Christina Radcliffe and Sarah Yardley - Review Date Jun 2016 (Senior Editor Sunil Bhandari)
- Supportive and palliative care should be offered on the basis of patient need and symptom burden, regardless of prognosis or underlying diagnosis
- Advance care planning is crucial to good supportive and palliative care; and should take place alongside renal replacement therapy and other disease-modifying treatments
- Common symptoms to ask about include: pain, breathlessness, weakness and fatigue, anorexia, depression and distress secondary to functional limitations
- Patients on renal replacement therapy over the age of 80 with a WHO performance score of 3 or more do not have a clear survival advantage in comparison to those managed conservatively and have increased rates of hospitalisation
- Conservative management requires active correction of reversible pathophysiology alongside symptom management and support for patients and families
- Professionals should explore ways to work collaboratively with patients and families in shared decision-making and can make use of varied resources to facilitate this co-operation
- Patients can benefit from joint working between specialists in nephrology and palliative care
- Morphine and diamorphine are not recommended in CKD stage 5 without dialysis; alternative opioids are preferred according to required route and stage of illness. Alfentanil and fentanyl are the preferred option for injectable opioid analgesia in ESRD
- Haloperidol is the first line recommendation for nausea and vomiting; glycopyronium or hyoscine butylbromide for excessive respiratory tract secretions; and midazolam for terminal agitation
- Reduced doses and increased dosing intervals are both important for all drugs (and metabolites) that are renally excreted
This is an overview of existing evidence regarding provision of supportive and palliative care in the context of end-stage renaldisease (ESRD). When possible, recommendations are based on high quality academic studies, specific to ESRD but this is a relatively under-researched area. In the absence of studies specific to renal disease best practice advice from recognised sources such as the UK Renal Association guidelines and the wider literature on supportive and palliative care is provided.
A systematic review of the clinical literature has been completed. Full details of the search and selection strategy are available from the authors, but in summary, electronic searches in Medline, Embase, CINHL and BNI identified a potential 323 papers for review from searches covering chronic renal disease, supportive and palliative care. Following title and abstract screening, 83 papers were selected for further consideration. Inclusion criteria at this stage were: CKD and/or supportive or palliative care, ESRD defined as stage 4/5 (MacGregor & Taal 2011), empirical data presented (of any type) or review, meta-analysis or evidence-based guideline. One of these papers was unobtainable and a further 34 were excluded on full text review due to poor methodological quality or because the inclusion criteria (as above) were not met. Therefore, 48 papers have been included.
Review of the evidence on drugs and prescribing was limited to clinical pharmacology relevant to palliation and renal disease. General principles of prescribing in renal impairment still apply in palliation, although the balance of risks and benefits of specific drugs may be weighed differently according to individual patient priorities. This guidance does not replace that contained within national formularies such as the:
- British National Formulary (https://www.bnf.org/bnf/index.htm)
- Electronic Medicines Compendium (https://www.medicines.org.uk/emc/)
- Palliative Care Formulary (https://www.palliativedrugs.com/index.html)
- Renal Drug Handbook (available at: https://www.rcgp.org.uk/shop/books/medical-disciplines/prescribing/the-renal-drug-handbook.aspx)
Prescribers should refer to these, plus the most up-to-date national guidance and local policies before prescribing medication for a particular patient.
In this review we use the term end-stage renal disease to describe the non-functional state of the kidneys regardless of cause, and the term chronic kidney disease (CKD) to refer to underlying pathology leading to this state.
Aetiology and pathology of end-stage renal disease (ESRD)
Renal disease can be divided into Acute Kidney Injury (AKI), Chronic Kidney Disease (CKD), and End-Stage Renal Disease (ESRD) (also known as Established Renal Failure, ERF). CKD leads to ESRD when renal function progressively deteriorates in an irreversible manner. Both AKI and CKD can be due to primary renal disease or secondary to other diseases or medical problems.
The UK Renal Association has produced a guide for CKD (https://www.renal.org/ckd). This provides online support for the diagnosis and renal management of CKD. Disease severity is divided into 5 stages according to eGFR (estimated Glomerular Filtration Rate). Detailed information about how the eGFR is calculated and relevant cautions related to it can be found in the UK Renal Association Guide (MacGregor & Taal 2011). The stages of CKD (MacGregor & Taal 2011) are detailed in the table below:
Table 1: Stages of CKD
Whilst patients at any stage may have a need for supportive and palliative care due to other reasons, and patients in need of supportive and palliative care may have any stage of renal failure, the population focused on here is those with stage 4 (severely reduced renal function) and 5 CKD (ESRD or ERF) who are those most likely to need supportive and palliative care because of their CKD. It should be noted that some patients with AKI will also need supportive and/or palliative care if their individual circumstances mean they are symptomatic, not able to receive active treatment for the underlying cause, or disease-specific treatment of the underlying cause of their AKI fails. Such patients should be discussed on an individual basis with specialist palliative care teams.
Patients who develop stage 4-5 CKD as a consequence of another life-limiting illness and who have a short prognosis are unlikely to be offered renal replacement therapy (RRT) (Mailoux 2010). Nonetheless, reversible causes of any deterioration in renal function (which may be ‘acute on chronic’) should be sought: these include exclusion or treatment of sepsis, renal tract obstruction, urinary retention, nephrotoxicity secondary to drugs, and cardiovascular instability.
Patients who have undergone kidney transplant are not specifically included in this review, aside from within groups conservatively managed or given dialysis (i.e. post-transplant failure). Therefore where the term RRT is used in this review, this indicates dialysis, which may be haemodialysis or peritoneal dialysis.
Definitions and terminology: supportive care, palliative care and conservative treatment
The terms ‘supportive care’ and ‘palliative care’ are not mutually exclusive and are often used interchangeably in the literature (Cramp & Bennett 2012). The use of the word ‘best’ as a precursor to either of these terms appears to stem from a desire to emphasise that these are not second rate options. Neither option implies there can be a lack of engagement from health professionals in the holistic care of patients, once disease-modifying treatment is no longer possible, or no longer adequately provides for all of their needs.
The National Council for Palliative Care (NCPC) defines supportive care broadly:
“Supportive care helps the patient and their family to cope with their condition and treatment of it – from pre-diagnosis, through the process of diagnosis and treatment, to cure, continuing illness or death and into bereavement. It helps the patient to maximise the benefits of treatment and to live as well as possible with the effects of the disease. It is given equal priority alongside diagnosis and treatment.” (NCPC 2013)
From this definition it can be seen that ‘supportive care’ places emphasis on integrated management and care plans that address quality of life and symptom control regardless of illness trajectory or stage of disease. Increased use of the term may also have occurred as it carries less direct ‘end-of-life’ connotations than the term ‘palliative care’ in society, meaning that it is a less threatening idea to introduce as part of a healthcare package alongside active disease treatment. Nonetheless there are similarities between this definition and that offered by WHO for palliative care:
“Palliative care is an approach that improves the quality of life of patients and their families facing the problem associated with life-threatening illness [our italics], through the prevention and relief of suffering by means of early identification and impeccable assessment and treatment of pain and other problems, physical, psychosocial and spiritual.
- Provides relief from pain and other distressing symptoms;
- Affirms life and regards dying as a normal process;
- Intends neither to hasten or postpone death;
- Integrates the psychological and spiritual aspects of patient care;
- Offers a support system to help patients live as actively as possible until death;
- Offers a support system to help the family cope during the patients illness and in their own bereavement;
- Uses a team approach to address the needs of patients and their families, including bereavement counselling, if indicated;
- Will enhance quality of life, and may also positively influence the course of illness;
- Is applicable early in the course of illness, in conjunction with other therapies that are intended to prolong life, such as chemotherapy or radiation therapy
- Includes those investigations needed to better understand and manage distressing clinical complications.” (WHO 2013)
Palliative care can, therefore be seen to encompass supportive care and go beyond it in the context of life-threatening (advanced, or progressive) illness where there is an expectation of death in the foreseeable future. Specialists in palliative care have important roles, not only for the delivery of direct patient care in this context, but also in supporting and collaborating with other teams, such as those providing expertise in nephrology (Murtagh 2014).
In the context of ESRD it is important to realise that the supportive and palliative care needs of patients should be considered regardless of whether they are receiving renal replacement therapy or conservative management.
There is no standard definition of ‘conservative management’ published in the current literature but it usually includes a combination of measures that involve active management of the effects of renal failure as well as efforts to reduce or slow disease progression. In essence, comprehensive conservative management can be summarised as ‘holistic patient-centred care which includes active symptom management, shared decision-making, detailed communication including advance care planning, psychological support, social and family support, and cultural and spiritual domains of care, but not dialysis’(Fliss Murtagh, personal communication 2014). Models of care delivery for conservative management may include regular clinic attendance and multidisciplinary team review with input from nephrologists, specialist nurses, dieticians, social workers and others; in order to ensure active management of fluid and electrolyte balance, anaemia, blood pressure and cholesterol and renal bone disease, alongside measures targeted at quality of life, symptom control and advance care planning (Hussain et al 2013, Carson et al 2009, MacGregor & Taal 2011) (See section on conservative care in Clinical Issues).
Fassett et al (2011) identified three intertwined facets of best practice in the assessment and management of CKD. These are:
- Patho-physiological condition: from diagnosis to initiation through to withdrawal of RRT with efforts aimed at preventing decline in renal function and preventing or treating serious co-morbidities
- Symptoms and quality of life: including addressing pain, weakness and fatigue, anorexia, depression, loss of roles and limitation of mobility
- Psychological condition of patient, their spouse and family and the relationship with the renal team: with attention to uncertainty/ certainty of prognosis, shared understanding, responses to disease progression, channels of communication, decision-making and advance care planning.
