NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Clinical Guideline Centre (UK). Care of Dying Adults in the Last Days of Life. London: National Institute for Health and Care Excellence (NICE); 2015 Dec 16. (NICE Guideline, No. 31.)
8.1. Introduction
Maintaining hydration at the end of life is both controversial and emotive. There was significant media coverage surrounding the Liverpool Care Pathway and relatives' concerns about people dying from dehydration and suffering with distressing symptoms as a result of inadequate fluid intake. Suspicion was also raised that fluids were withheld and even denied to dying persons in order to hasten death. Concerns raised in the More Care Less Pathway review30 related to poor communication with the dying person and their relatives surrounding the issue of hydration. There was also the suggestion that not providing hydration (whether via the oral or parenteral route) caused more distress to dying people and their relatives than the person's lack of ability or desire to eat and drink.
Clinically assisted hydration refers to the practice of providing fluids in the form of a drip, usually either intravenously or subcutaneously (a process known as hypodermoclysis) or via a nasogastric tube or gastrostomy to prevent dehydration. It does not include assisting a person to drink via the oral route. The subcutaneous route has some advantages at the end of life as it can be administered in many different settings, including in the community without the need for staff trained in venous access but its efficacy is unclear.
Practice varies widely across the UK regarding the use of clinically assisted hydration at the end of life and decisions are often setting dependent and based on individual clinician preference and patterns in the person's medical history. The RCP National Care of the Dying Audit for hospitals85 showed that artificial hydration was in place for 29% of people at the time of death. There is a lack of clear evidence based guidance on whether clinically assisted hydration is effective in improving symptoms in the dying process for those in the last few days of life. The Committee therefore posed a review question which sought to establish if quality of care could be improved by the use of this intervention.
8.2. Review question: In patients in their last days of life, is clinically assisted hydration effective in improving symptoms and general comfort?
For full details see review protocol in Appendix C.
Table 37
PICO characteristics of review question.
8.3. Clinical evidence
One Cochrane review was identified.39 This was assessed for inclusion and fulfilled our criteria for this topic. It was therefore decided to include this as evidence using the GRADE approach and to conduct an update search for the inclusion of further studies that may have been published since the Cochrane review's search cut-off (April 2014). No additional papers were included from this search.
The Cochrane review identified 6 primary studies;14,15,20,69,101,102 these are summarised in Table 38. Using the GRADE approach, the evidence from these studies is summarised in the clinical evidence profile below Table 38 and Table 39. See also the study selection flow chart in Appendix E, study evidence tables in Appendix H, forest plots in Appendix K, GRADE tables in Appendix J and excluded studies list in Appendix L.
Table 38
Summary of Cochrane review included in this report.
Table 39
Summary of studies included in the review.
The following control comparisons were made:
- The other 3 papers had a prospective controlled trial design and compared a control group given at least 1000 ml fluid per day with those given oral hydration alone.69,70,102,101 The route of clinically assisted hydration included IV and subcutaneous. One paper added hyaluronidase to the subcutaneous fluid given in the experimental arm.101
Differences in outcome reporting across the studies meant that meta-analysis was inappropriate. The papers used different terminology to describe clinically assisted hydration, including medically assisted hydration, artificial hydration and parenteral fluids. For the purposes of clarity and consistency the term clinically assisted hydration will be used throughout this review. This reflects the language used by the General Medicine Council (GMC).
8.3.1. Summary of included studies
Table 39 provides a summary of the studies included in the Cochrane review. No further studies were identified.
Table 40 and Table 41 provide the clinical evidence summary of outcomes assessed using GRADE.
Table 40
Clinical evidence summary: Clinically assisted hydration versus placebo.
Table 41
Clinical evidence summary: Clinically assisted hydration versus oral hydration only.
8.3.2. Narrative evidence
Symptom sedation/drowsiness - Waller 1994
Waller 1994 had consciousness level as a secondary outcome.102 It was not possible to extract data from the paper; but the paper reports there was no statistical difference between the intervention and the control group.
Adverse Events - Morita 2005
Morita 2005 reported adverse events from positive fluid balance. The study reported bronchial secretion.69 It was not possible to extract data from the paper; but the paper reports there was no statistical difference between the intervention and the control group.
Survival - Cerchietti 2000
Cerchietti 2000 reported survival between a hydration and usual care comparison.20 It was not possible to extract data from the paper; but a narrative description reports that there was no significant difference in survival between groups.
8.4. Economic evidence
Published literature
No relevant economic evaluations were identified.
See also the economic article selection flow chart in Appendix F.
Unit costs
In the absence of recent UK cost-effectiveness analysis, relevant unit costs are provided below to aid consideration of cost effectiveness.
