Competing interests: The authors have no competing interests to declare.
Authors' contributions: RGF, LT, RB, RM, SC and IKR are responsible for writing the manuscript. All authors have read and approved the final manuscript.
Professor Robert G Fassett, Renal Research, Royal Brisbane and Women's Hospital, Brisbane, QLD 4029, Australia. Email: email@example.com
Patients with end-stage kidney disease have significantly increased morbidity and mortality. While greater attention has been focused on advanced care planning, end-of-life decisions, conservative therapy and withdrawal from dialysis these must be supported by adequate palliative care incorporating symptom control. With the increase in the elderly, with their inherent comorbidities, accepted onto dialysis, patients, their nephrologists, families and multidisciplinary teams, are often faced with end-of-life decisions and the provision of palliative care. While dialysis may offer a better quality and quantity of life compared with conservative management, this may not always be the case; hence the patient is entitled to be well-informed of all options and potential outcomes before embarking on such therapy. They should be assured of adequate symptom control and palliative care whichever option is selected. No randomized controlled trials have been conducted in this area and only a small number of observational studies provide guidance; thus predicting which patients will have poor outcomes is problematic. Those undertaking dialysis may benefit from being fully aware of their choices between active and conservative treatment should their functional status seriously deteriorate and this should be shared with caregivers. This clarifies treatment pathways and reduces the ambiguity surrounding decision making. If conservative therapy or withdrawal from dialysis is chosen, each should be supported by palliative care. The objective of this review is to summarize published studies and evidence-based guidelines, core curricula, position statements, standards and tools in palliative care in end-stage kidney disease.
The role of palliative care in end-stage kidney disease (ESKD) is well developed in the UK, USA, Italy and Canada.1–9 Palliative care in ESKD is important in the contexts of conservative therapy (choosing a non-dialysis pathway), withdrawal of therapy and in symptom control. Advanced care directives and end-of-life decisions overarch these pathways. There is a recognized need for education regarding provision of palliative care in dialysis patients.10 However, there is no clear pathway to palliative care,11 considerable variation in the provision of palliative care services for ESKD patients12 and little evidence upon which to develop standards of renal palliative care in ESKD.13 There has been an increase in the elderly accepted onto dialysis in Australia. In 2004, 244 (445 per million population) new patients were accepted on dialysis in the 75–79 year age group. This increased to 277 (504 per million) in 2008. In the 80–84 year age group 103 (267 per million) started dialysis in 2004, which increased to 187 (442 per million) in 2008 and in the >85 year group 32 (107 per million) started dialysis in 2004, which increased to 58 (159 per million) in 2008.14 Despite this, the Caring for Australasians with Renal Impairment (CARI) Guidelines do not address palliative care.15 In addition, many elderly assessed for dialysis either do not progress16 or die before they would have required dialysis therapy.17
We will review the existing literature on palliative care provision in ESKD in the contexts of conservative therapy and withdrawal from dialysis. The available observational, retrospective and case studies are summarized in Table 1. There are no reported randomized controlled trials.
Table 1. Studies investigating palliative care in kidney disease
Study population (n = subject numbers)
Duration of follow-up
DOPPS, Dialysis Outcomes and Practice Patterns Study; N/A, not available; NHS, National Health Service.
End-of-life planning should be an expected part of care supported by provision of palliative care
PALLIATIVE CARE SUPPORTING CONSERVATIVE CARE IN ESKD
The literature reporting on withdrawal of dialysis extends back many years and has been the focus of palliative care in ESKD until recently.34 However, the emphasis on making a choice between conservative (non-dialysis therapy) as an alternative to active (dialysis) treatment pathway before the need to start dialysis is gaining importance with some recent studies reporting comparable outcomes between these pathways in the elderly with multiple comorbidities.18,30 These studies may enable renal multidisciplinary teams to provide evidence-based advice to patients before committing to ESKD therapies.22,30 There is increased recognition in critical care medicine that a holistic approach is required to support end-of-life decisions,35 and in renal medicine the role of palliative care is also gaining importance.11,13 The interrelationships of these issues are summarized in Figure 1.