The need for palliative care alongside usual care may occur at any time in the illness trajectory and may arise from unresolved concerns in any of these three strands. Gayomali et al (2011) et al describe how patients with ESRD are often referred to palliative care services late in their illness trajectory when earlier referral could improve QoL, symptom control and advance care planning (ACP). In addition referrals were often made in this study, not by nephrologists, but by other physicians when the patient presented to them - for example, due to an acute deterioration or intercurrent illness.
Arguably usual care should include attention to all aspects of these three strands and generalists in palliation (who may be specialists in nephrology) should seek to address issues that arise, drawing on specialist palliative care advice and support in complex cases. This expectation is endorsed in the UK Renal Association Guidelines for planning, initiating and withdrawal of RRT (MacGregor & Taal 2011).
In the current renal palliation literaturethere are gaps in understanding regarding the epidemiology of CKD, renal replacement therapy and conservative care. Articles identified during this review of the literature demonstrated that the incidence and prevalence of CKD and, therefore, for end-stage renal disease is rising. Between 2001 and 2008 there were 11,293 people who had chronic renal failure recorded as an underlying cause of death. A further 94,520 deaths mentioned CRF as a contributing factor (NEOLCIN 2010). In 2006, the number of patients receiving dialysis or a transplant in the UK reached 43,901 (UK renal registry 2007). In 2012, there were 54,824, with an incidence and prevalence of adults accepted for RRT in the UK being 108 and 861 per million population (UK renal registry 2013). In the same year, the US renal data system reported over 430,000 people on dialysis (US Renal Data System 2013). The reasons for the rising numbers of patients requiring RRT are multi-factorial. Worldwide, many countries have aging populations who are living for longer with one or more chronic diseases. Both primary and secondary renal disease is prevalent in these populations.
In addition we identified two published studies that sought to quantify all deaths from end stage renal disease by undertaking population based research regarding all ESRD-related deaths, including dialysis, dialysis withdrawal and conservative management. Hemmelgarn et al (2012) conducted a community cohort study in Alberta, Canada of 1,816,824 adults with eGFR baseline of 15ml/min/1.73 m2 or higher, and not currently receiving renal replacement therapy. Measurement of adjusted rates of treated (dialysis / transplantation) kidney failure, untreated kidney failure and death were made. During a median follow up of 4.4 years, 5.36% of the population died, 0.18% were treated for kidney failure and 0.17% developed kidney failure that was managed conservatively (i.e. an almost equal number to those treated). Conservative management was increasingly common with increasing age of patients at all rates of eGFR (Hemmelgarn et al 2012).
More recently similar findings emerged when Sparke et al (2013) sought to estimate the total incidence of kidney failure in Australia, including patients receiving conservative management. This was a data linkage study of observational datasets recorded nationally. In the period 2003-7 there were 21,370 incident cases of kidney failure recorded, with similar numbers receiving intervention (renal replacement therapy / transplantation: n=10,949) to those receiving conservative management (n=10,421) (Sparke et al 2013). As with Hemmelgarn et al’s 2012 study older patients were disproportionately more likely to be managed conservatively. Both of these studies report notably higher incidence of conservative care than anticipated.
For people who develop CKD, a decision needs to be made whether or not RRT in the form of dialysis is an appropriate intervention. This decision-making is complex (see decision-making in Clinical Issues). Most studies of choices made are limited to patients referred to specialist renal services, and therefore data on patient populations who are not referred for consideration of RRT is limited. However, studies suggest a further 20% of patients with advanced CKD are managed conservatively through General Practitioners and other generalists not referring to specialist services (Murtagh et al 2007a, Smith et al 2003).
Indicators of prognosis
In the presence of high co-morbidity there may not be any survival advantage to dialysis, and these patients are likely to experience a high symptom burden regardless of management (Douglas et al 2009). Young et al found that life expectancy decreased as age and time on dialysis increased, with a rapid fall in five year survival in patients over the age of 45 years (Young et al 2009).
Key Point: Patients over 65 years who start dialysis have a 5 year median survival of 14.5% (Ansell et al 2008).
It has been repeatedly found that severity of co-morbid conditions and functional capacity are more important than age in predicting survival and morbidity of patients on dialysis but when Chandra et al (1999) investigated the potential cost–benefit balance of denying dialysis to high risk patients (who requested it) they concluded that no significant savings could be made without losing potentially long term survivors, an effect deemed unacceptable. No more recent economic analysis was identified in the literature.
Falling serum albumin levels have been found to be a consistent and strong predictor of death in many studies: a serum albumin level of less than 3.5g/dL is associated with a 1 year mortality of approximately 50% (Renal Physicians Association and American Society of Nephrology 2000).
For patients on dialysis estimates of residual renal functions can be made from interdialytic weight gains (higher weight = lower urine output). Patients with even 3-4% residual function can survive weeks to months but prognosis is variable, particularly with varying co-morbidities. However, while this and functional scales are helpful indicators at population level and may provide a general guide, there are no sufficiently accurate models of prognosis for specific individuals that can be used in clinical-decision making at present (Cohen et al 2006). Death can occur either through co-morbidities or directly due to renal disease through volume overload, uraemia and hyperkalaemia (with cardiac dysrythmias). In these patients, there is a high risk of breathlessness and uncomfortable peripheral oedema (Farrington & Chambers 2010).
Survival in ESRD: RRT v conservative management
In the most recent study of RRT versus conservative management, Hussain et al (2013) compared survival, hospital admissions and palliative care access of patients aged over 70 years with CKD stage 5 according to whether they chose RRT or conservative management. This retrospective observational study of patients attending a pre-dialysis clinic identified 172 patients who chose conservative management (39%) and 269 who chose RRT (61%). In this study, 7% of patients changed from RRT to conservative management (due to personal preference or significant coincident events such as diagnosis of metastatic cancer) and 0.6% changed from conservative management to RRT (due sudden deterioration precipitated by acute inter-current illness).
Those in the RRT group survived longer than the conservative group when survival was calculated from the time eGFR was <20mL/min, <15mL/min or <12mL/min but stratification of factors influencing survival also found that there was
Key Point: no survival advantage in receiving RRT for those who were older than 80 years or with a WHO performance score of 3 or more.
The effect of RRT on survival was also significantly reduced in patients with high Charlson Comorbidity Index scores. The relative risk of an acute admission to hospital (RRT v conservative management) was 1.6 (p<0.05, 95%CI =1.14-2.13). 47% of the conservative management group and 69% of the RRT group died in hospital. 76% of the conservative management group access community palliative care services; none of the RRT patients did so. At 2 years 46% of the RRT and 20% of the conservative group were alive. Overall, age, Charlson Comorbidity Index score, WHO performance score and independent living all correlated with survival. Those in the conservative group were older than the RRT group but did not differ in gender, ethnicity or underlying cause of CKD. The conservative group were also more likely to have a lower albumin, diabetes, hypertension, chronic lung disease and dementia. They tended to have poorer performance status and were more likely to be institutionalised or live with family who were primary carers for them (Hussain et al 2013).
This study mirrors the findings of previous studies internationally. For example, Carson et al found that median survival in an observational study of a single centre in the UK of 202 elderly (>/=70yrs) with ESRD who chose conservative management (n=29) or RRT (n=173) was 37.8 months (range 0-106) for RRT and 13.9 months (range 2-44) for conservative patients. RRT patients had higher rates of hospitalisation. Conservatively managed patients were more likely to die at home or in a hospice These authors concluded that dialysis prolonged survival for elderly patients with significant comorbidities for up to 2 years but did not increase hospital free days. Again the conservatively managed group were older but neither the burden of comorbidity nor the prevalence of diabetes / other clinical markers was significantly different between the groups (Carson et al 2009).
Smith et al (2003) conducted a cohort study of patients with ESRD who underwent multidisciplinary assessment and counselling about treatment options in a renal unit. RRT was recommended to 258 patients and palliation to 63 in a 54 month period. Those who received a recommendation for palliation were more functionally impaired, older and more likely to have diabetes. They were also less likely to die in hospital. Out of the total 321 patients, 34 died during the follow up period (57 months); 10 directly of renal failure. 10 patients who had been recommended palliation opted instead for dialysis: in these patients median survival after initiation of RRT was 8.3 months which was not statistically significantly longer than survival in the rest of the palliative cohort (6.3 months from putative date of dialysis initiation if this option had been chosen). The lack of impact on survival raises difficult questions about patient choice, information giving and ethical management of resources given the burdens and cost of dialysis.
Opting for conservative management
O’Connor & Kumar’s review of the literature regarding conservative management was conducted in 2012. They found that conservative management resulted in a median survival of at least 6 months (range 6.3-23.4) across 13 studies. As with other researchers the survival advantage of dialysis was least clear in the elderly and was reduced in patients with comorbidities, most notably ischemic heart disease. The patients in the primary studies of this review reported high symptom burden although it was unclear whether this resulted in any difference in quality of life compared to patients on dialysis due to a lack of age-matched comparison studies (O’Connor & Kumar 2012).
Tse (2009) conducted a retrospective study of the characteristics of patients choosing palliation rather than dialysis (n=36 out of 96 interviews for advance care planning). Patients opting for palliation tended to be older, included more widowers, were more financially dependent on family, had a higher incidence of diabetes, and higher modified Charlson comorbidity index scores. They needed more assistance to walk and were more likely to have a caregiver at home. Reasons for declining dialysis included perceived physical burden, social burden and psychological burdens. They received palliation for a median duration of 132.5 days (range 4-437).