Table 42
Cost of setting up subcutaneous assisted hydration.
Table 43
Cost of setting up IV assisted hydration.
The costs reported above do not include the cost of treating adverse events due to the intervention.
8.5. Evidence statements
Clinical
Clinically assisted hydration versus placebo
There was moderate to very low quality evidence identified from 2 RCTs (n=129 and 51) conducted in hospice and home settings on a population of people with cancer who were not severely dehydrated. Both RCTs compared 1 litre subcutaneous fluid with a placebo of 100 ml subcutaneous fluid. The RCT's reported no clinical difference in clinically assisted hydration over placebo on quality of life of people, patient wellbeing, survival time or on relief of symptoms (including anxiety, dyspnoea, pain, nausea and sedation or drowsiness and delirium), between the groups.
Clinically assisted hydration versus usual care
Moderate quality evidence was identified in 1 small RCT comparing clinically assisted hydration titrated to need or oral hydration only in a population of people with cancer in the last days of life. They reported no increased adverse procedural events or over-hydration in people with clinically assisted hydration over usual care alone. There was also very low quality evidence from this study that there was no clinical difference in survival between the groups.
Low and very low quality evidence was identified from non-randomised control trials (NRCT). An observational study comparing subcutaneous fluids with hyaluronidase titrated to need to usual care (n=26) reported no clinical benefit of clinically assisted hydration over usual care on patient-reported wellbeing or symptom relief from anxiety, dyspnoea, pain, nausea, and sedation or drowsiness.
This was reflected in another large NRCT (n=226) that showed that clinically assisted hydration gave no clinical benefit in preventing delirium over oral hydration only. This NRCT also found no increased clinical risk of oedema, ascites or pleural effusions, no increased clinical benefit in hydration status or clinical difference in laboratory tests between hydration and usual care.
Economic
No relevant economic evaluations were identified.
8.6. Recommendations and link to evidence
| |
|---|---|
| Relative values of different outcomes | The Committee was most interested in wellbeing ratings as a proxy for quality of life and symptom control as a surrogate outcome for comfort. Adverse events, such as pain at injection site, local swelling and oedema, were also considered to be critical. The Committee also considered length of survival as an important outcome. Laboratory results, while discussed, were not prioritised as important outcomes in the protocol as these were excluded from the scope. |
| Trade-off between clinical benefits and harms | The clinical evidence identified showed no overall improvement in wellbeing and symptom control associated with clinically assisted hydration. Clinically assisted hydration did not lead to more frequent adverse events over placebo or usual care, and survival length did not increase or shorten when using clinically assisted hydration, but was of limited quality. Committee consensus was that some adverse events do occur in practice including cannula site discomfort, line infections and worsening oedema or heart failure when there is fluid overload. However, the Committee discussed the equivalence in efficacy between clinically assisted hydration and usual care or placebo. The Committee was divided on whether or not the addition of another intervention in the last hours or days of life would be perceived as beneficial by the people important to the dying person. Some members of the Committee considered such a procedure to be invasive, whereas others thought that it could possibly alleviate distress. They also noted that providing an intervention that was invasive and that was not likely to provide any clinical benefit could also add an element of discomfort for the dying person. The experience of the Committee was that there is benefit in some circumstances, such as in the case of managing thirst or managing delirium caused by dehydration, but this was not captured by the evidence. |
| Trade-off between net health benefits and resource use | No economic evidence was identified for this question. The Committee considered the cost of providing clinically assisted hydration when making recommendations; this cost was dependent on the care setting (hospital or community) and more resources, especially in terms of staff time, are required for providing clinically assisted hydration in the community setting. Also, the cost was different between subcutaneous and IV assisted hydration due to the different equipment, staff level and time needed. Since the clinical evidence did not show any overall clinical benefit associated with clinically assisted hydration and, in addition, clinically assisted hydration may cause discomfort to the patient, the Committee considered that providing clinically assisted hydration could, in some cases, unnecessarily increase costs. However, some benefits of clinically assisted hydration may not have been captured in the clinical evidence, such as improvement in thirst or delirium or the psychological benefit to the relatives and those important to the dying person who could otherwise feel distressed should clinically assisted hydration not be provided. For this reason, the Committee considered that, in some circumstances, the cost of the intervention could be outweighed by its benefits. |
| Quality of evidence | Evidence for the reported outcomes varied from moderate to very low quality. The evidence from the RCTs was considered to be moderate and, as such, was given more weight in the discussion. The Committee questioned the overall validity of the evidence available due to risk of bias in study design in all papers and the imprecision of a large proportion of the outcome measurements. They noted that the RCTs were terminated early due to recruitment or financial problems and were therefore underpowered. Two of the identified studies only included participants diagnosed as having mild to moderate dehydration15,20 and 2 studies excluded people with severe dehydration.14,15 Three studies excluded participants with symptoms of fluid overload (for example, ascites or congestive heart failure). Most of the observational studies identified did take account of possible confounding characteristics which may have influenced results. The studies were performed in settings outside of the UK, and the Committee raised that this may not be representative of the range of settings in which most dying persons within the scope of the Committee will enter their last days. The majority of evidence was obtained in cancer populations, which may not be representative of people dying from other causes. |