Pre-dialysis education is considered an essential part of the preparation for ESKD management36–39 as it acts to inform the choices made by patients and their carers and enhances shared care planning with multidisciplinary teams.5 Patients and their families may be unwilling or unable to choose not to commence treatment or to withdraw from it40 and therefore information about palliative care options is an important inclusion in pre-dialysis education. Hence, in addition to discussing dialysis modality options and transplantation, discussion of a conservative approach supported by palliative care should be offered to those particularly of advanced age and/or with multiple comorbidities. Although some observational and retrospective studies have been published18,19 and are summarized in Table 1, there are limited studies available upon which to base such discussions.
The issue of conservative therapy was addressed in an observational cohort study where patients approaching dialysis who had undertaken a multidisciplinary assessment were recruited over 54 months.18 Investigators looked for features that influenced clinicians to advise a conservative approach rather than starting dialysis. The patients were followed for 3–57 months on the basis of the therapy option selected, dialysis or palliative care. Of 321 patients recruited, 258 were recommended for renal replacement therapy and 63 for palliative care. The patients that were recommended to take a palliative care pathway had greater functional impairment, were older and more often diabetic. Of the 63 patients, 34 recommended for palliative care died, 26 of these from kidney failure. Ten patients recommended for palliative care actually chose dialysis but had a median survival of only 8.3 months. This was not significantly longer than those that actually chose the palliative care pathway. In this group of patients the decision to accept either dialysis or palliative care had no significant effect on survival.
A retrospective study of 129 stage 5 CKD patients over 75 years of age who attended pre-dialysis multidisciplinary clinics assessed patient survival defined as time from reaching an eGFR of <15 mL/min until death or the end time point of the study.19 There were 52 patients in the dialysis group and 77 in the conservative treatment group. The survival of the dialysis group was significantly greater than that of the conservative treatment group both at 1 and 2 years. However, when adjusted for comorbidities, particularly ischaemic heart disease, there was no such advantage seen.
Survival, scored using the validated Stoke comorbidity grade, was assessed in a prospective observational study of patients, managed through a multidisciplinary team, who chose not to undertake dialysis.20 Seventy-three patients were recruited with a median age of 79 years. The median survival was 1.95 years and 1 year survival was 65%. The Stoke comorbidity grade independently predicted survival. Based on these results the authors advocated pre-dialysis multidisciplinary care supporting conservative therapy particularly for elderly patients with comorbidities. The Stoke comorbidity grade may provide prognostic information for predicting survival that will help multidisciplinary teams counsel ESKD patients approaching dialysis.
Nursing home patients
To be able to offer accurate advice to nursing home patients of advanced age and/or multiple comorbidities, it is necessary to know how outcomes compare between conservative therapy and dialysis treatment. A recent study attempted to address this issue, The US Renal Data System, and was used to identify residents of nursing homes that started dialysis over a 2 year 4 month period. The outcomes for residents of nursing homes in the USA were poor with a mortality rate of 58% in the first year and 29% having decreased functional status. Pre-dialysis functional status was only maintained in 13%.30 This highlights the importance of offering palliative care with its associated focus on symptom control.41 In an associated editorial the paucity of data in this area was noted. Increased comorbidity can predict death in dialysis patients.42 However, unless there are data comparing quality and quantity of life in ESKD therapy compared with conservative management we struggle to identify those that would most likely benefit from such therapy. More studies are required to particularly enable us to define which patients will benefit from conservative rather than dialysis therapy.41 In addition, it is important to adequately inform patients of potential outcomes to assist them with their decisions.