In an observational uncontrolled analysis of 11 patients all aged over 75, who had declined dialysis despite being eligible, and so were managed conservatively it was found that GFR dropped by a mean of 2.4 ml/min per year. These patients were reported to have comparable quality of life to patients on dialysis but to be functionally less well than non-uraemic age matched controls. Further research is needed to test if this finding is replicated; and assess if it has clinical significance with respect to prognostication and symptom management (De Biase et al 2008).
The morbidity and mortality rates experienced by patients with ESRD are significant (https://www.usrds.org/adr.aspx). Patients on RRT are maintained in a persistently uraemic state and once RRT is withdrawn, or if patients with CKD do not receive it, uraemia will get progressively worse (Chambers Ch15).
Key Point: In a retrospective review of 35 patients (1992-1999) who withdrew from dialysis and were followed daily by a palliative care team, mean survival was found to be 10 days.
The commonest symptoms during this period were confusion, agitation, pain and dyspnoea. A third of patients were cognitively impaired at the time of the withdrawal decision (Chater et al 2006). Similar findings were reported by Cohen et al (1995) in a study of symptom burden of patients who chose to discontinue dialysis.18 patients discontinued after a mean duration of 43.6 months on haemodialysis and they lived a mean 9.6 days after termination. The quality of these days was variable underscoring the need for supportive and palliative care.
See sections on decision making and conservative care in Clinical Issues.
Decision-Making in Dialysis
This section summaries evidence behind decision-making in chronic renal failure, addressing decision-making about both starting and stopping dialysis, and exploring alternative options in ambulant patient settings.
Decisions are required on a more rapid basis when there is a new presentation of ESRD due to life threatening illness.
Dialysis: To Start or Not to Start
Dialysis is usually considered in the renal clinic as the Glomerular Filtration Rate (GFR) falls to below 20 mls/min (Carson 2009) as time needs to be spent educating patients and their carers about modalities of renal replacement therapy (RRT); then securing access - aiming to commence RRT as the GFR drops to below 10 ml/min. The decision about whether or not to start dialysis - and which modality to choose - should be a multi-disciplinary decision which ideally involves the patient and their family or carers.
There are practical, ethical and legal aspects to consider when deciding whether to start any therapy and dialysis is no exception. The increasing reporting of eGFR rather than creatinine alone, the Renal National Service Framework (NSF) (DH 2005), increasing patient expectations and technological advances, have all contributed to increasing numbers of patients receiving RRT. Many units are expanding by up to 10% per annum (Burns & Carson 2007). Despite improvements in technique, the annual mortality rate for patients undergoing dialysis is 20-25% (Davison 2010). The 90 day mortality in those commencing RRT is high (Burns & Carson 2007) and for some patients commencing RRT is not the correct choice. While it is possible to predict those patients who are at higher risk of doing poorly on RRT, it is much more difficult to model individual patient outcomes (Feest 2010).
Evidence suggests that there are several factors which alter a patient’s prognosis; and these should be taken into account when discussing whether or not to start dialysis.
Renal failure is a disease of the elderly (Burns & Carson 2007) and due to an ageing population worldwide, there are increasing numbers of elderly patients reaching end stage renal disease. The average age at the start of RRT in the UK in 2012 was 64.6 years (UK renal registry 2013). Age in itself should not be a barrier to RRT (Murtagh & Sheerin 2010), but increasing age commonly coincides with increasing frailty and co-morbidity (MacGregor & Taal 2011). As discussed above, a recent study comparing patients over 70 years of age who underwent RRT with those managed conservatively found that RRT does confer a significant survival advantage, however this survival advantage is lost for patients aged over 80 years or with a WHO performance status of 3 or more (table 2) (Hussain et al 2013). A high Charleston Co-morbidity Index also correlated with reduced survival benefit.
Table 2: WHO performance status
- 0 = Fully active, able to carry out all pre-disease activities without restriction
- 1 = Symptomatic but completely ambulatory (restricted in physically strenuous activity but ambulatory and able to carry out work of a light nature)
- 2 = Symptomatic, <50% time in bed or chair during day (Ambulatory and capable of self care, but unable to carry out work activities, up and about >50% of the time)
- 3 = Symptomatic, >50% time in bed or chair during day, but not bedbound (capable of only limited self care, confined to bed or chair 50% or more of waking hours)
- 4 = Bedbound (Unable to carry out self care, completely confined to bed or chair)
Guidelines in both the USA (Galla 2000) and UK (MacGregor & Taal 2011) suggest that the decision to commence RRT should be made in conjunction with the patient and carers where appropriate. Some patients will elect not to commence RRT. The reasons for this have been explored in qualitative studies and include: the need to make an autonomous decision; the arduous nature of dialysis; difficulties in getting to the dialysis centre; previous experience of family or friends on dialysis; perceptions that dialysis indicated death sentence; repeated venepuncture and inability to make a proactive decision (Noble et al 2009).
In a Canadian based questionnaire study of patients receiving dialysis, patients were reported to rely heavily on their nephrologist and family practitioner for medical information. The majority (60.7%) stated that they regretted having started dialysis, and 52% reported that it was their physician’s wish that they commence dialysis, with 14% following a family wish (Davison 2010). It is therefore important that patients, their families, and physicians are aware of the alternatives to dialysis. Hussain et al (2013) also reported that only a small number of patients switched treatment modality following commencement of either dialysis or conservative treatment, highlighting the need to make appropriate decisions early on in the disease process.
In a thematic synthesis of qualitative studies on decision-making in treatment for end stage renal disease, Morton concludes that there is often insufficient information given to patients about methods of RRT and the alternative of conservative care. They recommend the formal introduction of expert patients to assist in information giving, improved timing of information giving (prior to any urgent need for RRT) and improvement of care pathways for those who do not wish to undergo RRT - as methods to improve the delivery of patient choice (Morton et al 2010).
Following this study Morton et al (2012a, 2012b) conducted a discrete choice study in adults with stages 3-5 CKD (n=105) and caregivers (n=73) who were educated about dialysis treatment. Predictors in the study included preferences for and trade-offs between life expectancy, number of hospital visits, ability to travel including subsidised travel, hours per treatment, treatment time of day, and flexibility of treatment schedule. Home-based dialysis was chosen by patients in 65% of the choice sets presented, hospital based dialysis in 25% and conservative care in 10%. For caregivers the figures were 72%, 25% and 3% respectively. Patients were willing to forgo 23 months of life expectancy with home-based dialysis to decrease their travel restrictions (i.e. to be able to go away on short breaks). It should be noted that this was only a study of expressed preference rather than actual choice but nonetheless suggests patients may weigh quantity-of-life against quality differently to professionals and caregivers.
The NHS has developed an information giving website for patients which can be of use to aid them in decision making about renal replacement therapy which can be found at: https://sdm.rightcare.nhs.uk/pda/established-kidney-failure/introduction/. This can be used in conjunction with conversations with renal teams to empower patients.
It should be emphasised that there can be no absolute reasons for withholding RRT, and each decision needs to be made on an individual patient basis. The number of patients commencing RRT differs from country to country, and from unit to unit, but expert consensus of physician opinion states that withholding RRT may be appropriate in certain settings; including, for example, patients in a persistent vegetative state, patients with end stage heart, lung or liver disease or metastatic cancer - who are confined to bed or chair and need help with activities of daily living (Mailoux 2010).
In 2006, 24.5% of deaths of patients on dialysis in the USA were identified as being preceded by dialysis withdrawal (Murtagh et al 2007a). However, given the international variations in defining what does or does not constitute ‘dialysis withdrawal’ it is difficult to arrive at a clear estimate of how common it really is and current estimates should be interpreted with caution (Murphy et al 2014). Attempts to ascertain the risk factors for dialysis withdrawal, have concluded that being female, older, having a higher educational level and progressive disease are patient factors which may be risk factors for stopping dialysis (Murtagh et al 2007a). Societal, religious and cultural factors are also likely to have an impact.
The issue of withdrawal may be brought up by the patient, clinician, dialysis staff or family. Ideally, where the discussion is raised by clinicians, this should be done early enough to allow time and further discussion to take place. It is important to ensure that the patient is ready to have the conversation, and that support is available. If a patient begins to discuss withdrawal from dialysis, it is important to exclude depression, frustration or symptoms as potential motivations for the request. Although autonomy should be respected, it is appropriate to explore the reasons for the request, understanding of the implications and that alternatives are explored (Murtagh et al 2007a).
The General Medical Council views withdrawing therapy no differently from withholding (GMC 2012), however, this is not always the opinion of all individuals including clinicians. Discussions about stopping dialysis need to be handled sensitively. It should be recognised that prognosis is not always easy to judge when withdrawing dialysis, and patients have been known to live for some months after withdrawal of dialysis if they have some residual renal function (Cohen et al 1995, Carson et al 2009).
If the decision is made to withdraw therapy, it is important to ensure that quality palliative care is available to support the patient and their family - toachieve a quality end of life. In studies investigating the patient experience, prognostic uncertainty and a sense of abandonment have been identified as themes when patients withdraw from dialysis (Noble et al 2008). A small number of studies have investigated the quality of dying after withdrawal of dialysis, and have found the majority of families report it isgood. However, symptom burden is likely to be high at end of life (Murtagh et al 2007a). In retrospective notes audits of patients withdrawing from dialysis, average prognosis is generally found to be in the region of ten days but with a considerable range (Chater et al 2006) (See also indicators of prognosis section above In Introduction and Background).