| Other considerations | Being able to eat and drink is a basic human right and need but often as death approaches, the desire and ability to take in food and fluids diminishes. Some people, if they are able, will want to continue to eat and drink right up until the point of death; others have a prolonged dying phase lasting several hours to days or even weeks when they may not be able to drink for various reasons such as reduced conscious level (possibly due to sedative drugs), dysphagia, nausea and vomiting or extreme weakness. They may develop symptoms of dehydration including dry mouth, thirst, confusion and agitation, particularly if there are associated conditions such as hypercalcaemia and opioid toxicity due to impaired renal clearance. This can cause considerable distress to the patient and those important to them particularly if hydration is not adequately assessed and managed. In an unconscious or semi-conscious patient the Committee discussed a need for careful assessment of hydration status by experienced staff to establish if the dying person is distressed with symptoms of dehydration (for example, becoming agitated or restless especially if on high dose opioids) or fluid overload (for example, increasing breathlessness). The Committee felt this may be done in conjunction with those important to the dying person and a best interests decision should be made (as per MCA guidance), taking into account the dying person's previous stated preferences, as to whether clinically assisted hydration should be continued, stopped or commenced. The Committee was also aware of the General Medical Council's guidance in 201037 on clinically assisted hydration and nutrition. Members of the Committee felt that this guidance is a good starting point for inexperienced practitioners as it provides general background information as well as helpful specific ethical guidance regarding people who lack capacity and people in a persistent vegetative state. It is important to differentiate dying from dehydration that is potentially reversible, for example, drugs given for pain relief that may have a sedating effect, meaning that a person may be unable to maintain an adequate oral intake. In the case of dying with symptoms of dehydration from an irreversible disease process, the aim of treatment is to relieve symptoms of dehydration and associated distress as a comfort, rather than to prolong life. The Committee discussed the role of blood tests to indicate hydration status in the last days of life as they were aware that this was setting dependent. The Committee felt that there was not always additional benefit to performing these tests in the last days of life. They agreed that the principle should be that these tests not be routinely undertaken as hydration status could be assessed clinically. However, if laboratory test results are present then they may guide decisions around clinically assisted hydration but no recommendation was made as this was outside the remit of this guideline. Oral hydration The Committee was keen to note that the management of hydration in the dying person should always be individualised and, wherever possible, be by oral means. They felt it important to make a recommendation that indicated that this should be encouraged. The Committee were aware of the risks of maintaining oral hydration at the end of life, when the swallowing reflex becomes progressively impaired. The Committee highlighted the need to assess the dying person for any problems with oral hydration, whilst being mindful of their preferences. The Committee felt that the decision to take oral fluids should be with the dying person where possible, but that clinicians or care providers should take steps to prevent aspiration. The Committee felt that it is often appropriate to continue to offer and assist a dying person with oral fluids if that is their wish, but some health care professionals may feel uncomfortable with managing the risk associated with this aspect of care. Offering oral fluids may include small sips of ice water or chips of ice; being ice cold may reduce the risk of this fluid being aspirated, causing distressing choking or coughing, as the dying person cannot forget it is there and inhale it. Supporting the dying person to drink may also include encouragement with a straw, a teaspoon, the right beaker or thickened fluids. The Committee felt that if enteral tubes were already in situ they could be used to provide fluids. They noted that certain clinical conditions may prevent the oral route, such as dysphagia, but that taking fluids, as required, by drinking would be the least invasive approach to managing the symptoms associated with dehydration. Care of the mouth and lips The Committee agreed that encouraging those important to the dying person to be part of oral care was also relevant and considered it important to include a recommendation that ensured that professionals encouraged this directly with those important to the dying person where desired. The Committee discussed the importance of frequent mouth care, such as cleaning teeth or dentures if they wish too. In addition, the Committee discussed the unconscious dying person and that lip care for these people was important, particularly as visible dry, cracked or bleeding lips could be distressing to those important to the dying person. The speech and language therapist co-opted expert advisor identified that good oral care is important. The availability of artificial saliva as part of the management of a dry mouth was considered by the Committee, and they were aware that certain products are pork based and therefore not appropriate for some populations. The Committee noted that the input of speech and language therapists could be beneficial in the management of oral care at the end of life, where specialist advice is required. The Committee considered the above points but as no evidence review was conducted in these areas, they chose not to make any specific recommendations. The Committee noted that there has been an MHRA alert regarding the use of oral hygiene sponges (oral swabs with a foam head). Foam heads of oral swabs may detach from the stick during use and may present a choking hazard for patients. The MHRA advice is to:
The co-opted speech and language therapist noted that many trusts have issued this statement for use by their staff, rather than withdrawing the product all together. The problem is that there is no obvious alternative, and if they are not used, the risk to oral hygiene from not having oral care is higher than the risk associated with using the mouth care sponges. Clinically assisted hydration A recommendation was made to review the need for clinically assisted hydration in people in the last few days of life preferably daily. It was felt that an assessment of hydration should be occurring routinely as part of general nursing and medical care. However, the Committee wanted to emphasise that there should be on-going monitoring and review of the dying person's hydration status and that once a decision is made to give or not give clinically assisted hydration, it is not “set in stone”. Any individual's hydration needs may change as their condition changes depending on their underlying disease and other factors, such as medications. The Committee chose to make a recommendation to discuss the risks and benefits of clinically assisted hydration with the dying person and those important to them because many criticisms of the Liverpool Care Pathway made reference to the issue of hydration and concerns of dying slowly of dehydration, and there had been little or no communication from healthcare staff about providing fluids. The Committee felt it was important to emphasise that clinicians weigh up the risks and benefits of clinically assisted hydration on a case by case basis for any dying person. They wanted to emphasise that open and honest discussions should take place around the decision making process, ensuring that the dying person and those important to them are aware of the uncertainty about its benefits and the effect on survival. The Committee felt this discussion alone may alleviate concerns (even if a decision is then made to not give clinically assisted hydration) as the dying person and those important to them will have a better understanding of hydration issues and the aims of care in the last few days of life. It is important that the dying person and those important to them are aware that the benefits of giving assisted hydration are for relief of distressing symptoms of dehydration and that fluids are not being administered to prolong life, except when there is uncertainty about whether the person is dying or there is potential for recovery (for example, in a patient who has suffered a stroke). The dying person and those important to them should also be made aware of any potential harms of clinically assisted hydration, such as the need for intravenous cannulation or insertion of a subcutaneous needle, pooling of fluids in subcutaneous tissues, fluid overload causing increased respiratory secretions and the possibility of moving care setting if clinically assisted hydration cannot be provided in the community. The Committee was aware from their clinical experience that family members may have preconceptions around the provision or ‘withholding’ of clinically assisted hydration in relation to the possibility of prolonging the dying process or hastening death. The Committee noted recent systematic reviews of the qualitative literature (Cohen et al., 2012;25,26 del Rio et al. 2012,28 and Gent et al., 2014,38) that report on the emotional, spiritual and comfort aspects of clinically assisted hydration. This literature also identifies similar key misconceptions amongst professionals as well as lay people about clinically assisted hydration.38 The lack of clear evidence about the benefits of clinically assisted hydration in the last few days of life and effects on survival meant that the Committee was unable to make a clear recommendation about when to use it or not as all decisions should be made on a case by case basis after weighing up the risks, harms and benefits to the individual. They acknowledged that the issue of hydration is very emotive and influenced by cultural and societal beliefs. The Committee felt strongly that a dying person and those important to them should be made aware of the uncertainty surrounding the effects of clinically assisted hydration on survival due to the evidence showing equivocal results. The Committee was therefore keen to make a recommendation about advising the dying person and those important to them that there is uncertainty whether: a) giving clinically assisted hydration will prolong life or the dying process and b) not giving clinically assisted will hasten death. The Committee hope that through education and implementation of this recommendation, this will lead to a change in practice and dispel some of the misconceptions that withholding hydration will speed up death or giving hydration will improve survival and therefore prolong suffering, as neither of these statements is based on robust evidence. The Committee noted that when reviewing the need for clinically assisted hydration, the wishes and preferences of the dying person and those important to them are respected, in order to capture discussions around culture and religion and any relevant decisions stated in the advance care plan. The Committee chose to state ‘respected’ to mean that in some cases it may not be possible, but that they would always be considered and taken into account. The Committee discussed that the evidence in this area had limitations, but did show equivalence in efficacy between clinically assisted hydration and usual care or placebo. The Committee noted that the clinical evidence showed no increase or shortening in survival (that is, the provision of clinically assisted hydration is