The increasing acceptance of the elderly onto dialysis programmes has heightened the interest in and study of the process of end-of-life decision making, supported by palliative care, in ESKD.43 This is particularly relevant as the morbidity and mortality seen in ESKD in its latter stages is very high.44 Mortality in ESKD is mainly a consequence of cardiovascular disease, which may be 10- to 100-fold greater than age- and gender-matched controls in the general population,45,46 or may be due to a higher prevalence of other causes such as pneumonia.47,48 However, one study in dialysis patients found older dialysis patients had a lower excess mortality in the first 3 years of therapy than younger patients.49 This can make individual survival and quality-of-life predictions difficult in the elderly. Despite this, the overall mortality is high and the assessment of the benefit of dialysis in the elderly is difficult. Available studies do suggest dialysis is still life extending in the elderly.19,50 However, in the retrospective study by Murtagh et al. the survival advantage conferred by dialysis was abrogated by comorbidities such as ischaemic heart disease.19 In a small prospective randomized controlled trial in those over 70 years a low protein diet delayed dialysis and was associated with an equivalent mortality when compared with those who started dialysis.51,52 Factors identified as indicators associated with not opting for dialysis among octogenarians included social isolation comorbidities such as diabetes, late referral and Karnofsky score.50 In those selecting dialysis therapy, dependent predictors of death included poor nutritional status, late referral and functional dependence.50 Octogenarians also have been shown to lose independence after dialysis initiation.53 The quality-of-life benefits of dialysis therapy in the elderly remain unclear.18 In a small observational study in ESKD patients over 75 years of age conservative therapy was associated with a quality of life similar to haemodialysis.8
PALLIATIVE CARE AFTER DIALYSIS WITHDRAWAL
Withdrawal from dialysis is one of the commonest causes of death and represents 35% of dialysis deaths in Australia.54 The Dialysis Outcomes and Practice Patterns Study, reported differences in withdrawal from dialysis between and within countries and that this was correlated with nephrologists' opinions on these issues.31 The mortality rate among dialysis patients is very high and may be greater than in HIV and some cancers. In addition, their symptom burden and rate of hospitalization are very high.55 As more elderly patients are being accepted onto dialysis the focus of care needs to shift from the life extension aspects of dialysis care to relief of symptom burden and palliative care. Withdrawal from dialysis is a generally accepted process34 and provided it is supported by adequate palliative care, the subsequent death can be good.56 In the USA, end-of-life support for renal patients is well developed with a specific website that includes pain management guidelines.3 In a study of 131 patients who withdrew from dialysis, 79 were followed prospectively until they died.33 These patients had multiple comorbidities and their main symptoms in the last day of their life were agitation and pain. This study recommended mandatory end-of-life planning in ESKD management incorporating palliative care provision.
There is a documented underutilization of hospice facilities in ESKD patients in the USA where only 14% of all ESKD deaths occurred in patients using these facilities.32 Only 40% of ESKD deaths from withdrawal of dialysis entered a hospice for care. This study also demonstrated a cost saving associated with dialysis patients dying in a hospice after withdrawal from therapy. ESKD patients use a hospice at a rate of 25% compared with that seen in cancer patients.55
A pilot study reviewed the charts of 35 dialysis patients that withdrew from therapy and were followed by a palliative care team.23 The mean survival time from dialysis withdrawal to death was 10 days. Symptoms were reduced in the last day with palliative care input. The study suggested improved education of multidisciplinary nephrology staff was required.
A small Australian study assessed the abatement of medical treatment in ESKD that encompassed both withdrawal and non-initiation of dialysis treatment.11 This study included four patients that withdrew from dialysis, seven that did not initiate dialysis and five spouses of these patients. The participants undertook semistructured interviews from which the investigators gleaned there would be benefits from a greater discussion of end-of-life issues with acceptance of this as part of standard practice. These findings are supported by a study into the experience of patients after cessation of dialysis that found early palliative care referral could assist the patient and multidisciplinary team to manage areas such as pain and create opportunities to discuss palliative care options.23
Factors identified as indicators associated with dialysis withdrawal include poor functional status, functional dependency, gender, ethnicity, social isolation and comorbidities.24,34,57 Recently, Kurella Tamura et al. explored dialysis withdrawal preferences and found these varied with race, with blacks less likely to withdraw from dialysis than whites.58 Also they found the elderly did not have an increased preference for dialysis withdrawal whereas younger patients were less likely to record their preferences and be open to end-of-life discussion.58
SYMPTOM ASSESSMENT AND TREATMENT IN PALLIATIVE CARE IN ESKD
Symptom control is of paramount importance in ESKD patients on dialysis with pain being the most common.59 The use of the World Health Organization three-step analgesic ladder is effective in pain management in haemodialysis patients.59 A prospective cross-sectional pilot study compared symptom burden and quality of life between patients with advanced ESKD with an eGFR <17 mL/min and a contemporary cohort with terminal malignancy.29 Those patients with ESKD had similar symptom burden and reduced quality of life as the terminal malignancy group. This highlights that the palliative care needs of patients with ESKD are just as important as those with terminal cancer.
