangri GS. Existential and supportive careneeds among patients with chronic kidney disease. J PainSymptom Manage 2010;40:838e843.

7. Arulkumaran N, Szawarski P, Philips BJ. End-of-life carein patients with end-stage renal disease. Nephrol DialTranspl 2012;27:879e881.

8. Davison SN. Advance care planning in patients withchronic kidney disease. Semin Dial 2012;25:657e663.

9. Royal College of Physicians. Advance care planning.Concise Guidance to Good Practice series. London: RoyalCollege of Physicians, 2009. Available at: http://www.rcplondon.ac.uk/resources/concise-guidelines-advance-care-planning.

10. Detering KM, Silveira MJ. Advance care planning andadvance directives Arnold RM, Savarese DM, editors.

UpToDate 2018. Available at: https://www.uptodate.com/.Accessed February 19, 2018.

11. Mullick A, Martin J, Sallnow L. An introduction toadvance care planning in practice. BMJ 2013;347:1e6.

12. Bernacki RE, Block SD. American College of physicianshigh value care task Force. Communication about seriousillness care goals. JAMA Intern Med 2014;174:1994.

13. Krouse RS, Rosenfeld KE, Grant M, et al. Palliative careresearch: issues and opportunities. Cancer Epidemiol Bio-markers Prev 2004;13:337e339.

14. Houben CHM, Spruit MA, Groenen MTJ, Wouters EFM,Janssen DJA. Efficacy of advance care planning: a systematicreview and meta-analysis. J Am Med Dir Assoc 2014;15:477e489.

15. Brinkman-Stoppelenburg A, Rietjens JA, van derHeide A. The effects of advance care planning on end-of-life care: a systematic review. Palliat Med 2014;28:1000e1025.

16. Department of Health. National service framework forrenal services part two : Chronic kidney disease, acute renalfailure and end of life care. London: Department of Health,2005.

17. National Institute for Health and Clinical Excellence.Chronic kidney disease quality standard 2012. Available at:http://publications.nice.org.uk/chronic-kidney-disease-quality-standard-qs5. Accessed July 1, 2014.

18. Crail S, Walker R, Brown M. Renal supportive and palli-ative care: position statement. Nephrology 2013;18:393e400.Available at: http://onlinelibrary.wiley.com/doi/10.1111/nep.12064/full. Accessed August 27, 2014.

19. Luckett T, Sellars M, Tieman J, et al. Advance care plan-ning for adults with CKD: a systematic integrative review. AmJ Kidney Dis 2014;63:761e770.

20. Lim CED, Ng RWC, Cheng NCL, Cigolini M, Kwok C,Brennan F. Advance care planning for haemodialysis pa-tients. Cochrane database Syst. Rev 2016;(7):CD010737.

21. Lovell A, Yates P. Advance Care Planning in palliativecare: a systematic literature review of the contextual factorsinfluencing its uptake 2008-2012. Palliat Med 2014;28:1026e1035.

22. Lund S, Richardson A, May C. Barriers to advance careplanning at the end of life: an explanatory systematic reviewof implementation studies. PLoS One 2015;10:e0116629.

23. Holley JL, Davison SN. Advance care planning in CKD:clinical and research opportunities. Am J Kidney Dis 2014;63:739e740.

24. Wong G, Greenhalgh T, Westhorp G, Buckingham J,Pawson R. RAMESES publication standards: realist synthe-ses. BMC Med 2013;11:21.

25. Pawson R, Tilley N. Realistic Evaluation. London: SAGEPublications Ltd, 1997.

26. Higgins A, O’Halloran P, Porter S. The management oflong-term Sickness absence in Large public sector health-care organisations: a realist evaluation using MixedMethods. J Occup Rehabil 2015;25:451e470.

27. Archer MS. Structure, Agency, and the Internal Conver-sation. Cambridge: Cambridge University Press, 2003.

28. Cheyne H, Abhyankar P, McCourt C. Empoweringchange: realist evaluation of a Scottish Government

806 Vol. 56 No. 5 November 2018O’Halloran et al.

programme to support normal birth. Midwifery 2013;29:1110e1121.

29. Damschroder LJ, Aron DC, Keith RE, Kirsh SR,Alexander J, Lowery JC. Fostering implementation of healthservices research findings into practice: a consolidatedframework for advancing implementation science. Imple-ment Sci 2009;4:50.

30. Lim CED, Siow S, Ho KEE, et al. Advance care planningfor haemodialysis patients (Protocol). Available at. CochraneDatabase Syst Rev 2013:CD010737 http://doi.wiley.com/10.1002/14651858.CD010737. Accessed September 8, 2017.

31. Critical Appraisal Skills Programme. Critical AppraisalSkills Programme. Available at: http://www.casp-uk.net/casp-tools-checklists. Accessed August 30, 2017.

32. O’Halloran P, Scott D, Reid J, Porter S. Multimedia psy-choeducational interventions to support patient self-care indegenerative conditions: a realist review. Palliat SupportCare 2014:1e14. Available at: http://journals.cambridge.org/abstract_S1478951514001229. Accessed December 3,2014.

33. Noble H, Meyer J, Bridges J, Kelly D, Johnson B. Patientexperience of dialysis REFUSAL or withdrawalda review ofthe literature. J Ren Care 2008;34:94e100.

34. Briggs LA, Kirchhoff KT, Hammes BJ, Song MK,Colvin ER. Patient-centered advance care planning in specialpatient populations: a pilot study. J Prof Nurs 2004;20:47e58.

35. Perry E, Swartz J, Brown S, Smith D, Kelly G, Swartz R.Peer mentoring: a culturally sensitive approach to end-of-life planning for long-term dialysis patients. Am J KidneyDis 2005;46:111e119. Available at: http://www.sciencedirect.com/science/article/pii/S0272638605004610. AccessedSeptember 29, 2014.

36. Song M-K, Donovan HS, Piraino BM, et al. Effects of anintervention to improve communication about end-of-lifecare among African Americans with chronic kidney disease.Appl Nurs Res 2010;23:65e72.

37. Kirchhoff KT,HammesBJ, KehlKA, Briggs LA, BrownRL.Effect of a disease-specific advance care planning interventionon end-of-life care. J Am Geriatr Soc 2012;60:946e950.

38. Song M-K, Ward SE, Fine JP, et al. Advance care plan-ning and end-of-life decision making in dialysis: a Random-ized controlled trial Targeting patients and their surrogates.Am J Kidney Dis 2015;66:813e822.

39. Song M-K, Ward SE, Lin F-C, et al. Racial differences inoutcomes of an advance care planning intervention for dial-ysis patients and their surrogates. J Palliat Med 2016;19:134e142. Available at: http://online.liebertpub.com/doi/10.1089/jpm.2015.0232. Accessed May 23, 2017.

40. Seal M. Patient advocacy and advance care planning inthe acute hospital setting. Aust J Adv Nurs 2007;24:29e36.

41. Hayek S, Nieva R, Corrigan F, et al. End-of-Life careplanning: improving documentation of advance directivesin the outpatient clinic using electronic medical records.J Palliat Med 2014;17:1348e1352. Available at: http://online.liebertpub.com/doi/abs/10.1089/jpm.2013.0684.

42. Eneanya ND, Goff SL, Martinez T, et al. Shared decision-making in end-stage renal disease: a protocol for a multi-center study of a communication intervention to improveend-of-life care for dialysis patients. BMC Palliat Care 2015;14:30.

43. Weisbord SD, Carmody SS, Bruns FJ, et al. Symptomburden, quality of life, advance care planning and the poten-tial value of palliative care in severely ill haemodialysis pa-tients. Nephrol Dial Transpl 2003;18:1345e1352. Availableat: https://academic.oup.com/ndt/article-lookup/doi/10.1093/ndt/gfg105.

44. Hines SC, Glover JJ, Babrow AS, Holley JL, Badzek LA,Moss AH. Improving advance care planning by accommoda-ting family preferences. J Palliat Med 2001;4:481e489. Avail-able at: http://www.liebertonline.com/doi/abs/10.1089/109662101753381629. Accessed May 23, 2017.

45. Calvin AO. Haemodialysis patients and end-of-life deci-sions: a theory of personal preservation. J Adv Nurs 2004;46:558e566. Available at: http://doi.wiley.com/10.1111/j.1365-2648.2004.03030.x. Accessed May 23, 2017.

46. Davison SN, Simpson C. Hope and advance care plan-ning in patients with end stage renal disease: qualitativeinterview study. BMJ 2006;333:886.

47. Davison SN. Facilitating advance care planning for pa-tients with end-stage renal disease: the patient perspective.Clin J Am Soc Nephrol 2006;1:1023e1028. Available at:http://cjasn.asnjournals.org/cgi/doi/10.2215/CJN.01050306. Accessed May 23, 2017.

48. Bristowe K, Shepherd K, Bryan L, et al. The develop-ment and piloting of the REnal specific Advanced Commu-nication Training (REACT) programme to improve AdvanceCare Planning for renal patients. Palliat Med 2014;28:360e366. Available at: http://journals.sagepub.com/doi/10.1177/0269216313510342. Accessed May 23, 2017.

49. Calvin AO, Engebretson JC, Sardual SA. Understandingof advance care planning by family members of persons un-dergoing hemodialysis. West J Nurs Res 2014;36:1357e1373.Available at: http://journals.sagepub.com/doi/10.1177/0193945913514637. Accessed May 23, 2017.