unlikely to prolong the dying process), but given the low quality, there is a high level of uncertainty. Moreover, the Committee did not feel, from their clinical experience, that withholding clinically assisted hydration would hasten death. The Committee recommended a therapeutic trial of clinically assisted hydration based on consensus using their expert knowledge. However, they felt it important that a full discussion of the harms and benefits of clinically assisted hydration took place with the dying person or those important to them, as appropriate, as part of holistic care and that any preconceptions and concerns are addressed before initiating clinically assisted hydration. When considering who would benefit from clinically assisted hydration, the Committee discussed the following issues. They noted that existing comorbidities, such as heart failure, renal failure, or difficulties in swallowing, may influence a clinical decision to start or withdraw fluids. For example, a dying person with heart failure may become more breathless by administration of excessive fluid but not simply by parenteral fluids per se. The Committee also highlighted that people with cognitive disabilities (including those with dementia and learning difficulties) may find the intervention invasive and distressing. This point was re-enforced by the co-opted psycho-geriatrician expert and the Committee felt that this should be taken into consideration when starting clinically assisted hydration. On the other hand, acute delirium caused or aggravated by dehydration may benefit from hydration whether by oral or a clinically assisted hydration route. The principle of care would be to maintain adequate hydration in the dying person in the last days of life to minimise unwanted symptoms such as delirium. The Committee was aware from their experience that some cultures and religions have specific beliefs about clinically assisted hydration (for example, whether it is seen as a medical intervention or not and therefore a life prolonging measure) and recommended that this should be taken into account when considering the role of clinically assisted hydration. The Committee discussed that, for those people who are thought suitable for clinically assisted hydration after a careful weighing of the above factors, a trial of clinically assisted hydration could be initiated if the dying person was experiencing symptoms associated with dehydration (such as thirst or delirium). The Committee felt that the dying person should then be monitored at least once a day for evidence of clinical harm from the intervention; or to determine whether the hydration provided was sufficient. Whilst clinically assisted hydration was not found to cause clinical harm in the evidence reviewed, from their clinical background the Committee acknowledged that adverse events do occur, and that monitoring would be important to prevent this. If there is evidence of clinical harm then the Committee felt that any clinically assisted hydration should be reduced or stopped and this should be discussed with the dying person and those important to them. Equally, if there is no evidence of clinical benefit from a therapeutic trial of clinically assisted hydration and, in discussion with the dying person and those important to them, it is felt that fluids are not beneficial, they should be withdrawn. The group discussed whether significant discomfort at the infusion site would be a reason to discontinue clinically assisted hydration, but felt that in some cases it may be appropriate to change the infusion site. The Committee commented that, on occasion, parenteral fluids are given to relieve psychological distress of the dying person and those important to them despite no clear evidence of their effect on physical symptoms. This may be an acceptable indication, provided there are no harmful effects, but the Committee felt that it would be important to explore the dying person's beliefs and correct any misconceptions first. This was an area highlighted in the Neuberger report.30 The Committee noted from their experience that not providing clinically assisted hydration to the dying person may cause psychological distress to those important to them and that, often, clinically assisted hydration may be prescribed to alleviate distress of those important to the dying person, as well as to improve hydration in the person who is dying. They felt that this practice was acceptable especially if there was any concern that the dying person was distressed with symptoms of dehydration, provided that the dying person was not experiencing any harmful effects from administration of the fluids. The Committee also discussed that in their clinical experience parenteral fluid does not necessarily need to be administered continuously but could be provided as an overnight intervention only. They noted that it may only be necessary to provide relatively small volumes of fluid a day (for example, 1 litre) in 24 hours to relieve symptoms of dehydration. Factors such as the person's height and weight, amount of time unable to take oral fluids and presence of electrolyte disturbances, if known, should also be taken into consideration. It was also discussed that a trial of clinically assisted hydration should more readily be started when there is uncertainty that a person is dying and might recover but is currently unable to take oral fluids. This would be important to prevent death from dehydration in a potentially reversible condition. The Committee discussed whether a person dying at home would require a move to hospital for clinically assisted hydration, but agreed that this was not necessary and that it is possible to provide clinically assisted hydration at home if there were appropriate available resources. The Committee discussed that setting was not a barrier to providing clinically assisted hydration, but that this may require additional resources for implementation. Please also see the chapter on shared decision making which details the Mental Capacity Act. |
- Maintaining hydration - Care of Dying Adults in the Last Days of LifeMaintaining hydration - Care of Dying Adults in the Last Days of Life
Your browsing activity is empty.
Activity recording is turned off.
See more...