Alternatives to Dialysis: Conservative Care
If a patient elects not to commence dialysis, or is felt to be too frail for dialysis, teams have recognised that the patient may continue to have specialist needs. With an aim to meet these requirements, some teams have set up conservative renal clinics within secondary care to manage these patients with a joint approach from renal and palliative medicine specialists and a wider multi-disciplinary team. Services are generally locally-led, but examples include joint working with palliative care specialists, renal specialist nurses and physicians and multidisciplinary notes reviews.
The aims of conservative care include managing fluid balance, anaemia, bone health and blood pressure as well as managing symptoms and maximising quality-of-life (O’Connor & Kumar 2012). Median survival in trials assessing conservative management ranges from 6 to 23 months. Patients managed conservatively are known to have high symptom burden and therefore joint management with palliative care is often warranted (O’Connor & Kumar 2012).Studies have investigated symptom burden and prognosis (see epidemiology section in Introduction and Background). Dialysis patients spend a greater time as inpatients in hospital (25 vs. 16 days per patient per year) and are significantly less likely to achieve home or hospice death (O’Connor & Kumar 2012).
The outcomes of conservative management clinics have not been explored in a trial format, and most of the evidence comes in the form of retrospective case notes audit. However, there are anecdotal reports of success in improving advance care planning, deaths in usual place of residence, hospital bed days and symptom control. In one review of such a conservative clinic, interventions performed in clinic included: symptom control; management of complications of chronic renal failure such as fluid overload and anaemia; interventions to slow progression of renal disease such as blood pressure and dietary advice; decision making and information giving and future care planning (Murtagh et al 2006b).
Symptoms in ESRD
Patients with ESRD experience a similar symptom burden to patients with end-stage malignancy or other organ failures (Saini et al 2006, Cohen et al 2006, Janessen et al 2008). Although symptom burden and severity can be unpredictable there is some evidence that for patients with stage 5 CKD significant rises in symptoms occur in the last 2-3 months and certainly the last month prior to death (Murtagh & Sherrin 2010). Several authors have produced systematic reviews of evidence for the management of end-stage renal disease including symptom control. Within these reviews cross-sectional surveys were the commonest form of data collection. Three studies used the Memorial Symptom Assessment Scale Short Form (MSAS-SF) and one used a modified Patient Outcome Scale Symptom Module (POSs); both tools have been validated in palliative care populations. Other tools have also been suggested for clinical use in renal patients to assess symptom burden, for example, a modified Edmonton symptom assessment system which uses a visual analogue scale for 10 physical / emotional symptoms (Davison et al 2006) and the dialysis symptom index (Murtagh & Weisbord 2010) although the latter is not fully validated.
Numbers of symptoms in ESRD populations reported in the literature range from 6.8-17 and most commonly included the symptoms listed below (O’Connor & Kumar 2012, Douglas et al 2008, Douglas et al 2009, Murtagh et al 2007c). Symptoms may arise as a consequence of renal disease, dialysis or co-morbid conditions (Murtagh & Weisbord 2010). Focusing on symptoms arising from renal disease, Murtagh et al (2010, 2011) recruited 74 patients with stage 5 CKD from 3 UK renal units (mean age 81 years) into a prospective cohort study of longitudinal patient surveys of symptoms in order to map individual patient trajectories. 49 of these patients died during follow up, and the authors found a marked increase in symptom distress and health-related concerns in the last 2 months of life. Several symptom trajectories were identified (stable pattern, increasing, fluctuating) illustrating the need for individualised care as well as for health professionals to be alert to marked increases in symptoms as a potential sign of increasingly imminent death.
Common symptoms for patients with ESRD include the following (figures are given from Murtagh et al 2010, 2011 as these papers are from the most recent study which confirmed previous symptom prevalence findings in other studies):
Table 3: Symptom Prevalence in ESRD
Study prevalence (and/or prevalence in month prior to death):
- Lack of energy 76% (86%)
- Poor appetite (71%)
- Pruritis (82%)
- Drowsiness 65% (82%)
- Dyspnoea 61% (80%)
- Pain 53% (73%)
- Oedema (71%)
- Restless legs 65%
- Nausea (59%)
- Poor concentration (76%)
- Dry mouth (69%)
- Constipation (65%)
Other reported symptoms in the literature include weakness, cough, dizziness, skin changes, vomiting, feeling bloated, numbness in hands/feet, arthropathy, sexual dysfunction and change in taste (Murtagh et al 2011, Murtagh & Weisbord 2010).
Anorexia, nausea and lethargy are known to be direct consequences of uraemia. For patients receiving dialysis, and for whom the dialysis itself is not an intolerable burden, changes to their regime (e.g. to daily or nocturnal dialysis) may relieve symptoms associated with uraemia (Murtagh & Weisbord 2010). This has to be balanced against the symptom burden of dialysis. Yong et al (2009) found that patients on dialysis (n=134) had a mean number of 9.3 symptoms while those in receiving palliative care (n=45) there was a mean number of 8.2 symptoms. The most prevalent and intense symptoms in both groups were fatigue, cold aversion, pruritis, weakness and insomnia (Yong 2009).
Chater et al (2006) studied the symptom prevalence in patients following withdrawal of dialysis (n=35). The commonest symptoms found were confusion, agitation, pain and dyspnoea – all symptoms known to be common in other diseases during the last few days of life. This study reported that palliative care input reduced symptom prevalence in the last 24 hours of life by approximately half for pain, agitation and dyspnoea (Chater et al 2006).
Evidence-base for Pharmacological Interventions
The use of medication to alleviate symptoms should always be as part of a broader management plan that also incorporates appropriate non-pharmacological approaches. There are no drugs without side-effects and many of the medications used to palliate symptoms carry higher risk of adverse effects in the context of impaired renal function. Nonetheless, appropriate use of medication can be of great benefit to patients experiencing symptoms as a result of chronic or end-stage renal disease, whether this is the primary cause of their clinical deterioration or secondary to other life-limiting illnesses.
Principles for any drug should be considered:
- How is it metabolised?
- Are the metabolites active? Toxic?
- How does excretion occur?
This section considers:
- Available evidence to guide pharmacological treatment choices of the symptoms described above when such evidence exists
- When to consider monitoring renal function to guide prescribing
- Additional issues to consider in patients who need pharmacological approaches to palliation but who are still receiving renal replacement therapy
- Prescribing in the last few days of life
General principles of prescribing in renal impairment still apply in palliation, although the balance of risks and benefits of specific drugs may be weighed differently according to individual patient priorities. It is not our intention to replace the guidance contained within national formularies such as the British National Formulary and the Palliative Care Formulary and prescribers should always refer to these, plus the most up-to-date national guidance and local policies, before prescribing any medication for a particular patient.
It is generally accepted that opting for conservative management of end stage renal disease should still include active management of fluid balance, treatment of anaemia and correction of acidosis, hyperkalaemia, calcium/phosphate metabolism and blood pressure, at least initially (O’Connor & Kumar 2012). These issues are managed similarly to patients at earlier stages of chronic kidney disease. Such management can contribute to symptom control through physiological correction of the patient’s disease state; and so, it should not be abruptly or automatically stopped as a patient develops the need for further palliation or approaches the very end-of-life. Rather each aspect should be considered as part of a shared decision making process with individual patients; to weigh up on-going benefits in terms of symptom relief while also offering palliation.
Literature on evidence-based prescribing for common symptoms in end-stage renal disease
Patients with end-stage renal disease may experience pain from several causes including the underlying cause of their renal disease and co-morbidities. Therefore, it is important to establish, where possible, the cause and nature of any pain; so that specifically targeted treatment can be offered if appropriate. It is not uncommon for patients with end-stage renal disease to experience neuropathic pain (from diabetes or other causes) and / or pain from co-morbidities such as cardiac or respiratory disease or underlying malignancies. Fibrosing dermopathy, calciphylaxis, secondary hyperparathyroidism (causing bone pain) and amyloidosis are all examples of other causes of pain which are more common in patients with end-stage renal disease than in other patient populations. Anecdotal data on calciphylaxis suggests bisphosphonates or hyperbaric oxygen may be of use (Cohen et al 2006).
Dean (2004) and Cohen et al (2006) reported that there had been no formal trials of analgesia use in renal disease. The lack of evidence in this area remains significant. Therefore, clinicians are often reduced to using pharmacokinetic principles to guide prescribing. Below is a summary of the available evidence as it stands. In the absence of specific evidence, liaising with specialist palliative care service allows sharing of expertise and development of joint management plans.
WHO Ladder Step 1 Analgesics and Co-analgesics
NSAIDS are problematic due to the increased risk of bleeding which may be exacerbated in patients with CKD due to uraemic induced platelet dysfunction; and the side effects of increased blood pressure and fluid retention are also likely to be problematic. In addition there is a risk that renal function impairment will worsen due the effect of NSAIDS on renal haemodynamics.
Acetaminophen (paracetamol) is less problematic although the nephrotoxic effect is subject to debate (Wilcox 2013, personal communication). It should be dose reduced or avoided in hepatic failure (Cohen et al 2006). Murtagh et al (2007c) also recommend this at step 1 of the WHO analgesic ladder to a maximum of 3g per 24 hours when eGFR is <10ml per minute.
There are case reports in the literature of codeine causing prolonged narcosis in CKD (as a possible idiosyncratic reaction) (Davison et al 2010) although the validity of such reports is difficult to interpret.