50. Bristowe K, Horsley HL, Shepherd K, et al. Thinkingahead--the need for early Advance Care Planning for peopleon haemodialysis: a qualitative interview study. Palliat Med2015;29:443e450. Available at: http://journals.sagepub.com/doi/10.1177/0269216314560209. Accessed May 23,2017.

51. Goff SL, Eneanya ND, Feinberg R, et al. Advance careplanning: a qualitative study of dialysis patients and families.Clin J Am Soc Nephrol 2015;10:390e400.

52. Simon J, Murray A, Raffin S. Facilitated advance careplanning: what is the patient experience? J Palliat Care2008;24:256e264.

53. Anderson JE, Sikorski I, Finucane TE. Advance careplanning by or on behalf of peritoneal dialysis patients inlong-term care. Am J Kidney Dis 2006;48:122e127.

54. Calvin AO, Eriksen LR. Assessing advance care planningreadiness in individuals with kidney failure. Nephrol Nurs J2006;33:165e170. Available at: http://ovidsp.ovid.com/ovidweb.cgi?T¼JS&PAGE¼reference&D¼med5&NEWS¼N&AN¼16613411%0Ahttp://ovidsp.ovid.com/ovidweb.cgi?T¼JS&PAGE¼reference&D¼emed7&NEWS¼N&AN¼16613411.

55. Yee A, Seow YY, Tan SH, Goh C, Qu L, Lee G. What dorenal health-care professionals in Singapore think ofadvance care planning for patients with end-stage renal dis-ease? Nephrology (Carlton) 2011;16:232e238.

Vol. 56 No. 5 November 2018 807Advance Care Planning in Kidney Failure

56. Collins M, Lehane E. Perspectives on death, dying andadvance care planning from patients RECEIVING haemo-dialysis. J Ren Care 2013;39:5e11.

57. Janssen DJ, Spruit M, Schols JMG, van der Sande FM,Frenken L a, Wouters EFM. Insight into advance care plan-ning for patients on dialysis. J Pain Symptom Manage2013;45:104e113.

58. Song M-K, Ward SE. Disconnect between emergencycontacts and surrogate decision-makers in the absence ofadvance directives. Palliat Med 2013;27:789e792. Availableat: http://journals.sagepub.com/doi/10.1177/0269216312474486.

59. Schmidt RJ, Weaner BB, Long D. The power of advancecare planning in promoting hospice and out-of-hospitaldeath in a dialysis Unit. J Palliat Med 2015;18:62e66. Avail-able at: http://online.liebertpub.com/doi/abs/10.1089/jpm.2014.0031.

60. Feely MA, Hildebrandt D, Edakkanambeth Varayil J,Mueller PS. Prevalence and contents of advance directivesof patients with ESRD receiving dialysis. Clin J Am Soc Neph-rol 2016;17:17. Available at: http://login.ezproxy.library.ualberta.ca/login?url¼http://ovidsp.ovid.com/ovidweb.cgi?T¼JS&CSC¼Y&NEWS¼N&PAGE¼fulltext&D¼medp&AN¼27856490.

61. Miura Y, Asai A, Matsushima M, et al. Families’ and phy-sicians’ predictions of dialysis patients’ preferencesregarding life-sustaining treatments in Japan. Am J KidneyDis 2006;47:122e130. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0272638605015167. Accessed May23, 2017.

62. Rodrı́guez Jornet A, Ibeas J, Real J, Pe~na S, Martı́nezOca~na JC, Garcı́a Garcı́a M. [Advance directives in chronicdialysis patients]. Nefrologia 2007;27:581e592.

63. Nishimura A, Mueller PS, Evenson LK, et al. Patientswho complete advance directives and what they prefer.Mayo Clin Proc 2007;82:1480e1486.

64. Al-Jahdali HH, Bahroon S, Babgi Y, Tamim H, Al-Ghamdi SM, Al-Sayyari AA. Advance care planning prefer-ences among dialysis patients and factors influencing theirdecisions. Saudi J Kidney Dis Transpl 2009;20:232e239.

65. Davison SN. End-of-life care preferences and needs: per-ceptions of patients with chronic kidney disease. Clin J AmSoc Nephrol 2010;5:195e204.

66. Kataoka-Yahiro MR, Conde FA, Wong RS, Page V, Peller B.Advance care planning among Asian Americans and NativeHawaiians receiving haemodialysis. Int J Palliat Nurs 2010;16:32e40. Available at: http://www.magonlinelibrary.com/doi/abs/10.12968/ijpn.2010.16.1.46181.

67. Kurella Tamura M, Goldstein MK, P�erez-Stable EJ. Pref-erences for dialysis withdrawal and engagement in advancecare planning within a diverse sample of dialysis patients.Nephrol Dial Transpl 2010;25:237e242.

68. S�anchez-Tomero JA, Rodrı́guez-Jornet A, Balda S, et al.Exploring the opinion of CKD patients on dialysis regardingend-of-life and Advance Care Planning. Nefrologia 2011;31:449e456.

69. Brown MA, Crail SM, Masterson R, et al. ANZSN renalsupportive care 2013: opinion pieces [corrected]. Nephrology(Carlton) 2013;18:401e454. Available at: http://doi.wiley.com/10.1111/nep.12065. Accessed September 4, 2017.

70. Davison SN. The ethics of end-of-life care for patientswith ESRD. Clin J Am Soc Nephrol 2012;7:2049e2057.

71. Gomm S, Farrington K. End of Life Care in AdvancedKidney Disease: A Framework for Implementation. London:NHS Improving Quality, 2015. Available at: http://webarchive.nationalarchives.gov.uk/20160506191033/http://www.nhsiq.nhs.uk/11739.aspx.

72. Ontario Renal Network. Ontario Renal Network Pallia-tive Care Report. Toronto: Cancer Care Ontario, 2016. Avail-able at: http://www.renalnetwork.on.ca/common/pages/UserFile.aspx?fileId¼360709.

73. Russon L, Mooney A. Palliative and end-of-life care inadvanced renal failure. Clin Med J R Coll Physicians Lond2010;10:279e281.

74. Warwick G, Mooney A, Russon L, Hardy R. Planning,Initiating and Withdrawal of Renal Replacement TherapyUK Renal Association 6 th Edition Draft Version (Feb2013) Introduction Summary of clinical practice guidelinesfor Planning, Initiating and Withdrawal of Renal Replace-ment Therapy Summa. Bristol: UK Renal Association,2013. Available at: http://www.renal.org/docs/default-source/guidelines-resources/final-guideline-for-planning-initiation-withdrawal-of-rrt.pdf.

75. Leventhal H, Nerenz D, Steele D. Illness representationsand coping with health threats. In: Baum A, Singer J, eds.Handbook of psychology and health, IV. New York: Erlbaum,1984::221e252.

76. Leventhal H, Diefenback M. The active side of illnesscognition. In: Skelton J, Croyle R, eds. Mental representa-tion in health and illness. New York: Springer Verlag,1991::247e272.

77. Donovan HS, Ward S. A representational approach topatient education. J Nurs Scholarsh 2001;33:211e216.

78. Donovan HS, Ward SE, Song M-K, Heidrich SM,Gunnarsdottir S, Phillips CM. An update on the representa-tional approach to patient education. J Nurs Scholarsh 2007;39:259e265. Available at: http://doi.wiley.com/10.1111/j.1547-5069.2007.00178.x. Accessed August 23, 2017.

79. Ajzen I. The theory of planned behavior. Organ BehavHum Decis Process 1991;50:179e211. Available at: http://linkinghub.elsevier.com/retrieve/pii/074959789190020T.Accessed August 30, 2017.

Supplemental Table 1Empirical Studies

First Author, Country, andObjectives

Population, Setting, andIntervention

Design and MethodologicalRigor (Strong/Moderate/

Weak) Key ResultsHow the Intervention Was

Thought to Work

Contextual FeaturesThought to Influence

Implementation

Systematic reviewsLim (2016), Australia.Objectives: To determine

whether advance careplanning inhemodialysis patients,compared with no orless-structured forms ofadvance care planning,can result in fewerhospital admissions orless use of treatmentswith life-prolonging orcurative intent, and ifpatient’s wishes werefollowed at end of life.

Systematic review: RCTs orquasi-RCTs withhemodialysis patients.

Intervention: ACP is wherethe responsible physicianhelps the patient (withtheir family) to reflect ontheir future care, consultsthe patient on theirpriorities and preferencesand levels of care, managestransition from active topalliative care, and,together with anyspokesperson appointedby the patient, acts as anadvocate for the patient inthe event of incapacity,drawing on the patient’sdocumented preferences.

Systematic review.Rigor: Strong

(comprehensive searchstrategy and rigorousassessment of studyquality).

Two studies (three reports)that involved 337participants investigatedadvance care planning forpeople with ESKD. Neitherof the included studiesreported outcomesrelevant to the review.Study quality was assessedas suboptimal.

The results suggestedpatients were highlysatisfied with the quality ofcommunication andgreater levels of comfort.Discussion regarding ACPand end-of-life care didnot destroy hope or causeunnecessary discomfort oranxiety to patients.

Given the lack of evidencefor advance care planningfor people receivinghemodialysis, large scale,well-designed RCTsinvolving people withESKD are necessary todetermine the efficacy andvalue of advance careplanning for patients.