In a retrospective chart review of conservatively managed stage 4–5 CKD patients Murphy et al. assessed symptom burden using a short patient-completed assessment tool.26 Patients all attended a renal palliative care service over a 10 month period. Comorbidity data were collected and a modified patient symptom module was completed. Fifty-five patients who were managed without dialysis were reviewed and the symptom burden recorded was high. Using a tool that may lead to assessing more effective symptom treatments, revealed the extent of symptom burden in conservatively managed ESKD. It is also important to emphasize that a conservative, non-dialysis approach to ESKD management should not be a vacuum, but in fact can provide an intensive programme of multidisciplinary care and support. It also provides the patient and their family with the confidence that there will be no reduction in medical and nursing care.60
A study from Hong Kong assessed and compared the quality of life and symptom burden between patients on haemodialysis and peritoneal dialysis with palliative care ESKD patients with an eGFR <15 mL/min.22 This prospective observational study included 179 patients, 134 who had dialysis and 45 who undertook palliative care. Those that received palliative care had greater comorbidity and were older. There was no significant difference in symptom burden between groups and the quality of life was significantly reduced in both groups. In this setting there was little difference in symptoms and quality of life whether they had dialysis or palliative care.
EMOTIONAL FACTORS IN PALLIATIVE CARE
The palliative care process needs to consider acknowledging and dealing with this grieving both in the patient, their family and health-care providers. A study conducted by Badger exploring factors impacting on end-of-life transitions in critical care found two key areas of concern for nurses.61 These were the ‘complex emotions and frank indecisiveness expressed by patients’ families. Grief and loss are issues intertwined throughout the course of CKD and ESKD management.62 Although grief is clearly associated with death, it is also evident and experienced much earlier in the trajectory of an illness and is even felt immediately a new high impact diagnosis is realized. Clinicians may avoid discussing end-of-life decisions with patients for fear of causing undue anxiety.63 This is despite the patients desire to address the issues.
CULTURAL DIFFERENCES IN PALLIATIVE CARE
Cultural differences in the approach to end-of-life decisions, advanced care planning and withdrawal from dialysis have been addressed by Davison and Holley.43 Non-Western cultures, significantly represented in the Australian population, may have very different understandings of the medical system, health and disease. These cultural sensitivities need to be taken into account when discussing palliative care and end-of-life decisions.
DIALYSIS STAFF INVOLVEMENT IN PALLIATIVE CARE
Several studies have indicated that the beliefs and values of health professionals have a clear impact on the integration of palliative care into the management of ESKD patients. Twohig and Byock64 found that the focus of care remained on cure and prolongation of life and that ethical cultural and legal issues impact on the clinical decision to withdraw or withhold dialysis. In their study on physicians' decisions to withhold or withdraw life-sustaining treatment, Farber et al. reported that internists found it emotionally harder to withdraw rather than withhold treatment.65
In 2002, Siegler et al. reported inadequate communication and planning for patients with ESKD around palliative care transition, increased patient suffering.21 This was later supported by a survey conducted of staff directly involved in dialysis care including nurses and social workers and found there was a deficiency in end-of-life discussion with patients and poor communication of the discussions that had occurred with staff actually caring for the patients.66 Not only should dialysis patients selecting conservative management be clearly identified, those directly caring for the patient also need to be aware of the outcome of end-of-life discussions.
REVIEWS ON PALLIATIVE CARE IN ESKD
There have been previous reviews of palliative care in ESKD. Brown et al. reviewed palliative care in nephrology and issues covered under the palliative care umbrella.67,68 Germain and Cohen noted the increasing mortality of incident dialysis patients associated with more elderly accepted for dialysis.55 Haras highlighted the lack of advanced directives and palliative care among patients with ESKD and how senior nurses are well placed to initiate such care and discussion.69 Jablonski, reviewed misconceptions that may be barriers to incorporating palliative care into the routine management of ESKD.70 Holley reviewed palliative care management in ESKD with a focus on advanced care planning, referrals to hospices and bereavement.71,72 Lichodziejewska-Niemierko and Rutkoski focused on the provision of palliative care support from the time of diagnosis through to family bereavement and on symptom relief.73 Poppel et al. reviewed the Renal Palliative Care Initiative at a tertiary hospital and described the benefits to their patients.44 They also described the evolution of renal supportive care from an initial focus on dialysis withdrawal through its expansion to incorporate the full continuum of CKD.74 They highlighted the need to provide guidelines and tool kits to enable clinicians to achieve their goals in this population. Dialysis withdrawal has been reviewed by Murtagh et al.56 along with White and Fitzpatrick who highlighted the paucity of available data.75 These authors provide practical ways of handling the palliative care patient withdrawing from dialysis and emphasize the importance of advanced directives and thorough assessment before stopping treatment. The role and benefits of a comprehensive conservative management approach were reviewed by Burns and Carson.76 Price reviewed the role of the nephrology nurse in palliative care for patients highlighting the importance of early referral and shared care.77
EVIDENCE-BASED GUIDELINES, CORE CURRICULUM, POSITION STATEMENTS, STANDARDS AND TOOLS IN PALLIATIVE CARE
There are many resources available, developed predominantly in the USA and the UK, to support those enquiring about palliative care in ESKD. A selection of these is summarized below to illustrate the breadth of resources available.