The maximum recommended dose of Tramadol in patients on dialysis is 100mg bd due to reduced renal excretion leading to accumulation (Cohen et al 2006). Murtagh et al (2007c) recommend this at step 2 of the WHO analgesic ladder as it is less problematic than other options; despite the need for dose reduction and increased dosing intervals. These authors suggest a dose of 50-100mg bd and monitoring for side effects. It should be noted that uraemia lowers the seizure threshold and therefore there is an increased risk of Tramadol precipitating seizures. Renal failure and increasing age also contribute to increasing minor adverse effects, such as nausea and dizziness - and this may contribute to poor tolerance of Tramadol. Interactions with SSRIs may also be exacerbated, so there is an increased risk of serotonin syndrome.
Gabapentin has been used in neuropathic pain although it will accumulate in CKD. For patients on dialysis a loading dose can be given following each dialysis session (Davison et al 2010).
Tricyclic antidepressants are known to cause extensive anti-cholinergic side effects although are still used.
Morphine and diamorphine. Both of these drugs risk opioid toxicity in patients with CKD as the metabolites of morphine are all renally excreted (Mercadante & Arcuri 2004, Murtagh et al 2007c, Douglas et al 2009). Signs of toxicity include myoclonic jerks, narcosis, toxic agitation, and respiratory depression. The metabolite morphine-6-glucuronide (M6G) is also problematic. It crosses the blood brain barrier slowly and so re-equilibration with the systemic circulation can be prolonged. This explains why side effects of morphine on the central nervous system may continue despite stopping the drug or post dialysis. Overall these drugs are best avoided in CKD, but in severe pain unresponsive to other approaches, and the absence of alternatives, single doses may have to be considered (Mercadante & Arcuri 2004, Douglas et al 2009).
Oxycodone. This is a semisynthetic opioid which is metabolised by the liver. The effects of its metabolites have not been extensively studied in patients with renal failure. There is some evidence to suggest that in CKD the clearance of oxycodone and its metabolites is prolonged, although this does not automatically lead to clinical toxicity (Douglas et al 2009). Equally, in case reports significant toxicities have been reported (Douglas et al 2009). Within the UK, clinicians have tended to dose reduce and increase dose intervals if using oxycodone in CKD; but there is no specific evidence to support a particular regimen to do this adjustment. Clinical experience suggests that oxycodone is considered safer than morphine but this is anecdotal (Broadbent et al 2003, Douglas et al 2009).
Hydromorphone. The metabolites of Hydromorphone accumulate in CKD although - as with oxycodone - it may be better tolerated than morphine (Douglas et al 2009).
Buprenorphine. Buprenorphine is metabolised by the liver to metabolites that are excreted in bile and via the gastrointestinal tract. One of the metabolites is weakly active. It is unclear how important accumulation of this is, although there is no evidence of significant toxic effects in the existing literature (Boger 2006).
Fentanyl. This is a strong, short acting synthetic opioid with a relatively short half-life. It can be used as a transdermal patch or subcutaneously. It undergoes metabolism in the liver before both fentanyl and its metabolites are excreted by the kidneys. Given the inter-patient variability in the pharmacokinetics of fentanyl it should always be used with caution; but as it has nontoxic and inactive metabolites, it is one of the safer drugs to use in CKD (Davies et al 1996). If used in a continuous infusion, there does remain a risk of accumulation and increase in half-life which should be monitored (Douglas et al 2009, Murtagh et al 2007c, 2011).
Alfentanil. This is related to fentanyl but with a shorter duration of action and faster onset. Hepatic metabolism produces inactive and nontoxic metabolites; and studies have shown that there is no change in volume of distribution or half-life with CKD - nor have toxicities been reported in the literature. This makes it a very suitable drug from a pharmacokinetic perspective but it does have short-comings: many people are unfamiliar with its use, it is expensive and its short duration of action means that needs to be used in a continuous infusion for most patients - as it will not provide adequate relief from breakthrough pain or during titration if given as a stat dose (Douglas et al 2009, Murtagh et al 2007c, 2011). There is some concern that tolerance to Alfentanil occurs more rapidly than with other opioids, although this has not been evidenced in palliative care patients (King et al 2011) Newer evidence on alfentanil and fentanyl is emerging but this complex and yet to be fully translated into clinical recommendations (Fliss Murtagh, personal communication 2014).
There appears to be a consensus in practice to use hydromorphone (only evidence of safety is in patients on dialysis), alfentanil and fentanyl rather than other opioids in renal disease although the evidence base for this, particularly with respect to hydromorphone, is not clear cut. Since the review by Douglas et al (2009) was published, practice in the UK has increasingly become reliant on alfentanil rather than fentanyl for subcutaneous use.
Murtagh et al (2007c) recommend fentanyl, alfentanil (in CSCI only) and methadone as step 3 analgesics while also noting that there are good theoretical reasons for considering buprenorphine (but limited empirical evidence). Hydromorphone and oxycodone are not recommended by these authors due to the limited evidence base although they would still anticipate these drugs to be less problematic than morphine or diamorphine.
For fentanyl, Murtagh et al (2007c) recommend that the dose is reduced by 25-50% giving a starting dose of 25mcg Q4H prn and for methadone a dose reduction of 50-75%. It is generally recommended that methadone is only prescribed in specialist palliative care settings or with specialist advice. It is a synthetic opioid metabolised in the liver and excreted via the gastrointestinal tract and renally. Great variation in inter-individual pharmacodynamics and differences between acute and chronic dosing mean that it is essential treatment proceeds on an individual case-by-case basis.
Xander et al (2013) conducted a Cochrane review of pharmacological interventions for pruritus. The methodology of this review excluded data other than that from RCTs, but found 38 reports comprising 40 studies with 1286 participants and 30 treatments for inclusion in the analysis. This is perhaps indicative of both the significance of this symptom for patients and the lack of an effective treatment, possibly because of multi-factoral aetiology of pruritus. As with many symptoms, identification of a specific pathophysiological cause increases the chance of identifying an effective therapeutic option.
Paroxetine was found to be effective for some patients even with different underlying causes and gabapentin effective for some patients with underlying CKD as the cause. An alternative for the latter patient group was nalfurafine or naltrexone although these drugs are often excluded from use due to risks of reducing analgesia. The review was unable to reach any other conclusions on adverse effects (Xander et al 2013).
Pruritus may be due to secondary hyperparathyroidism, hypophosphatemia, increased calcium-phosphate deposition in skin, dry skin, anaemia, uraemia, iron deficiency or low grade sensitivity to products used in dialysis. If appropriate for dialysis to continue then initial management of pruritus should include assessment of whether increased dialysis may bring symptom relief. Standard renal therapies should be used to correct physiological abnormalities if this does not place undue burden on an individual patient (Cohen et al 2006). For patients for whom this is too burdensome, or otherwise no longer appropriate, then alternatives reported in other literature that can be considered include: oral antihistamines (Cohen et al 2006, Murtagh et al 2006a), naltrexone (Cohen et al 2006), ondansetron (Cohen et al 2006, Murtagh et al 2006a), thalidomide (Cohen et al 2006, Murtagh et al 2006a), capsaicin cream (Cohen et al 2006, Murtagh et al 2006a) and phototherapy with UV light (Cohen et al 2006).
Caution is advised as these suggestions are not based on trial evidence but drawn from expert consensus and wider anecdotal practice in palliative care. Overall, Murtagh et al (2006a) suggest (pragmatically given the limitations of the evidence base) the best option for the treatment of pruritus in CKD is to first exclude other systemic or dermatological causes (which might be amenable to treatment), then correct calcium, phosphate and PTH levels when possible, and to use emollients for dry skin. After this, capsaicin might be used if itch is localised, or an antihistamine (bearing in mind sedating side effects).
Anorexia may be caused by inadequate dialysis and uraemia. In addition, patients often experience taste disorders, gastroparesis and mechanical causes of anorexia which are not specific to end-stage renal disease. Cohen et al (2006) suggest zinc for taste deficiency, together with standard palliative care approaches to anti-emetics, avoiding medications which cause dry mouth and using appetite stimulants such as megace.
Assessment should be conducted to identify sleep apnoea, periodic leg movement disorder and restless leg syndrome in patients who are well enough to undergo further investigation and cause-specific management. General advice includes avoidance of caffeinated or alcoholic drinks prior to sleep and if necessary use of sedative medications. Standard medications such as zolpidem and benzodiazepines have been used (Cohen et al 2006).
Correction of anaemia may be appropriate. Malnutrition should also be addressed if patient able to undertake dietary change and depression should be excluded (Cohen et al 2006). Blood transfusion and Erythropoietin may be helpful.
Avoid laxatives which contain potassium or require high fluid intake e.g. macrogol (Movicol) (Murtagh & Weisbord 2010).
Diuretics may help if patient has residual renal function, and it is important that the patient is educated to continue salt and fluid restriction where appropriate (Murtagh & Sherrin 2010). Sometimes high doses of loop diuretics, with or without a thiazide diurectic, are useful. Potassium-sparing diuretics are not usually useful, as they are weak diuretics and risk hyperkalaemia. Short-term or intermittent ultrafiltration can also be used in predialysis patients, or those that have stopped dialysis.
There is no good evidence base regarding safety and effectiveness of pharmacological interventions but SSRIs are recommended as first line (with caution) with TCAs used for resistant symptoms. Levels of sertaline, fluoxetine and venlafaxine are not raised in ESRD but levels of paroxetine are raised. Other drugs are not recommended due to lack of evidence or risk of adverse events. Lithium has been used and is removed by dialysis, however should be used second line to anticonvulsants for bipolar disorders due to unpredictable pharmacokinetics (Cukor et al 2010).