Aiding patients to establishoptions and priorities forend-of-life care enablespreferences to bemaintained according toindividual wishes. ACPenables patients to preparefor death, strengthenrelationships with lovedones, achieve control overtheir lives, and relieveburden on others.

The most significant barriermay be lack of physicianintervention to initiate andguide advance careplanning discussions.

Luckett (2014), Australia.Review objectives:1) To identify what

interventions have beendeveloped, piloted, andevaluated.

2) To identify whichmeasures have beenused in intervention andother studies.

3) To establish evidencefor the efficacy ofinterventions.

4) To informunderstanding ofbarriers and facilitators

Systematic review: Sampleshad to be adults with aprimary diagnosis of CKDand/or families and thehealth professionals caringfor this group.

Intervention: ACP wasdefined as any formalmeans taken to ensurehealth professionals andfamily members are awareof patients’ wishes for carein the event they becometoo unwell to speak forthemselves.

Systematic integrative reviewusing narrative and meta-analysis. Rigor: Strong(comprehensive search,assessment of study quality,rigorous meta-analysis/synthesis).

Fifty-five papers wereincluded in the review.Seven interventions weretested; all were narrowlyfocused, and none wasevaluated by comparingwishes for end-of-life carewith care received. Oneintervention demonstratedeffects on patient/familyoutcomes in the form ofimproved wellbeing andanxiety following sessionswith a peer mentor.Qualitative studies showedthe importance of

Loss of cognitive capacity iscommon, leaving familiesand nephrologists todecide whether towithdraw dialysis, so thatpatients may continue ondialysis or be resuscitatedor ventilated, when theywould have chosenotherwise. ACP facilitatesan opportunity fordiscussion of treatmentpreferences, ensuringhealth care providers andfamily members are awareof patients’ wishes.

Interventions do not givecomprehensive attentionto patient, caregiver,health professional, andsystem-related factors.Implementation enhancedby educating staff on thebenefits of early ACP andensuring staff have timefor ACP as ‘‘core business.’’Patients tended to wait forhealth professionals toraise ACP and dramaticallyoverestimated their lifeexpectancy; nephrologistsdiscussed EOL issues

Appendix807.e1

Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

to implementation aswell as stakeholders’perceptions of idealapproaches.

instilling patientconfidence that theiradvance directives will beenacted and discussingdecisions about (dis)continuing dialysisseparately from‘‘aggressive’’ life-sustainingtreatments (e.g.,ventilation).

Of eight intervention studiesidentified, four were RCTs,two used a predesign/postdesign, and tworeported postdata withoutcomparison. Rating of biasidentified six of thesestudies as poor quality andtwo as fair.

based on prognosis butstruggled to identify asuitable juncture; nursesfelt uncomfortable raisingACP for fear of upsettingpatients, families, andsenior staff; judging theoptimal timing for theACP was difficult.Surrogates are influencedby their ideas rather thanpatient preferences,leading to poorcongruence concerningEOL care decisions such asdiscontinuation of dialysis.

Noble (2008), UK.Review questions:1) What are theexperiences of thoseopting not to havedialysis to treat renaldisease and their carers?

2) What is the differencein experience andtrajectory to death, forthose withdrawing fromdialysis compared withthose choosing not tostart dialysis?

Systematic review: Studieswere included if they hadbeen peer-reviewed andincluded adultpopulations with aprincipal diagnosis ofESRD.

No specific intervention butadvance care planning wasincluded.

Systematic narrative review:Papers were themed.

Rigor: Weak (quality ofstudies not assessed;analysis is not clearlyreported).

Fifty-one quantitative papersand three qualitativepapers included in thereview.

The presence of an advancedirective requesting thecessation of dialysis, inaddition to the familypreference to stop dialysis,led to most nephrologistssaying that they would stopdialysis. Americannephrologists were morelikely to stop than Germanor Japanese nephrologists.If the family disagreed withthe patient’s advancedirective, nephrologistswere less willing to stopdialysis.

Patients identified hope ascentral to the advance careplanning process in thathope helped them toidentify future goals ofcare and influenced theirwillingness to engage inend-of-life discussions.

Potential barriers to hopewere a reliance on healthcare professionals toinitiate end-of-lifediscussions and the needfor staff to focus on theirdaily clinical work.

Randomized controlled trialsBriggs (2004), USAObjectives: To examinefeasibility of patient-centered advance careplanning (PC-ACP) andreliability of datacollection instruments.

Patients aged 50 yrs or more(and surrogates) attendingheart failure, renal dialysis,and cardiovascular surgeryclinics.

Intervention: One-hour PC-ACP interview, deliveredby an experienced ACPfacilitator; focused onpatient understanding/misconceptions ofcondition and ACP; whatthe patient would want the

Pilot RCT: patient/surrogatepairs. Randomized to PC-ACP (n ¼ 13) or ‘‘usualcare’’ (n ¼ 14)dwrittenmaterial on advancedirectives, with referrals totrained advance careplanning facilitators ifrequested. Predialysispatients offered a class ondialysis treatment choicesand consequences.

The experimental group hadsuperior outcomes incongruence in statementof treatment preferences,decisional conflict, andpatient’s perceived qualityof communication; but notin surrogate’s perceivedquality of communication,patient knowledge of ACP,or surrogate knowledge ofACP.

Deeply involving thesurrogate as the patientclarifies their values andpreferences will enablethem to better representthe patient.

Interviewer skilled infacilitating difficultconversations regardingfuture medical care.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e2

Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives

Population, Setting, andIntervention

Design and MethodologicalRigor (Strong/Moderate/

Weak) Key ResultsHow the Intervention Was

Thought to Work

Contextual FeaturesThought to Influence

Implementation

chosen surrogate tounderstand and act on;summary of choices andneed for futurediscussions.

Rigor: Weak(Randomization notconcealed. Gender andsurrogate (spouse/child)imbalance in the groups.Small sample size).

Kirchhoff (2012); USAObjectives: 1) to identify

patient choices usingpatient-centeredadvance care planning(PC-ACP) comparedwith usual care, and 2)for those patients whodied, to assess whethertheir EOL care was inaccord with theirchoices stated on studyintake.

Patients (and theirsurrogates) with adiagnosis of end-stagecongestive heart failure(CHF) or ESRD; at risk ofserious complication ordeath in the next twoyears; receiving outpatientmedical care, from twocenters in Wisconsin.

Intervention: 1e1.5 hour PC-ACP interview, deliveredby a trained ACP facilitatorto assess the patient andsurrogate understandingof illness; provideinformation abouttreatment options andtheir benefits and burden;assist in documentation oftreatment preferences;help prepare surrogates tomake decisions that honorthe patient’s preferences;ending with thedocumentation of patientpreferences for care usingthe Statement ofTreatment Preferencesform.

RCT: 313 patients with CHF(n ¼ 179) or ESRD(n ¼ 134) and theirsurrogates who wererandomized to receiveeither PC-ACP (n ¼ 160;CHF ¼ 88, r ¼ 70) or usualcare (n ¼ 153; CHF ¼ 89,r ¼ 64).

Rigor: Moderate (clearobjectives, adequaterandomization, notblinded, good follow-up,underpowered).

Among ESKD patients,significantly moreexperimental patients(37.7%) chose to withdrawfrom dialysis than controls(17%)dhowever, this wasnot discussed in theinterview.

Hundred and ten deaths;74% (n ¼ 81) of thepatients made their owndecisions until death andtherefore received care inaccord with their wishes.Patients and surrogateswere usually appreciativeof discussions. They askedwhy no one had raisedthese issues earlier.

When we are doing disease-specific planning, we arereally helping prepareboth the surrogate and thepatient to face futuredecisions in a moreinformed and preparedway.

The facilitators (selectednurses, social workers, andchaplains) were trained indelivery of theintervention and observedduring the study. Theyreceived video-taping,evaluation, and feedback.

Health professionalsreluctant to hold sensitivediscussions. Routinizingthe discussions requires afacilitator and dedicatedtime, checking to see ifpreferences have changed.

Perry (2005); USAObjectives: To explore the

impact of peer mentors(dialysis patients trainedto help other patients)on end-of-life planning.

Patients from 21 dialysiscenters across Michigan.

Intervention: Peers weretrained through anadvance directive (AD)workshop and alsocompleted their own ADs.Eight contacts withpatients over a two- to four-month period: three face-

RCT: 203 participants inthree groups: usual care(n ¼ 81); printed materialon AD (n ¼ 59); peerintervention (n ¼ 63).

Rigor: Moderate(randomization brieflydescribed; nonblindedoutcome assessment).

Peer mentoring significantlyincreased the completionof ADs overall; mostprominently amongAfrican Americans, alsoimproving comfortdiscussing subjectivewellbeing and anxiety.Among white patients,printed material on ADs

Face-to-face relationship withsomeone with similarexperiences engenderstrust, which motivatescompletion of ADs.

African Americans may bemore open to peerinterventions becausetheir culture partakes oforal, rather than writtentraditions, which may bemistrusted.

807.e3

Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

to-face and five by phone.Discussion of chronicillness experiences, values,spirituality, fears, ADissues, and documents.

decreased reportedsuicidal ideation.

Song (2010), USAObjectives: To determinethe feasibility andacceptability of patient-centered advance careplanning (PC-ACP)among AfricanAmericans and theirsurrogate decision-makers.

African Americans over 18with stage 5 CKD receivingeither center hemodialysisor home peritoneal dialysisfor at least three monthsand their surrogates.