The UK Expert Consensus Group have developed evidence-based guidelines for symptom management in adults who are dying from ESKD.4 These guidelines developed from the Liverpool Care Pathway for the Dying Patient, which was used initially for terminal cancer but subsequently for stroke and heart failure patients. An Expert Consensus Group for patients dying with renal failure found those dying with renal failure had similar symptoms to those dying with terminal cancer hence the Renal Liverpool Care Pathway prescribing guidelines were developed with the aim of controlling these symptoms.78 The NKF KDOQI guidelines state Nephrologists should be familiar with the principles of palliative care and should not neglect hospice referral for patients with advanced kidney failure.2,5 The CARI guidelines do not address palliative care15 and formulating guidelines in the Australian context should be a high priority. However, the Kidney Health Australia website provides information for patients on conservative approaches both pre-dialysis and withdrawing from dialysis.79
National Kidney Foundation core curriculum in nephrology summarized the relevance of palliative care and its incorporation into dialysis units.5 It highlights the usefulness of advanced care planning in patients with ESKD and strategies to increase its use.
The American Society of Nephrology and the Renal Physicians Association produced a position statement on End of Life Care in 2002.1 This is a comprehensive document that addresses advanced care planning and directives, hospice care and palliative care. It also makes recommendations, which includes ensuring education of multidisciplinary renal team members in palliative care principles including advanced care planning, supporting the patient requesting dialysis withdrawal with palliative care referral and the development of renal unit policies and protocols to ensure advanced care planning occurs.
Clinical practice guidelines
The Renal Physicians Association and the American Society of Nephrology also provide a clinical practice guideline on dialysis initiation and withdrawal.80
Standards for providing Quality Palliative Care for all Australians were published in 2005.81 Although there is no specific reference to patients with kidney disease the standards provide guidelines that can be applied to all diseases. The standards do emphasize the need to encompass the patient and their family's wishes and needs in the decision-making process of care planning. In addition, access to palliative care services should be available independent of diagnosis and should be based on clinical need.
The only tool in the public domain that we could find was in the National Health Service National End of Life Care Program to enhance end-of-life care in those without cancer. It introduced the tool to support patients with kidney failure.6,82
DISCUSSION AND CONCLUSIONS
Palliative care support should be offered to patients selecting ESKD management options including a conservative treatment pathway or withdrawal from dialysis. The increased acceptance of the elderly with comorbidities, nursing home patients with their inherent poor outcomes emphasizes the importance of supporting end-of-life decisions with palliative care. There should be an associated focus on adequate symptom control, which has been poorly attended to in ESKD as evidenced from some studies. The strong emotional influence, including grief and loss, apparent in the literature for patients, family and health professionals, suggests that there is a real need for education and support in relation to palliative care planning for each of these groups. To do this effectively further rigorous studies are needed to provide a stronger evidence base upon which to advise patients and their families when faced with impending dialysis. Some countries such as the UK, USA, Italy and Canada are well advanced in providing treatment guidelines and resources once dialysis withdrawal is planned but a greater focus on the pre-dialysis phase is required. Multidisciplinary nephrology teams must ensure that patients and their families are accurately informed so they can choose between dialysis and conservative treatment supported by palliative care. The inclusion of palliative care guidelines for Australian nephrology through the CARI guidelines should be considered.
The National Health and Medical Research Council is the funder of this study through Grant B0016419.