Monitoring Renal Function to Guide Prescribing
Key Point: An eGFR of less than 30ml/min significantly alters drug metabolism and increases the risk of toxicity due to accumulation of drugs that are renally excreted (Douglas et al 2008); and therefore all prescriptions should be reviewed with knowledge of the GFR.
Pharmacological Approaches to Palliation in Combination with Renal Replacement Therapy
Literature on clearance of analgesics for patients on dialysis was last reviewed by Dean in 2004. This review found that the evidence base was lacking, although on the basis of limited evidence, recommendations were made for the use of hydromorphone or fentanyl rather than other opioids. The clearance of drugs usually excreted from the body though non-renal mechanisms (most commonly the liver) will be largely unaffected by dialysis. Weight, water solubility, protein-binding (itself affected by uraemia) and volume of distribution all affect the removal of drug molecules from the blood stream during dialysis.
Morphine is likely to be removed by most dialysis procedures although this may be with a time delay - and metabolites of morphine will still accumulate between dialysis sessions. Hydromorphone is thought to be removed by dialysis although the evidence base is not robust. Given the properties of oxycodone, this too should theoretically be removed by dialysis while the properties of fentanyl and methadone suggest these would not be removed. Dose adjustment is advised for codeine which has been found to cause adverse reactions at low doses (Dean 2004)
Prescribing in the Last Few Days of Life – Evidence from Literature Specific to Final Phase of Dying
There is a relative absence of data on prescribing in the final stages of life in end-stage renal disease. Common prescribing principles established in the wider discipline of palliative care stand: patients should be involved in treatment choices as much as they are able to do so; the goal of medications should be to alleviate symptoms and distress as part of a holistic approach to support and management - and patients should have access to as required medication for common symptoms at the end of life (Cohen et al 2006)
In 2009, Douglas et al published their review of evidence for symptom management in adults who were dying with advanced kidney disease. This review, combined with the views of an expert consensus group in the UK, was used to develop guidelines for end-of-life care in this patient population. Given the evidence available the authors recognised that the resultant guidelines were based on level 3 or 4 evidence. These guidelines were endorsed by the UK Department of Health in 2008. At the time the Liverpool Care Pathway was recommended by the UK Government as a tool to improve care of patients in the last days of life, including those with CKD. However, a recent review of the LCP has suggested that a more individualised end of life care plan is more appropriate (Neuberger 2013).
Nonetheless the principles of good symptom control for patients at the end-of-life stand, including the provision of as required medications when required to prevent or control pain, dyspnoea, terminal restlessness and agitation, nausea and vomiting and respiratory tract secretions. All the symptoms are common in the last few days of life regardless of the life-limiting illness which has led to this point (Douglas et al 2009). Douglas et al (2009) focused on the use of subcutaneous medications in patients who had never had, or had stopped dialysis. Their advice is summarised below.
Nausea and Vomiting
Haloperidol is recommended as the drug of choice, based on a combination of clinical experience and its known activity at the chemoreceptor trigger zone (stimulation of which is thought to be the mechanism of uraemia induced nausea). As the metabolites of Haloperidol may accumulate in renal failure, it is sensible to commence at a reduced dose such as 0.5-1mg nocte. Douglas et al (2009) suggested that refractory nausea and/or vomiting should be treated with low dose levomepromazine as a second line antiemetic, as metoclopramide carries increased risk of extrapyramidal reactions and cyclizine may induce hypotension and tachyarrythmias (Douglas et al 2009).
Respiratory Tract Secretions
Douglas et al (2009) recommend that hyoscine hydrobromide is avoided as this drug is known to cross the blood brain barrier; and may lead to excessive drowsiness or paradoxical agitation in patients who are elderly or have co-morbidities, regardless of renal function. Patients with impaired renal function are generally more sensitive to drugs which cross the blood brain barrier so such effects might be exacerbated. Instead either glycopyronnium (glycopyralate) or hyoscine butylbromide 20mg as required can be used for excessive respiratory tract secretions. Glycopyronnium should be commenced at a reduced dose of 100mcg PRN due to evidence that it will accumulate in CKD.
It is standard practice in the UK to use midazolam subcutaneously to relieve agitation in the last days of life. The evidence base for this, rather than alternatives, has not been established through trials. Unbound levels of midazolam will be higher in patients with CKD; and so Douglas et al (2009) recommended that both dose reduction, starting at 1.25mg PRN, and increased dosing intervals were used. Levomepromazine 3.125-6.25mg SC may be a suitable alternative if symptom control is not achieved or excessive side effects occur. Benzodiazepines can worsen delirium and psychosis. So if these are thought to be the cause of agitation, an anti-psychotic agent such as haloperidol may be best.
Pain and Dyspnoea
Opioids are used for both symptoms. Different drugs within this group are known to have different toxicities in renal failure. However, Douglas et al (2009) did not find definitive evidence to recommend a single opioid above all others. Their work is summarised in the section on pain above, therefore, takes into account the evidence reviewed by them, plus their clinical expertise and practical considerations.
Advance care planning (ACP) is a voluntary process of discussion about future care between an individual and their care providers. This discussion allows the clinical team to explain the likely course of events and allows the individual to express their preferences for future care (Henry and Seymour 2013).
Any member of the health care team can commence discussions about advance care planning, and ACP may involve discussions about wishes for what happens as health deteriorates, or what does not happen, as well as place of care. The medical team should also aim to have early conversations about ceilings of care, for example discussions about whether intensive care therapy would be appropriate. It is preferred that this occurs prior to any clinical need, and guidelines suggest that this is best done by senior team members, preferably those who know the patient, and during daylight hours (GMC 2010).
There are a number of documents available to support the communication of a patient’s wishes to appropriate members of the health care team and these are discussed below.
Although ACP is becoming more common, it is by no means universal. In a retrospective notes audit of deaths on a renal unit, it was found that the average length of stay prior to death was 22 days and discussions regarding preferred place of care are documented in only 11% of patients. Patients are involved in decision making in a small number of cases, which may reflect the frailty of this population (Noble & Rees 2006). In another retrospective study, while preferred place of care was discussed and documented in 48% of cases, 50% chose home and only 14% chose hospital; however actual place of death was hospital for 79% and home for only 21% of patients (McKeown et al 2008) This shows that it is not adequate to merely record preferences, but requires proactive action to achieve them.
Most renal patients welcome discussions about advance care planning, but expect these to be physician initiated. Patients modify their preferences for life sustaining therapy on their expected functional and cognitive ability. As a result appropriate times to review advance care planning include: at a change in health; during an acute illness; and in out-patients after discharge from hospital (Young 2009).
Statement of Wishes and Preferences
A ‘statement of wishes and preferences’ is a summary term embracing a range of written or recorded oral expressions; by which people can tell people about their wishes and preferences in relation to future treatment and care - or explain their feelings, values and beliefs that govern how they make decisions. These records are not legally binding, but should be used where a person lacks capacity to guide best interest decisions (also known as overall benefit decisions) (Henry and Seymour 2013). There is no standardised document to record this statement, though individual units may devise such a document. Statements of wishes and preferences may include personal preferences; such as where one would wish to live, or one's values (for example whether the wishes of a spouse should be taken into account when making decisions).
ADRT (Advance Decision to Refuse Treatment)
This is legally binding in England and Wales as long as it is valid and applicable to the current circumstances. It must be documented in written form and witnessed. If the treatment may be life sustaining (such as dialysis) then the ADRT must state “even if my life is at risk” or similar words. An ADRT is made by a patient with capacity, and comes into effect when they no longer have capacity. It applies only to a refusal of treatment. There is no legal or ethical framework that allows a patient to demand treatment. If an ADRT is valid and applicable, then to go against it constitutes assault. This is not the case where there is doubt and treatment is necessary to avoid death or serious injury.
An ADRT is invalid if:
- The person still has capacity to make the decision and chooses differently
- The person withdrew the decision
- After making the ADRT, the person appointed as a Lasting Power of Attorney for health and welfare, is given the attorney authority to make the same decisions as those covered in the ADRT
- The person has done something that clearly goes against the advance decision which indicates that they have changed their mind.
An ADRT is not applicable if:
- It does not apply to the situation in question and in the current circumstances
- The proposed treatment is not the treatment specified in the advance decision
- The circumstances are difference from those that may have been set out in advance
- There are reasonable grounds for believing that there have been a change in circumstance which would have affected the decision if the person had known about them at the time they made the advance decision
- A more recent ADRT has been made, or a LPA has been appointed more recently.
If an ADRT is not valid or applicable, and the person lacks capacity and there is no LPA, then decisions are made by clinicians in the patient’s best interests. It is necessary to take an ADRT into account when making a best interests decision (Regnard & Dean, 2010).
Lasting Power of Attorney (LPA)
The ability to appoint an attorney to make health and welfare decisions was created in the Mental Capacity Act 2005. The LPA must be chosen when an individual has capacity but can only act when the patient lacks capacity to make the decision in question. The attorney(s) may or may not have power to make decisions about life sustaining treatment, therefore the paperwork pertaining to the LPA should be reviewed. If the LPA has been approved by the Court of Protection and relates to decisions about health and welfare, then the decisions they make should be treated as if they were that of the patient themself.