Intervention: One-hour PC-ACP interview, deliveredby a trained ACPfacilitator. Addressing: 1)representationalassessment of participants’beliefs about their illness;2) exploration ofmisunderstandings aboutCKD and life-sustainingtreatment, includingdialysis; 3) creation ofconditions for conceptualchange; 4) introduction ofreplacement information;and 5) summarization ofthe discussion.

Pilot RCT: 19 patient/surrogate pairs.Randomized to PC-ACP(n ¼ 11) or ‘‘usual care’’(n ¼ 8)dwritteninformation on advancedirectives provided by anurse or social worker whoencouraged patients tocomplete an AD andaddressed their questionsabout life-sustainingtreatment options.Questions about medicaloptions were referred tophysicians.

Rigor: Moderate (clearlyfocused question,randomization acceptable,outcome data collectorsblinded, very small samplesize).

Significantly greatercongruence and quality ofcommunication in theintervention dyads.

No significant difference inpatient-clinicianinteraction, patient’sdifficulty in makingchoices, surrogate’scomfort in decision-making, or psychospiritualwellbeing of the patientand surrogate.

A shift among participantstoward using all possiblelife-sustaining treatmentsto prolong their lives.

PC-ACP prom tes skillsrelated to in tiatingdiscussions, ssistingindividuals theidentificatio of values andgoals relate to theirhealth care nddeveloping ducationaland organiz tional systemsthat are eff tive inimplementi g anindividual’s lan of care.

African Americans typicallyprefer continuing life-sustaining treatment.

Song (2015); USAObjectives: To test theeffect of an ACPintervention onpreparation for EOLdecision-making fordialysis patients andsurrogates, and forsurrogates’ bereavementoutcomes.

Patients dialyzed for at leastsix months and at risk ofdying in the next12 months, and theirsurrogates, from 20 dialysiscenters in North Carolina.

Intervention: Apsychoeducationalintervention to helppatients clarify EOLpreferences and preparethe surrogate for makingdecisions on the patient’sbehalf, with a trainednurse facilitator at thecenter and a follow-upsession at home two weekslater, with completion of a‘‘goals of care’’ documentto indicate patientpreferences and ADs.Facilitator adherence toprotocol monitored.

RCT: 210 dyads randomizedto intervention (n ¼ 109)and control (n ¼ 101)groups.

Rigor: Strong (concealedrandomization, blindedoutcome assessment,sample size adequate,although small forbereavement measures).

Dyad congruence andsurrogate decision-makingconfidence were better inthe intervention group,but patient decisionalconflict did not differbetween groups.

Forty-five patients diedduring the study.Intervention surrogateshad less anxiety,depression, and post-traumatic distress thancontrols.

ACP can help reparepatients and surrogates fortreatment d cision-makingat the end life byproviding in ividualizedinformation neffectivenes ofmechanical pports at theend of life; y helping thepatient to c rify values;and by activ ly involvingthe surroga to helpprepare for he emotionalburden of d cision-making.

Intervention sessions wereaudited for fidelity andrefresher training offered.

African Americans have beenreported as less amenableto ACP, but the focus ondecision-making ratherthan advance directivescould have made this moreacceptable.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e4

Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

oiainnd, aeaecnp

p

eofdo

ssublaetete

Supplemental Table 1Continued

First Author, Country, andObjectives

Population, Setting, andIntervention

Design and MethodologicalRigor (Strong/Moderate/

Weak) Key ResultsHow the Intervention Was

Thought to Work

Contextual FeaturesThought to Influence

Implementation

Song (2016)*; USAObjectives: To compare

the efficacy of an ACPintervention onpreparation for end-of-life decision-making andpostbereavementoutcomes for AfricanAmericans and whiteson dialysis.

*A secondary analysis ofdata from Song (2015)

Patients dialyzed for at leastsix months and at risk ofdying in the next12 months, and theirsurrogates, from 20 dialysiscenters in North Carolina.

Intervention: apsychoeducationalintervention to helppatients clarify EOLpreferences and preparethe surrogate for makingdecisions on the patient’sbehalf, with a trainednurse facilitator at thecenter and a follow-upsession at home two weekslater, with completion of a‘‘goals of care’’ documentto indicate patientpreferences and ADs.Facilitator adherence toprotocol monitored.

A secondary analysis of datafrom a randomized trialcomparing an ACPintervention (SharingPatient’s IllnessRepresentations toIncrease Trust [SPIRIT])with usual care wasconducted. There were420 participants and 210patient-surrogate dyads(67.4% AfricanAmericans).

Rigor: Moderate (the studywas not powered to test theinteraction of theintervention with race).

Among African Americans(but not whites), theintervention was superiorto usual care in improvingdyad congruence andsurrogate decision-makingconfidence two monthsafter intervention and inreducing surrogates’bereavement depressivesymptoms.

ACP can help preparepatients and surrogates fortreatment decision-makingat the end of life byproviding individualizedinformation oneffectiveness ofmechanical supports at theend of life, by helping thepatient to clarify values,and by actively involvingthe surrogate to helpprepare for the emotionalburden of decision-making.

It may be that theintervention aligns withAfrican Americans’familial, religious, andcommunal frame ofreference; whites in thecontrol group seemed tobenefit simply from beingasked thought-provokingquestions repeatedly.

Non-RCT intervention studiesEneanya (2015), USAObjectives: To test the

shared decision-makingrenal supportive carecommunicationintervention tosystematically elicitpatient and caretakerpreferences for end-of-life care so that careconcordant withpatients’ goals can beprovided.

Hemodialysis patients whoare at high risk of death inthe ensuing six months,attending 16 dialysis unitsassociated with two largeacademic centers.

Intervention: Nephrologistsand social workers willcommunicate prognosisand provide EOL planningin face-to-face encounterswith patients and familiesusing a social workecentered algorithm.Follow-up sessions with thesocial worker will takeplace monthly to providefurther support,education, information,and referral to resourcessuch as hospice.

Protocol for a prospectivecohort study with aretrospective cohortserving as the comparisongroup.

Protocol: No results The interventionsystematically elicitspatient preferences forEOL care, so thatconcordant care can beprovided.

Hayek (2014), USA.Objectives: To examine the

Hundred and fifty-sevenpatients aged >65 yrs,

Design: Test-retest ofnonequivalent

Of the initial pre-intervention sample

By adding AD to the patient’sactive problem list, its

E-mail and visual alertscontribute to information

807.e5

Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

effectiveness of anelectronic medicalrecord (EMR)-basedreminder in improvingAD documentationrates.

CHF, COPD, AIDS,malignancy, cirrhosis,ESRD, or stroke attendinga hospital outpatientclinic.

Intervention: Adding a visualreminder of ‘‘AdvancedDirectives Counseling’’ tothe EMR problem list ofoutpatients with high riskfor morbidity/mortality.

groupsdinitial cross-sectional estimation of thepercentage of patients withdocumented ADs; retestedtwice over the followingyear.

Rigor: Weak (design open toconfounding, fair samplesize. Lack of detail onphysicians’ AD discussionswith patients).

(n¼100) none had ADdocumented. 588 patientswere screened and 157(26.7%) were eligible.Over 6 months, 64 of thesepatients were seen inclinic; 38 had AD on theirproblem list. By the end ofthe study, 29 (76%) chartswith the EMR reminderhad documentation of anAD, compared to 3(11.5%) of those without.Results suggest that EMRreminders may improveAD documentation rates.

importance is highlightedand seen as equivalent asthe patient’s medicalproblems, thereforeencouraging physicians toaddress them.

overload, leadingphysicians often to eitherignore or avoid them. Theintervention in the presentstudy was simple, cost-effective and did notrequire any extra financialor human resources.Combining AD discussionswith the outpatient visitworkflow serves as areminder to the healthcare provider.

Seal (2007), AustraliaObjectives: To explain the

role of patient advocacyin the advance careplanning process.

Nurses on the palliative care,respiratory, renal andcolorectal pilot wards, andthe hem-oncologycoronary care, cardiology,and neurology/geriatriccontrol wards.

Intervention: The RespectingPatient Choices Program(RPCP) to improve use ofAD through a frameworkof educative, patientinformation, safety andquality systems, and policysupport for ACP, plusequipping mainly nurses,through a comprehensivetwo day training course,with skills and resources tofacilitate the process.

A prospectivenonrandomizedcontrolled trial, with focusgroups. 74 in theintervention group; 69 inthe control group.

Rigor: Weak (nonvalidatedquestionnaires;nonrandomized trial withnonequivalent controlgroup).

Statistically significantdifferences in favor of theintervention wards innurses’ self-reportedwillingness, efficacy, andjob satisfaction in relationto upholding patientsinformed choices abouttheir end-of-life treatment,supported by focus groupfindings.

Being able to provideinformation to patientswhile they were still wellenough to talk with theirfamilies stimulatedcommunication, enablingpeople to determine theirfuture care and easingtheir concerns about lossof control. The programprovided a legitimateplatform for nurses toempower patients and toadvocate on their behalf.

Organizational endorsementof ACP with framework ofRPCP encouraged nursesto be patient advocates.

Pressure from medical staffand relatives to treat at allcosts can prevent thepatients’ wishes beingrespected.

Weisbord (2003), USA.Objectives: To identify

symptom burden,health-related quality oflife (HRQoL) andadvance directives inextremely illhemodialysis patients toassess their suitability forpalliative care and toascertain theacceptability of palliativecare to patients andspecialist healthprofessionals.