Do Not Attempt Cardiopulmonary Resuscitation (DNACPR)
At the natural end of a terminal illness, cardiopulmonary resuscitation is not a treatment option since it will not succeed. In this situation the decision not to resuscitate should be communicated to all clinicians looking after the patient (Regnard & Dean 2010). In renal failure, the decision may be difficult, as there may be instances where reversible factors may contribute to cardiac arrest such as hyperkalaemia. Decisions should be based on an individual assessment, and not made on a clinician’s assessment of the patient’s quality of life. Decisions should normally be made by the most senior clinician in charge of the patient’s care, but should involve discussion with the wider team, patient and family where appropriate. Whether or not to offer CPR is a medical decision. A patient with a DNACPR order in place should still receive care, symptom control, and treatment appropriate to their condition.
In order to discuss CPR with patients, information is required about the likely success rates. Unfortunately this is limited in the renal population, and available data shows variable results. For the general population experiencing in hospital cardiac arrest, a meta-analysis has shown the rate of survival to discharge as 13.4%. In a study of over 200 dialysis patients experiencing in hospital cardiac arrest, 8% of patients were reported to be alive six months after the arrest. A study of outcomes in patients who experienced a cardiac arrest at a community dialysis unit showed that 60% died within 48 hours of cardiac arrest (Young 2009).
In the future, many areas will have a co-ordinated system to allow communication of the needs and preferences for a patient with palliative care needs. This is commonly known as an Electronic Palliative Care Co-ordination System (EPaCCS) and it may be appropriate to ensure that patients are on this system. In the meantime, it is helpful to ensure that patients who are at risk of deterioration and death in the coming year are well known to their General Practice; and are being managed proactively to address their current needs and to plan for their anticipated demise. This may be facilitated by a general practice supportive and palliative care register or according to the Gold Standards Framework.
Within the hospital environment, efforts are being made to ensure that patients are recognised as approaching end of life so that their needs are met; and they are appropriately communicated with and involved in decisions including about their own priorities for care and preferred place of care. The NHS Improving Quality "Transforming End of Life Care in Acute Hospitals: 'Route to Success' Programme" encompasses this using 5 key enablers: One of the key enablers is the AMBER care bundle which is used to ensure that appropriate decisions are made and communicated regarding escalation of care, where patients may not recover from an episode of illness and are thought to have a short prognosis.
Renal teams should recognise the psychological burden that RRT and consequent changing physical health has on a patient and their family. Axelsson et al reported that the feeling of sorrow was prevalent and that patients receiving RRT accepted a restricted and “heavy” life, for a chance of living longer. Patients had questions of meaningfulness of life, and lived with a constant threat of deterioration, increasing dependence and increasing losses (Axelsson et al 2012). In the face of these difficulties, it is not surprising that many patients suffer mental health co-morbidity, particularly depression and anxiety.
Depression can be difficult to identify as a result of biochemical abnormalities, symptoms and side effects which may mimic depression. However, it is thought that approximately 10-20% of patients with end stage renal disease have a major depressive disorder and up to 40% have significantly elevated depressive affect. The suicide rate is also modestly higher than that of the general population (Cukor et al 2010). There are validated questionnaires available to assess both anxiety and depression in the ESRD population, and there is some evidence for value of both psychological and pharmacological therapy for these conditions.
The level of social support available to a patient on RRT is thought to have an integral role in their outcomes. Those with more support tend to have improved access to healthcare, better treatment compliance, reduced depression and other psychological distress, and increased patient reported quality of life (Cukor et al 2010). However, an ageing population commencing RRT, together with decreasing mobility and loss of the nuclear and extended family structure, means that patients may present with very limited social support. It should also be recognised that the social support may come from the dialysis unit itself due to regular attendance and familiarity with staff and fellow patients.
Coping in the context of end stage renal disease has been investigated using validated questionnaires including the Jaloweic Coping Scale (JCS40). In patients with end stage renal disease, the most common coping strategies were “Try to maintain some control”, “Try to find out more about the situation”, “Hope for improvement”, “Accept the situation as it is”, and “Try to look at the problem objectively”. Patients with ESRD who were not yet on dialysis had significantly more worry and nervousness than those on dialysis (Klang et al 1996).
Coping strategies have also been explored in the spouses of patients on RRT. The most widely used coping strategy was that of optimism and this was also regarded as the most effective strategy. Spouses of transplant patients had significantly higher quality of life scores than those of CAPD or HD patients and also had significantly better emotional wellbeing, slept better, and were more satisfied with their sex life (Lindqvist et al 2000).
It is important to remember that care giving may be a positive experience. In a Dutch study, positive aspects of care giving and family well being were reported. Suggestions have been made that this may be due to sharing of problems and having someone to talk to. This study also found that levels of caregiver strain were higher for spousal carers. It was hypothesised that this may be due to changing relationship with the patient, new tasks, reduced social activities and life changes. Interestingly, caregiver strain was not correlated with level of dependency or the presence or absence of a qualified carer; but caregiver strain was associated with levels of depression and psychological distress in the patient - suggesting that providing emotional support may be one of the most challenging roles of the carer (Janssen et al 2012).
Sexual function in renal failure is poorly investigated in research, and probably poorly addressed in clinics and dialysis units. It has been reported to be a major concern for patients, with 43% of patients reporting reduced interest in sex, and 47% reporting difficulties in getting aroused (Cukor et al 2010).
A number of spiritual issues arise as a result of being diagnosed with a life-threatening illness. Patients are confronted with mortality and loss and may question the meaning of life and suffering (Hanks et al 2010). Patients facing illness and death may find strength in spirituality. It is important to recognise that this is wider than religious belief and can encompass the nature of meaning and reflection. Support is also delivered by the wider health care team in addition to chaplains. As a consequence of the individual and subjective nature of spirituality, it can be difficult to define (Hanks et al 2010). It may be useful to ask simple questions in a patient assessment such as “What gives you strength and hope?” or “What keeps you going day to day?” to elicit the beginnings of discussions about spirituality.
Elliott et al investigated coping with end stage renal disease through qualitative interviews with patients and their relatives. They found that several themes pertained to religious beliefs and practices, particularly in the area of “finding meaning of a life on dialysis.” Respondents reported that they had beliefs that there is a time to die; and that for dialysis patients, this time may be nearer than for others. Most respondents in this American, Caucasian, patient population were clear that their religious beliefs did not mean that they could not stop dialysis at any time. Faith often offered hope and countered fear in the group contemplating withdrawal from dialysis. Community support and connections as a consequence of religious belief were important to this patient population (Elliott et al 2012).
It should be recognised that patients may require spiritual support, informally via the multidisciplinary team, or more formally via chaplaincy, to address existential and religious concerns and worries - and to offer support, as their condition deteriorates and they approach death.
Talking about difficult issues with patients and their carers is a fundamental skill in medicine and nursing, and an inevitable part of the health care professionals work; but one in which few have received sufficient help and training. In order to protect patients, health care professionals may censor information and this often leads to increased fear, anxiety and confusion. A patient with a short or uncertain future needs honest communication in order to adapt and plan for the future (Hanks et al, 2010).
It is important to ascertain the information needs of a patient and their family. Research has shown that most patients want more information than they are given. It is important to recognise that patients may not take in all of the information that they are given; and therefore it may be necessary to offer written information, prompt sheets or recordings of consultations. It may also be necessary to give multiple opportunities to ask questions. Peer education from fellow patients has also been suggested as a method of aiding decision making surrounding dialysis (Moustakas et al 2012).
Barriers to effective communication include: inadequate communication skills training; fear of provoking distress and difficult emotions; containing one’s own emotions; being blamed for failure; over identification with the patient; confronting one’s own death fears (Hanks et al 2010).
Focus groups with patients on RRT and their families have suggested that they crave more education, including about dialysis, the likely trajectory of the illness, side effects of medications, pain control and stopping dialysis. They also desire open communication about end of life planning. Though patients and their families often differ in their opinions about when this should occur (Berzoff et al 2009).
In a systematic review of decision making in older adults with renal failure, shared decision making was found to be important, and patients relied on the clinical team to provide the right information at the right time. The development of trust between a patient and their health care team helped in facilitating this communication (Moustakas et al 2012).
Self-care is important when caring for dying patients and health care professionals should recognise the need for support for themselves.
Management Innovations in the Literature – Systems and Service Development
There is increasing recognition amongst both renal and palliative care teams that the needs of patients with end stage renal disease who are choosing not to receiving dialysis, or who are failing on dialysis, and those of their families, are not always met. As a consequence, service development in this area is on-going. The Renal Association guidelines on planning, initiation and withdrawal of renal replacement therapy suggest early identification of patients who have chosen not to dialyse (or are deteriorating on dialysis) to allow end of life care to be addressed (MacGregor & Taal 2011).
Palliative care teams have historically been mainly involved in the care of cancer patients. However, this is changing, and there is increasing recognition of their role in the management of chronic non-malignant illnesses including renal failure. Palliative care teams can be involved in the care of patients on renal inpatient wards. In a retrospective case notes audit of patients in one hospital referred to a hospital specialist palliative care team with renal failure as the primary diagnosis, 44 referrals were made over a two year period. These included patients on RRT, patients deciding to withdraw from RRT, patients on a conservative management programme and patients with acute kdiney injury, dying despite RRT (McKeown et al 2008).
In an audit of deaths on a renal unit in London, 27% of patients were referred to the hospital palliative care team, and reasons given for referral included symptom control, advice and support, hospice referral and counselling. Appropriate medications for symptom control were prescribed in 78% and documentation related to efficacy was found in 69% of patients. Non-essential medications were stopped in 69% of patients. A number of symptoms were described in the dying patient and there was evidence in the case notes that these symptoms were resolved in only 50% of cases; indicating that referral to palliative care teams may have been under-utilised, or that more education may be necessary for the renal teams caring for these patients. It is also recognised that the palliative care team may provide support for the renal team who deal with dying less frequently and who may have a long relationship with the patient concerned (Noble & Rees 2006).
It is important to note that the palliative care team does not need to be involved with a patient’s care just by virtue of their deterioration, but may be able to offer advice on symptom control, advance care planning and psychological support.
Several teams worldwide have reported on the development of joint renal and palliative care clinics to review patients on an outpatient basis. Murtagh reports a renal palliative care service where patients are reviewed jointly by a palliative care physician and a renal nurse, together with a regular multi-disciplinary case notes review of patients following a conservative management pathway (Murtagh et al 2006b). The main aim of this service was to maximise quality of life for the patients. This was achieved by prolonging renal function, controlling symptoms and supporting patients and families as well as helping patients to plan for the future. Interventions included: symptom management; management of complications of renal failure such as anaemia, fluid overload and acidosis; interventions to slow progression of renal failure such as diabetes and blood pressure management and patient education; review of decision making regarding dialysis; future care planning; family discussion and education and referral and liaison with the specialist palliative care team in the community setting (Murtagh et al 2006b).
Another UK based study describes the use of a dedicated multidisciplinary clinic to manage the needs of patients who choose “non-dialysis therapy”. This MDT consists of a consultant nephrologist, renal dietician, anaemia specialist nurse, ward link nurse, social worker, community nephrology nurse and a complementary therapis; and the clinic has links with general practice and palliative care teams. The team were able to increase haemoglobin levels through increased use of intravenous iron and Darbepoetin. The paper does not report whether this led to a reduction of symptoms as a consequence of anaemia. The study found that blood pressure control to target levels was difficult in this patient population as patients found it difficult to tolerate aggressive reduction in blood pressure, and therefore a pragmatic approach was necessary. This was not a controlled trial, but numbers of hospital admissions seem fairly low and median length of stay for all but three patients was ten days. Of the 28 patients who died during follow up, 20 died at home which probably represents a significant improvement on expected rates (Wong et al 2007).
Internationally, renal palliative care services have also been reported. In Hong Kong, Chan et al report the collaborative approach between palliative and renal services. They employed a palliative care specialist, palliative care nurses, nephrologists, a psychologist, a social worker, pastoral care workers and volunteers to assess and review patients who have foregone or withdrawn from dialysis. The renal palliative care team were able to review patients in outpatient settings, as inpatients and at home, providing good continuity of care. The focus of the care was on symptom control, psychological support, future care planning and bereavement care. The team reported this enhanced working relationships and co-working as well as providing a focus on palliation of symptoms and distress for patients and families (Chan et al 2007).
In Hing Kong, Tse (2009) reported a shared decision model between nephrology, a medical social worker, and the patient and caregiver in order to formulate an advance care plan with defined treatment goals in a medical outpatient setting. This was introduced as an active process to avoid passive withholding of dialysis. If patients decline dialysis, a conservative programme is initiated which involves reviews by palliative medicine, social work and nursing and crosses care borders.
There is clearly some benefit from collaborative working between palliative care and renal teams in terms of improved relationships, mutual education and holistic patient care. However, there is also a need for more evaluation of these services, as well as an on-going need to share innovative practice both locally and nationally.
Further studies of medications in patients with ESRD (receiving dialysis or conservative management) could clarify best practice in prescribing.
There is a lack of quality randomised trials about the effectiveness of shared care and other service developments.
In addition, more information is needed about variability in dialysis practice, criteria for conservative management, and reasons for dialysis withdrawal.
New health services research is required into both the effectiveness and cost-effectiveness of complex interventions in the care of patients with ESRD receiving palliation. This includes: psychological and social interventions, information/shared decision-making interventions (including advance care planning), models of conservative care delivery, screening and active symptom management (in whole renal or dialysis populations), and early palliative care intervention (Germain et al 2011).
Thanks to Dr Sarah Maclaran (Consultant in Palliative Medicine, UHCW Coventry) and Dr Fliss Murtagh (Reader in Palliative Medicine at the Cicely Saunders Institute, King's College London), for reviewing earlier drafts of this document
Burns A and Carson R (2007). Maximum conservative management: A worthwhile treatment for elderly patients with renal failure who choose not to undergo dialysis. Journal of Palliative Medicine; 10(6): 1245-7
Carson RC, Juszczak M, Davenport A et al (2009). Is maximum conservative management an equivalent treatment option to dialysis for elderly patients with significant co-morbid disease? Clinical Journal of the American Society of Nephrology; 4: 1611-9
Chandna SM, Schulz J, Lawrence C et al (1999) Is there a rationale for rationing chronic dialysis? A hospital based cohort study of factors affecting survival and morbidity. British Medical Journal; 318: 217-23
Cukor D, Farrell EM, Cohen LM et al (2010) Psychological and psychiatric considerations in patients with advanced renal disease IN Chambers EJ, Brown EA & Germain MJ (eds)Supportive Care for the Renal Patient. 2nd ed. Oxford, OUP. Chapter 9.
De Biase, V, Tobaldine, O, Boaretti, C et al (2008). Prolonged conservative treatment for frail elderly patients with end-stage renal disease: the Verona experience. Nephrol Dial Transplant: 23;1313-1317
Douglas C, Murtagh FEM, Chambers EJ et al (2009). Symptom management for the adult patient dying with advanced chronic kidney disease: A review of the literature and development of evidence based guidelines by a United Kingdom expert consensus group. Palliative Medicine; 23:103-10
Farrington K & Chambers EJ (2010) Death and end-of-life care in advanced kidney disease. IN Chambers EJ, Brown EA & Germain MJ (eds)Supportive Care for the Renal Patient. 2nd ed. Oxford, OUP. Chapter 15
Feest, T (2010). What determines a good outcome? The selection of patients for renal replacement therapy. IN Chambers EJ, Brown EA & Germain MJ (eds)Supportive Care for the Renal Patient. 2nd ed. Oxford, OUP. Chapter 5
Hussain JA, Mooney A and Russon L (2013). Comparison of survival analysis and palliative care involvement in patients aged over 70 years choosing conservative management or renal replacement therapy in advanced chronic kidney disease. Palliative Medicine; 27(9): 829-39
King S, Forbes K, Hanks GW et al (2011). A systematic review of the use of opioid medication for those with moderate to severe cancer pain and renal impairment: A European Palliative Care Research Collaborative opioid guidelines project. Palliative Medicine; 25(2): 525-552
Lindqvist R, Carlsson M and Sjödén PO (2000). Coping strategies and health related quality of life among spouses of continuous ambulatory peritoneal dialysis, haemodialysis and transplant patients. Journal of Advanced Nursing; 31(6): 1398-408
Morton RL, Tong A, Howard K et al (2010). The views of patients and carers in treatment decision making for chronic kidney disease: systematic review and thematic synthesis of qualitative studies. British Medical Journal; 340: c112
Morton RL, Turner RM, Howard K et al (2012). Patients who plan for conservative care rather than dialysis: A national observational study in Australia. American Journal of Kidney Diseases; 59(3): 419-27
Murphy EL, Murtagh FEM, Carey I et al (2009). Understanding symptoms in patients with advanced chronic kidney disease managed without dialysis: Use of a short patient completed assessment tool. Nephron Clinical Practice; 111: c74-80
Murtagh FEM, Addington-Hall JM, Edmonds PM et al (2007b). Symptoms in advanced renal disease: A cross-sectional survey of symptom prevalence in stage 5 chronic kidney disease managed without dialysis. Journal of Palliative Medicine; 10(6): 1266- 76
Murtagh FEM, Chai MO, Donohue P et al (2007c). The use of opioid analgesia in end stage renal disease patients managed without dialysis: Recommendations for practice. Journal of Pain and Palliative Care Pharmacotherapy; 27(2): 5-16
Murtagh FE, Addington-Hall J, Edmonds P et al (2010). Symptoms in the month before death for stage 5 chronic kidney disease patients managed without dialysis. Journal of Pain and Symptom Management; 40(3): 342- 52
Murtagh F & Weisbord, SD (2010) Symptoms in renal disease; their epidemiology, assessment and management IN Chambers EJ, Brown EA & Germain MJ (eds)Supportive Care for the Renal Patient. 2nd ed. Oxford, OUP. Chapter 7
Murtagh FEM, Sheerin NS, Addington-Hall J et al (2011). Trajectories of illness in stage 5 chronic kidney disease: A longitudinal study of patient symptoms and concerns in the last year of life. Clinical Journal of the American Society of Nephrology; 6: 1580-90
Renal Physicians Association and American Society of Nephrology (2000) Shared Decision-Making in the Appropriate Initiation of and Withdrawal from Dialysis, Clinical Practice Guideline Number 2. Washington, D.C, Renal Physicians Association
Smith C, Da Silva-Gane M, Chandna S et al (2003). Choosing not to dialyse: Evaluation of planned non- dialytic management in a cohort of patients with end stage renal failure. Nephron Clinical Practice; 95: c40-46
Sparke C, Moon L, Green F, et al (2013). Estimating the total incidence of kidney failure in Australia including individuals who are not treated by dialysis or transplantation. American journal of kidney diseases; 61:413-9
Xander C, Meerpohl JJ, Galandi D et al (2013). Pharmacological interventions for pruritis in adult palliative care patients (Review). Cochrane Database of Systematic Reviews, Issue 6. Art. No.: CD008320. DOI: 10.1002/14651858.CD008320.pub2.
Wilcox, A (2013). Advanced Course in Pain and Symptom Management. Personal Communication.