Nineteen patients whoreceived outpatient dialysisthree times per week in ahospital, for at least threemonths, and had modifiedCharlson comorbidityscores of P8.

Single-group test-retest study.Each patient completedsurveys to assess symptomburden, HRQoL, andprior advance careplanning.

Palliative care specialists thenvisited patients twice andgeneratedrecommendations.Patients again completedthe surveys, and dialysischarts were reviewed toassess nephrologists’compliance withrecommendations andnephrologists’

Patients: mean age 67 yrs,majority male, six of the 19patients died before end ofthe study.

Advance care planning:Twelve patients (63%) hadnever spoken to theirnephrologist about theirwishes for care at the endof life; five (26%) hadnever spoken to theirfamily or friends. Six(32%) had executed aliving will, and of these,four had informed theirnephrologist verbally andthree had presented a copy

Patients have significantsymptom burden andimpaired HRQoL. Only asmall number officiallydocumented their wishesfor their future care.

Nephrologists did notimplementrecommendations relatedto addressing advancedirectives or to pursuingworkup for ongoingmedical problems.

Patients may discuss theseissues with nurses,dieticians, and socialworkers, so incorporatingpalliative care may dependon their input.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e6

Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives

Population, Setting, andIntervention

Design and MethodologicalRigor (Strong/Moderate/

Weak) Key ResultsHow the Intervention Was

Thought to Work

Contextual FeaturesThought to Influence

Implementation

documentation ofsymptoms reported bypatients on the symptomassessment survey.

Rigor: Weak (small, single-group test-retest study).

of the document to theirnephrologist. Seventeenpatients (89%) reportedhaving someone theydesired to make medicaldecisions for them at theend of life; six (32%) hadformally appointed thisperson in writing as asurrogate decision-maker.

Postintervention assessment:no differences wereobserved in symptoms,HRQoL, or number ofpatients making advancedirectives as a result of theintervention.

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

Observational studiesAl-Jahdali (2009), SaudiArabia.

Objectives: To examine thepreferences for CPR andend of life medicalintervention amongSaudi HD patients.

Hundred patients from twohospitals hemodialyzed fortwo yrs or more (meanduration 6.0 years).

Multicenter cross-sectional,observational study.

Data collection:questionnaire completedduring dialysis capturingdemographics, education,employment, family size,knowledge of CPR,mechanical ventilation,and ICU admission; andresponses to scenarios onEOL decisions, medicalinterventions, CPR, ICUadmission, and surrogatedecision-makers.

Rigor: Moderate(questionnaire trialed on asample of patients,appropriate use of statisticsto interpret data).

More than 95% of patientshad little or no knowledgeabout CPR, intubation,and mechanicalventilation; majoritywanted their physician tomake decisions about CPRif they did not have an AD,and only 22% believed thatdecisions should be madeby family members.Overall, physicians shouldadvise patients about theirdisease/prognosis anddiscuss the effects of futuremedical interventions.Informed patients canmake rational choicesabout EOL issues.

Limited patient educationabout CPR and ventilatorysupport.

ACP preferences are rarelydiscussed with patientsand their families.Physicians may lack time,training, or beuncomfortable with EOLdiscussions.

In Saudi society, ADs are notroutinely discussed (notfor religious reasons but)because physicians feardistressing patients.

Anderson (2006), USA.Objectives: To establish the

preferences and directivesof severely ill dialysis

Hundred and nineperitoneal dialysis (PD)patients in an academicnursing home.

Design: Retrospective chartreview of PD patients.Patients were followed upto death or loss to follow-

60.6% of patients werewomen, 54.1% were white,45.0% African American,and 0.9% Asian.

Nonadherence with DNHand ACPR plans may haveoccurred due to decisionsabout hospitalization

The near-completeparticipation with ACP isattributable to thefacility’s explicit

807.e7

Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

patients/surrogates andwhether these change withtime or are influenced bypatient’s functional status;and whether ACPsestablished by patientscorrelate with clinicaloutcomes.

up during a 14-year period(1986 to 2000).

Rigor: Strong (all dataextracted systematicallyand appropriate use ofstatistics).

ACPs were obtained for 108patients. Do not attemptresuscitation (DNAR)status was significantlyassociated with increasedage, lower ADL scores,coronary disease,amputation, anddementia. In compliancewith ACP: patientsindicating do nothospitalize (DNH) spentas much time in hospital asthose planning to accepthospitalization. Seven of46 patients designatedattempt CPR (ACPR) hadCPR attempted. For thesechronically ill patients, ageand functional statusstrongly influence DNARand DNH plans. ACP wasnot conclusive indetermining events duringacute illness.

during times of acuteillness with patients whoretained capacity (or theirsurrogates) who changedtheir minds. CPR notbeing attempted forACPR patients may bebecause CPR is rarelyperformed in nursinghomes.

requirement for ACP, thecommitment of staff inACP participation, andhigh rates of face-to-facediscussions with patientsand family. AfricanAmerican patientsparticipated in ACP asoften as whites due todiscussions and inclusionof family members.

Calvin (2006), USA.Objectives: To develop aninstrument to assessreadiness of patients withchronic kidney disease todiscuss advance care plans.

Questionnaire assessed forcontent validity by a panelof four experts in end-of-life care, and a patientpanel of five personscurrently being treatedwith HD in an inpatientsetting. Thirty-iteminstrument piloted withanother sample of 10patients on HD.

Design: Development of aquestionnaire.

Rigor: Weak (content validityand reliability assessed inthe instrument but onlyusing a small sample.Inadequate reporting ofresults from the pilotquestionnaire).

Content validitydMostpatients rated the items asvery relevant and veryclear, but the professionalpanel ratings were morevaried.

ReliabilitydCronbach’salpha (0.88) for the 30-item questionnaire.Patients also reported theitem was ‘‘helpful and notdifficult’’ and ‘‘gave mesomething to thinkabout.’’

The questionnaire has thepotential to ease theinitiation of ACPdiscussions for health careprovidersdADcompletion may remainlow among patients withkidney failure becausehealth care professionalsare not assessing patientreadiness to discuss ACP.

Collins (2013), Ireland.Objectives: To investigate theviews of HD patients ondeath and dying; beliefsabout what is worse thandeath; likelihood oftalking about death anddying and issues related toend-of-life care and ACP.

Fifty HD (minimum threemonths) adult patientsattending a hospitaldialysis unit.

Design: Cross-sectional,questionnaire (modifiedversion of the EOL surveydeveloped by Life’s EndInstitute) survey with aconvenience sample of alladult patients.

Rigor: Moderate (clearobjectives, adequatelysized conveniencesample).

Most participants werecomfortable discussingdeath, but not alwaysamong family, with mostwanting medicalinterventions to extendlife. The control of painand other symptoms,being physicallycomfortable and being atpeace spiritually were themost important issues atend of life.

Generally, dialysis patientsbelieved that theirchances of survival were asgood as anyone else’s anddid not see themselves ashaving a terminal illness,so issues relating to deathand dying were rarelyconsidered. The majoritywanted medicalintervention to extendlife.

ADs have no legal status inIreland, and very fewpeople are aware of theirexistence.

Health care professionalsdeveloped goodrelationships with patientsallowing the discussion ofsensitive topics andproviding accurateinformation supportedinformed decisions.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e8

Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

Davison (2010), Canada.Objectives: To evaluate EOL

care preferences of CKDpatients, highlight gapsbetween current EOL carepractice and patients’preferences, and to helpprioritize and direct futureinnovation in EOL carepolicy.

Hundred CKD patients(stages 4 and 5) in dialysis,transplantation, orpredialysis clinics in auniversity-based renalprogram.

Design: Survey developmentand implementation(n ¼ 584).

Survey developed fromliterature review aiming toidentify patients’preferences andexpectations of EOL care,such as place of death,symptom treatment, andadvance care planning.Preferences forhypothetical clinical EOLscenarios were alsoincluded.

Rigor: Strong (surveyrigorously reviewed byexperts and piloted in fourdialysis units; largesample).

Patients relied on decision-making by the nephrologystaff for EOL care; painand symptommanagement, ACP, andpsychosocial and spiritualsupport.

ACP: Patients werecomfortable discussingEOL issues with family andnephrology staff, yet lessthan 10% had done so,and only 38.2% hadcompleted an advancedirective

Patients had poor knowledgeof palliative care optionsand of their illnesstrajectory; 84.6% felt wellinformed but only 17.9%felt that their health woulddeteriorate over the nextyear. 61% regretted theirdecision to start dialysis.

EOL care that patients mostwanted was bettereducation and support forstaff, patients, andfamilies, greaterinvolvement of family,more focus on pain andsymptom management,what to expect clinicallynear the end of life,palliative care, and hospiceservices.

CKD patients need to bebetter informed with EOLcare issues; symptommanagement was asignificant priority forpatients but pain in CKDis both underrecognizedand undertreated.

Current EOL clinicalpractices do not meet theneeds of patients withadvanced CKD.

Doctors and nurses are vitalfor providing emotional,social, and spiritualsupport. However, a focuson technology has drawnphysicians away, leavingemotional and spiritualsupport mainly to nursingstaff and otherprofessionals such associal workers andspiritual counselors.

Feely (2016), USA.Objectives: To retrospectively

examine the prevalenceand contents of ADs usedby HD patients.

Review of medical records ofall HD patients (n ¼ 808)in an academic medicalcenter.

Design: Retrospective reviewof medical records of allHD patients over a five-year period (2007e2012).Data collected:Demographics, prevalenceof ADs, and contentanalysis of those ADs withemphasis on managementof EOL interventions

Forty-nine percent had ADs;10.6% mentioned dialysis;and 3% specificallyaddressed dialysismanagement at EOL.Patients who had ADs wereolder. For HD patients whohad ADs, more addressedCPR, mechanicalventilation, artificial

Patients may have had ADsbecause they are moreaware of their reduced lifespan and are encouragedby health professionals,especially, when they havecomorbidities and see arange of professionals.

Patients receiving in-centerHD come into contactwith several health careprofessionals and createmultiple opportunities forthis high-risk group tocomplete ACP.

807.e9

Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

including dialysis.Rigor: Strong (large sample

size, clear objectives, andresults well described anddiscussed).

nutrition and hydration,and pain managementthan dialysis.

Janssen (2013), TheNetherlands

Objectives: To investigate HDpatient preferences forlife-sustaining treatments;and to examine the qualityof patient-physiciancommunication aboutEOL care, includingbarriers and facilitators tothis communication.

Eighty ESRD HD patientsattending dialysisdepartments in sixhospitals.

Design: Cross-sectional,observational studyexamined palliative careneeds using a conveniencesample of ESRD patients.

Validated questionnaires,administered at home orduring dialysis sessions, tomeasure comorbidity,quality of life, willingnessto accept life-sustainingtreatment, and quality of/barriers tocommunication.

Nephrologists completed aquestionnaire on ACP foreach patient to evaluate ifthey discussed prognosis ofsurvival, preferences forCPR and mechanicalventilation, the EOL, andpalliative care with thepatient.

Rigor: Moderate (smallconvenience sample butvalidated questionnairesused).

All patients communicatedtheir choices for life-sustaining treatments andplace of death. Thirtypercent of patientsdiscussed life-sustainingtreatments with theirnephrologist, and qualityof patient-physiciancommunication aboutEOL care was rated poor.Barriers and facilitators toend-of-life carecommunication wereidentified, e.g., barrier’s ‘‘Iwould rather concentrateon staying alive than talkabout death’’; andfacilitators ‘‘my doctorcares about me as aperson.’’

Sixty-five percent of patientsreported that they werenot ready to talk about thecare they want if they getvery sick, making itunlikely that they wouldinitiate ACP discussions.

Patient-physiciancommunication aboutend-of-life care is poor:prognosis, dying, andspirituality areinfrequently discussed.

Most patients stated thatthey wanted CPR but arelikely not aware of lowsuccess rates with CPR.

In this setting, continuity ofcare was lacking andpatients did not knowwhich doctor was going tobe caring for them at theEOL, which hindereddevelopment of ACP.

Rodriguez Jornet (2007),Spain.

Objectives: To investigatedialysis patients’ wishes forhealth care if theircondition becameuntreatable; and toprovide patients theopportunity to create anAD.

Hundred and thirty-fivedialysis patients (aged 18e95 yrs) on dialysis for morethan three months and notexpected to die withinthree months.

Design: Questionnairegathering data ondemographics, medicalstatus, and preference forinterventions in the eventof deterioration.

Rigor: Weak (statistics not agood fit for the design andno details regardingvalidation of thequestionnaire).

Sixty-four patients (47.8%)did not want CPR,respirator, tube feeding, ordialysis in case of coma,persistent vegetative state,severe dementia orterminal illness.Preference fornoninterventionassociated with older age.

ACP occurs with both thephysician-patient and alsowithin the family especiallywith their children asdecision-makers. Reasonsfor not responding to thequestionnaires werepatient reluctance to thinkabout those situations andfear of their answers beingmisinterpreted.

Patients were more likely tocreate an AD if they hadexperience of a familymember previously beingill.

Physicians are somewhatreluctant to talk about theissue of ADs with patients.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e1

0Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

Kataoka-Yahiro (2010), USAObjectives: To ascertain theattitudes to death anddying and completion ofACP among AsianAmerican (AA) and nativeHawaiian (NH) patientsreceiving hemodialysis forstage 4e5 CKD; and todiscover with whom suchpatients prefer to discussACP.

Fifty outpatients from fourdialysis centers; patientswere of AA or NHethnicity; stage 4e5 CKD.

Design: Descriptive, cross-sectional survey design. Aconvenience sample of 50participants aged 30e82 yrs was interviewed face-to-face.

Data collection: EOL survey(43-item questionnaire)along with demographicinformation such asgender, age, ethnicity,marital status, level ofeducation, currentemployment status, andannual household income.

Rigor: Moderate (limitedsample and surveyquestions were notdeveloped from the theoryof planned behaviorframework; therefore,concepts can only bedescribed and exploredand compareddescriptively).

Most participants perceiveddying as an important partof life and werecomfortable talking aboutdeath but expressedconcerns and fears aboutEOL issues.

Aspects of ACP, such asplanning a funeral service,getting finances in order,and completing the willwere important.

Most participants’ attitudesabout ACP were positive,but less than half (40%)had completed ACP.

AAs and NHs would prefer todiscuss ACP with familymembers rather thanhealth or legalprofessionals.

If participants have strongbeliefs about death, dying,and ACP and have apositive evaluation of ACP,they will more likely havea positive attitude towardACP.

The authors commentedthat family involvementand discussion, andcollective decision-making, are importantamong AAs and NHs.

Kurella Tamura (2010), USA.Objectives: To investigatewhether preferences forwithdrawal andengagement in ACP variesby age, race, or ethnicity.

Sixty-one patients (black,Latino, Asian, white) fromtwo dialysis units.

Questionnaire designed toascertain dialysiswithdrawal preferences infive health states.Questionnaire and dialysisunit records used to findout engagement in ACP,completion of advancedirectives, do-not-resuscitate (DNR) or do-not-intubate orders.

Rigor: Weak (no informationon who administeredquestionnaires; smallsample and subgroups).

Fifty-seven percent ofparticipants reported EOLdiscussion with a healthcare provider; 38% hadcompleted an advancedirective, and 10% had aDNR order. Blacks andAsians were more likely toengage in EOLdiscussions, but no blackor Latino participantscompleted DNR orders ascompared with 10% ofAsians and 24% of whites.

Blacks less likely to preferdialysis withdrawalcompared with otherethnic groups; thisdifference not explainedby age, education,comorbidity, and other

The results show that in thisdiverse population, ageand race/ethnicity hadclear patterns ofassociation with treatmentpreferences and withengagement in ACP.

Unexpectedly, youngerpatients and minoritiesdid not have lower rates ofengagement in each ofthe various aspects ofACP; rather the specificpattern of engagementdiffered by age and race/ethnicity groups.

ACP may be less likely to becompleted in cultureswhere patients rely onfamily members toadvocate for their carepreferences.

807.e1

1Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

confounders. Rate ofengagement in EOLdiscussions higher thandocumentation of ACP forall age and most ethnicgroups.

Miura (2006), Japan.Objectives: To assess howaccurately family membersand physicians can predictpatients’ wishes aboutmedical care andtreatment (includingdialysis and CPR) undervarious medical scenarios.

Four hundred fifty dialysispatients in 15 hospitalsand their family membersand physicians.

Design: A questionnairesurvey administeredamong 450 dialysispatients in 15 hospitals,along with their familymembers and physicians

Data collection: Surveydetermined preferencesfor CPR and dialysistherapy under variouscircumstances.

Rigor: Moderate (large,convenience sample;questionnaire was notvalidated, and patients andfamily had opportunity todiscuss answers).

Three hundred ninety-eight(88%) patients/familymembers/physiciansresponded. Sixty-eightpercent of family memberscorrectly predictedpatients’ preferences forCPR. Sixty-seven percentpredicted patients’preferences for dialysis ifthey were severelydemented. Sixty-ninepercent predictedpatients’ preferences fordialysis if they hadterminal cancer.Corresponding figures forphysicians were 60%, 68%,and 66%. However, Kappacoefficient analysisindicated that neitherfamily members norphysicians more accuratelypredicted their patients’wishes about life-sustainingtreatments than expectedby chance alone. (Allkappa coefficients <0.4.)

Patients with specific EOLpreferences should leavebetter advance directives.

Only half of patientsexpected their familymembers to correctlypredict their wishes forCPR or continuation ofdialysis therapy. Theseresults suggest that it isimpossible to knowpatient preferenceswithout propercommunication.

Japanese culture includesthe belief that one’swishes are intuitivelyknown to others;physicians may bepaternalistic; and familiesmay see withdrawal oftreatment as abandoningtheir duty to the patient.

Nishimura (2007), USA.Objectives: To determinehealth care preferencesexpressed by patients inADs and to identifycharacteristics of patientswho completed them.

Four hundred seventy-sixparticipants who wereinpatients or outpatientsof a hospital, with a rangeof conditions (only 32 withrenal insufficiency), whosubmitted an AD.

Design: A computer-generated random sample(of >25,000 patients) wasselected. Data wereextracted from thehospital’s electronicpatient information systemon AD characteristics (e.g.type of AD and number ofpages), and contents (e.g.,designation of health careagent, decision-makingpowers given to the agent,and specific health carerequests) were recorded.

Median age of patients whenthey submitted an AD was67 yrs.

Of the 476 patients, 434 hada designated health careagent, and 340 hadgranted the agent powersto consent for procedures,to access information andto withhold and withdrawlife-sustaining treatments.Three hundred thirty-ninehad submitted a combinedAD (incorporating a livingwill and a power ofattorney).

Knowledge of what types ofinformation patientsinclude (and do notinclude) would highlightthe EOL concerns thatare important to patientsand possibly point outinadequacies in thecurrent use of ADs.

The most effective ADshould allow a patient toname a health care agent,to state their preferences,is detailed yet easy to use,and is disease specific.

A combined AD seemedmore convenient for thepatients; long ADs coulddiscourage patients; avariety of form typesshould be offered as someprefer writing moredetailed comments andothers not.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e1

2Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

Rigor: Strong (largerandom sample and ADsassessed by at least tworeviewers).

Most patients expressed adesire for pain control(308). For the clinicalsituations of dying orpermanentunconsciousness, mostpatients (371) expressedthe preference to avoidgeneral life-support butdid not explicitly addresscommon life-sustainingtreatments including CPR,HD, mechanicalventilation, artificialnutrition and hydration,blood transfusion.

Sanchez-Tomero (2011),Spain.

Objectives: To examinepatient familiarity with anadvance directivedocument (ADD) and theproportion with an ADDor a legal representative;acceptance of AD indialysis units, and patients’opinions regarding theADD and their attitudes toEOL care.

Hundred and fifty-fourdialysis patients in sevenhospital or peripheraldialysis units.

Patients receivedinformation explainingAD and completed aquestionnaire designed todiscover their opinions ofthe ADD.

Rigor: Moderate (unclear ifquestionnaires werevalidated; however, largesample size, number ofsites, and appropriatestatistical analyses).

Questionnaires weredistributed to 265 patientsand 154 (58.11%)responded.

Only 7.9% of patients had anADD, and 6.6% hadformally appointed arepresentative, but 60.9%had empowered anotherperson to represent themand their wishes, shouldthey lose capacity. Over65% did not wantmechanical ventilation,tube feeding, continueddialysis, or resuscitation ifthey had cardiac arrest.Most patients were clear asto their wishes for futurecare, but stating these in awritten document was rare.Fifty-four percent said theywould like to create anADD, whereas 43.7% saidthey would not; 51.6% saidthat an ADD should becreated before dialysistreatment begins, whereas65% said that this option

The patients in this studyhad clear wishesregarding their EOL care,despite the lack of ADdocumentation. However,the authors commentedon the importance ofhaving these wisheswritten in an official ADD,which leads to a greaterlikelihood that they arerespected.

In Spain, familiarity withADDs is very limited, andpatients may be suspiciousthat making an ADD willlead to doctorswithholding necessarytreatments. Hospitalprotocols would helpfacilitate ADDs.

807.e1

3Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

should only be given topeople if they ask for it.

Schmidt (2015), USA.Objectives: To review theeffect of advance careplanning on withdrawalfrom dialysis, use ofhospice, and location ofdeath.

Sixty-five former dialysispatients who died over fiveyears, who attended a ruraloutpatient dialysis unit.

Retrospective review ofdialysis and medicalrecords.

Rigor: Weak (small sample,single center, little detailon data collectionmethods).

Sixty-five deaths occurredduring the five-year reviewperiod. Of these, 46 (71%)occurred out of hospital,and 33 (51%) followedwithdrawal from dialysis.

Advance care planning wasassociated with a lowincidence of death inhospital and, among thosewho withdrew fromdialysis, high use ofhospice.

Patients were routinelyasked if they had an ADand completed a ‘‘scopeof treatment’’ form.Physicians discussedwithdrawal of dialysis withthose considered likely todie within six months.

Hospice care was activelyrecommended byphysicians.

A reluctance of many ruralpatients to travel thenecessary distance fortertiary care clearlycontributes to the lowprevalence of in-hospitaldeath, and all patientswho died in hospital didso after an acute eventprompted hospitalization.

Song (2013), USA.Objectives: To describehow chronically seriouslyill patients can be assistedto identify a surrogate whowould act as their decision-maker if their conditionsuddenly deteriorates andwhether these surrogateswere the same as theirdocumented emergencycontacts.

Ninety-four AfricanAmerican or Caucasianpatients attendingoutpatient dialysis centersin North Carolina for atleast six months.

A descriptive cross-sectionalstudy using eligibilityassessment and baselinedata from an ongoingefficacy trial of an EOLcommunicationintervention.

Rigor: Moderate (smallconvenience sample).

Three of the 94 patientsalready had a surrogatedocumented in an AD.The remaining 91 came tothat decision after workingthrough guiding questionswith the recruiter.

Thirty-two patients (34%)named a spouse, 29(30.9%) named a child, 17(18.1%) named a sibling,five (5.3%) named aparent, three (3.2%)named a friend, and eight(8.5%) named anotherperson. Of the 94surrogates, 60 (63.8%)were listed as emergencycontact; 34 (36.2%) werenot.

No differences insociodemographiccharacteristics betweenthose who appointed adifferent person for theirsurrogate versus those whoappointed the emergencycontact as a surrogate.

Without a named surrogatedecision-maker, peopleacting as emergencycontacts may findthemselves becoming thedecision-maker, despiteneither the patient northe emergency contactagreeing to this.

Some states demandnotarization for anadvance directive, and thismay cause a reliance onemergency contacts fordecision-making.

Yee (2011), Singapore.Objectives: To explore theknowledge, attitudes, andexperience of renal healthcare professionals inSingapore with ACP forpatients with ESRD.

Five hundred sixty-two renalphysicians, renal nurses,renal medical socialworkers (MSWs), andother allied healthprofessionals working ingovernment hospitals,private sector, and

Cross-sectional survey: 41-item questionnaire withfour sections:demographics ofrespondents, knowledgeof, attitudes to, andexperience of ACP.

Survey ‘‘distributed’’ to

Six hundred twenty surveysdistributed; 562 returned(90.6% response rate).Hundred percent of MSWsconsidered ACPdiscussions part of theirrole, compared with 82.4%of doctors, 37.1% of

Physicians fear upsettingfamily, report lack of time,and perceive thatSingaporeans may not beready for ACP. In Asiansocieties, the family playsan important role inhealth-related decision

(Continued)

Vol.

56No.

5Novem

ber2018

807.e1

4Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

voluntary groups offeringsubsidized dialysis.

health care professionalsworking in renal units.Authors do not state howthis was done, for example,by post or other means.

Rigor: Strong (good samplesize, high response rate,statistics were appropriateand well explained).

nurses, and 38.1% of otherhealth professionals.MSWs and doctors hadhigher knowledge scoresthan the other groups.Nurses were leastconfident in discussingACP and most afraid ofdistressing patients andfamilies, whereas MSWswere most confident. Fordoctors, the main barriersto discussing ACP wereshortage of time, concernsabout family reaction, andtheir belief that patientswere not ready to discussACP.

and may sometimesconflict with patient’spreferences. ACP can beconfused with AD (andAD with euthanasia) byprofessionals andpatients. Nurses havemost contact time withpatients but most see ACPas part of the physician orMSW role.

Qualitative studiesBristowe (2014), UKObjectives: To develop andpilot a REnal-specificAdvancedCommunicationTraining (REACT)program to address theneeds of ESKD patientsand renal professionals.

Patients with ESKD andprofessionals from twolarge renal units inLondon teachinghospitals.

Two-part study: 1)development of REACT byone multiprofessionalfocus group and patientexperience survey (n ¼171) and 2) piloting of theprogram with nine nurses/health care assistants andseven renal consultants.

Rigor: Weak (genericsatisfaction survey; smallsample for evaluation oftraining).

Patient survey: 71% feltconfident to ask questions;65% felt treatment optionshad been clearlyexplained; but only 50%felt that their concernswere taken seriously; 48%felt that they had beensufficiently involved intreatment decisions. Focusgroup identified need forbetter information aboutEOL and patient/carerperspective; skills andpractice for difficultdiscussions; opportunity todebrief.

REACT was foundacceptable, but nosignificant increase in staffconfidence.

Inadequacies in EOL carefor renal patients must beaddressed, but renalprofessionals needtailored training andsupport to optimize theirskills and addressconcerns in managingACP discussions.

Challenges remain inprioritizingcommunication trainingin busy renal units andidentifying funding toprovide high-qualitytraining with adequatefollow-up.

Bristowe (2015), UKObjectives: To investigateexperience of patientsstarting HD, impact onquality of life, and

Patients with ESKD from twoNHS hospitals.

Qualitative study;semistructured interviewswith 20 patients, from twoNHS hospitals, purposivelysampled for age, time on

Themes: Looking back,emotions of commencinghemodialysis; currentexperiences, illness, andtreatment burdens; and

Given a changingpopulation, there is aneed for a culture shiftfrom a ‘‘disease-focused’’model toward a ‘‘holistic

‘‘Conveyor belt’’ culture,with prioritization of‘‘getting you on and off’’rather than caring for theindividual. This

807.e1

5Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

preferences for futureand EOL care to informunderstanding andtiming of ACP.

dialysis, and symptomburden.

Thematic analysis.Rigor: Strong (careful data

collection, analysis, andfindings well described).

looking Ahead, facing therealities.

Challenges: Gettinginformation,communicating with staffand ‘‘conveyor belt’’culture of hemodialysisunits, lack of opportunityto discuss their future ifhealth deteriorated, andvariable involvement intreatment decisions.Discussion of issues wasmore acceptable to somethan others.

care-based’’ approach,normalizing discussionsabout preferences,priorities, and future care.

experience wascompounded for thosereliant on hospitaltransport, for which therewas often a lengthy wait,with difficult recovery thenext day.

Staff avoid discussion offuture due to fear ofcausing distress.

Calvin (2004); USAObjectives: To explore

what happens ashemodialysis patientsmake decisions aboutEOL treatment and todevelop a substantivetheory to explain thisphenomenon.

Hemodialysis patientsattending three outpatientcenters in central Texas.

Grounded theory;semistructured interviewswith 20 participants.

Rigor: Moderate (insufficientdata presented to supportthe findings).

Patients chose to focus onliving rather than dying. Asubstantive theory of‘‘personal preservation’’was developed. Thisconsists of three phases:knowing the odds forsurvival, definingindividuality (beating theodds, discoveringmeaning, being optimistic,and having faith in ahigher force), andpersonal preservation(being responsible andtaking chances).

When health care providersdo not know whathemodialysis patientsprefer with regard to life-sustaining treatment, theyfeel obligated to doeverything possible topreserve patients’ lives.

Patients protectedthemselves by trustingtheir family members withEOL decisions anddistancing themselvesfrom death.

Some patients saw advancedirectives as, primarily, forothers’ legal protection,as opposed to their ownautonomy.

Calvin (2014), USAObjectives: To explore

hemodialysis patients’family members’understanding of EOLdecision-makingprocesses. Aimed toaddress 1) familymembers’ constructionsof ACP and 2) familymembers’ perceptionsof health care providers’roles and responsibilitiesin ACP.

Family members of personsattending hemodialysisoutpatient facilities inHouston.

A qualitative descriptivedesign, with a Glaserianapproach to the analysis.

Eighteen family memberswere recruited primarilyfrom outpatient dialysisfacilities and interviewedindividually.

Rigor: Strong (design andrecruitment appropriate;analysis and findingsclear).

Family members feel a call toduty to protect patients by1) sharing burdens, 2)normalizing life, and 3)personalizing care.

Participants and patients livewith burdensomeuncertainties. Participantsspoke of health careproviders’ general lack ofcare and struggled toarticulate health careproviders’ responsibilitiesin ACP because they hadnot experienced ACP.

Participants avoidedthoughts of death andnormalized life to protectpatients. Personalizingcare involved beingpresent during encounterswith health care providers

To honor patients’ valuesand medical treatmentpreferences, health careproviders need tocommunicate with familymembers because of thepatient’s loss of decision-making capacity.

Lack of trust in health careproviders. Patients, theirfamilies, and health careprofessionals are reluctantto discuss EOL treatmentpreferences or formulateadvance directives.

(Continued)

Vol.

56No.

5Novem

ber2018

807.e1

6Advan

ceCare

Plan

ningin

Kidn

eyFailu

re

Supplemental Table 1Continued

First Author, Country, andObjectives Population and Setting

Design, Methods of DataCollection, and

Methodological Rigor(Strong/Moderate/Weak) Key Results Explanation of the Results

Contextual FeaturesThought to Influence

Implementation

and respecting wishes.Davison & Simpson

(2006), CanadaObjective: To understand

hope in the context ofACP from theperspective of patientswith ESRD.

Patients with ESRD selectedfrom the renalinsufficiency, hemodialysis,and peritoneal dialysisclinics at a Canadianuniversityeaffiliated renalprogram.

Twenty-three individualinterviews with 19 patients(14 expected to requiredialysis within the next12 months).

Rigor: Strong (design andrecruitment strategyappropriate, data analysis,and findings clear).

Participants’ hopes wereshaped by personal values.They perceived ACP as ameans of enhancing hopeby providing informationearly in the illness thatfocuses on the impact ondaily life, empoweringpatients, and enhancingrelationships with staff andloved ones. These were keyfactors in sustainingpatients’ ability to hopeand helped them toimagine possibilities for afuture that were consistentwith their values andhopes.

Comprehensive care ofpatients with ESKDrequires skill in ACP to layout a set of values andprocesses for approachingEOL decisions andidentify preferences forfuture goals of care.

Participants were willing todiscuss EOL but expectedprofessionals to initiateand guide them throughthe process of ACP. Theydid not feel they shouldhave to ask for prognosticinformation.

Reliance on healthprofessionals to initiateEOL discussions and thedaily focus of clinical carewere perceived to bepotential barriers to hope.

Davison (2006b), CanadaObjectives: To obtain views

of patients with ESRDon the salient elementsof ACP discussions andto discover theirpreferences regardinghow ACP could befacilitated by the healthcare team.

Patients with ESRD fromrenal insufficiency,hemodialysis, andperitoneal dialysis clinics;Northern Alberta RenalProgram at the Universityof Alberta.

Ethnographic studyconsisting of personalinterviews with 24 patients.Participants purposivelyselected on theirwillingness to discuss issuesand were stratified by age,gender, and dialysismodality.

Rigor: Strong (researchdesign and recruitmentstrategy appropriate;findings and analysisclear).

Patients’ views on benefit ofACP varied significantlyand reflected beliefs, fears,and wishes for EOL care.

Patients wished for moreinformation earlier intheir illness, andinformation giving wasviewed as a critical elementof the ACP process.

Patients also requiredinformation about theimpact of interventions ondaily life.

Patients requested thathealth care professionalsshould initiate ACP, use laylanguage and beempathetic, and affirmself-worth in patients.

Argument for shifting thefocus away from thetraditional informationgiving and documentcompletion model ofadvance directives andreplacing it with arelational, patient-centered process thatfocuses on broader goalsof care.

Tension between patientpriorities for ACP(focused on how careplans affect them andtheir loved ones in dailylife) rather than that ofclinical practitioners(statistical effectiveness oftreatments).

Physicians clearly seen bypatients as havingresponsibility forinitiating and guidingACP. No standards of careregarding when to initiateor how to conduct ACPdiscussions.

Goff (2015), USAObjectives: To inform the

development of anintervention focused onimproving ACP. To elicitrecommendations forhow dialysis teams

Patients and family membersattending two dialysis unitsin Massachusetts and NewMexico.

Applied grounded theoryapproach. In-depthinterviews with 13 patientsand nine family/friends;interviews were audio-taped, professionallytranscribed, and analyzed

Most patients and family/friends felt optionsavailable for EOL careshould be discussedthrough ACP, but fewpatients or family/friendsreported discussing

Patients’ life experiencesinfluenced their dialysisexperiences and theirviews of ACP, for example,the suffering theyobserved in the dialysisunit affected their views.

Negative dialysisexperiences, includinglack of a personalconnection to thenephrologist and dialysisteam, dissatisfaction withcommunication, and

807.e1

7Vol.

56No.

5Novem

ber2018

O’H

alloranetal.

should discuss ACP. in an iterative process.Rigor: Strong (research

design and recruitmentstrategy appropriate. Dataanalysis and findings clear)

prognosis, life goals, oroptions for care withhealth care professionals.

Several patients and family/friends had discussed ACPwith the primary carephysician, but this was notshared with thenephrologist.

Patients less critical of theirdialysis experience hadhigher education/incomelevels re older.

disenfranchisement maydiminish trust in thedialysis team.

Patients and family/friendswith lower socioeconomic

Hines (2001), USAObjectives: To determinewhether differences invalues and preferencesfor the ACP process maybe reasons familymembers areinadequately informedto act as surrogates.

Hemodialysis patients fromtwo regionsdNew Yorkand West Virginia, USA.

Face-to-face and telephonesurveys using structuredquestionnaires.

Participants: 242 dialysispatients and theirdesignated surrogates

Rigor: Moderate (rationalefor how interviewquestions were decidedupon not clear).

Ninety percent of patientsdesignated a familymember as their surrogate.Generally, conversationsabout EOL issues did notincrease surrogateknowledge of patients’values or preferences.

Surrogates wanted writtenand oral instructions moreoften than patients wantedto provide them.

Compared with patients,surrogates were less likelyto want to prolong thepatient’s life if it entailedsuffering and were moreconcerned about beingcertain before stoppinglife-sustaining treatments.

DespitemembmakesurrogknowlvaluesPatienthe detheirexerc

Abbreviations: RCT ¼ randomized controlled trial; ACP ¼ advance care planning; ESKD ¼ end-stage kidney disease; CKD ¼ chronic kidney disease; EOL ¼ encentered advance care planning; CHF ¼ congestive heart failure; AD ¼ advance directive; EMR ¼ electronic medical record; HD ¼ hemodialysis; CPR ¼ ca

Vol.

56No.

or we

and/or minority statusfelt disempowered anddiscouraged from ACP.

Some staff minimized theseriousness of dialysis, andstaff could do more toreduce feelings ofhumiliation anddependency that patientsexperience as dialysispatients.

reliance on familyer surrogates toEOL decisions,ates often lackedge of the patient’sand preferences.ts may overestimategree of autonomysurrogates want toise.

Physicians may assume thatpatients who reporthaving talked with theirfamily will haveadequately preparedthem for surrogatedecision-making.

d of life; ESRD ¼ end-stage renal disease; PC-ACP ¼ patient-rdiopulmonary resuscitation.

5Novem

ber2018

807.e1

8Advan

ceCare

Plan

ningin

Kidn

eyFailu

re