Forum.
LITERATURE REVIEW
While there is an extensive literature about nutrition and fluid management in palliative patients, none of this relates specifically to sedated, imminently dying patients, and very little of it is original research. There are a large number of articles that reflect the opinion of the author(s)6,7-12, a handful of observational studies/surveys13-18 and some reports about individual patients who have lived for varying periods without food and/or fluids5-7,19,20. In contrast with the IRA prisoners, the patients documented in these reports, who refused both food and fluids in the medical setting, died peaceful deaths6,19-22. The general level of comfort, dignity, and absence of distress is notable. They lived for periods ranging from a week to several weeks.
When reviewing the evidence for use of artificial nutrition in cancer patients, Lichter found no evidence of improvement in treatment response rates, and concluded that “increased calorie intake by whatever route is unlikely to provide real benefit or alter the clinical outcome of cancer patients with weight loss”23.
A retrospective chart review in patients with motor neurone disease compared symptoms and length of life in patients fed via naso-gastric tubes with patients given only oral feeds24. The artificially fed group experienced more symptoms associated with oropharyngeal secretions and also experienced more hunger (the latter in some cases not until commencement of artificial feeding). The artificially-fed patients did not live any longer.
The provision of nutrition or hydration by artificial means is a medical intervention that carries with it the potential to cause harm through side effects or complications. The risks associated with placement of nasogastric or gastrostomy tubes and intravenous hyperalimentation are well documented in the literature25.
There is no evidence that artificial nutrition prolongs survival in patients with advanced cancer, and in view of the associated risks current accepted practice in many centres is to not routinely offer artificial nutrition to the dying patient. There is however a diversity of opinion about fluid administration.
The only systematic review of the literature on fluid status in the dying was published in 1997 by Viola et al26. From a search of the literature between 1966 and 1996, they identified original articles about fluid therapy or fluid status in dying or terminally ill patients. They found only six descriptive studies of low methodological quality, none of which specifically addressed the issue of fluid or nutrition in sedated, imminently dying patients. Viola et al concluded that there was insufficient evidence to draw firm conclusions regarding fluid management in palliative patients. They expressed the opinion that the only option is to assess each patient’s individual circumstances in order to arrive at a recommendation. The paper concluded with some fundamental questions that have yet to be answered, including, “What is ‘normal fluid status’ in the dying?” and “What impact does fluid status have on mental status near death?”
Ganzini et al recently published the results of a questionnaire completed by nurses working in hospices in the State of Oregon, USA, where Physician-Assisted Suicide is legal18. One hundred and two nurses reported that they had cared for a patient who had voluntarily and deliberately refused all food and fluids with the primary intention of hastening death. The authors concluded that on the basis of nurses’ reports, patients in hospice care who refuse food and fluids usually die a “good death” within two weeks. The authors noted several limitations to their study, including the possibility of recall bias and limitations in understanding the directions in the mailed questionnaire. The responses were based on nurse recall (sometimes up to four years after the event) rather than on chart review. Ganzini et al compared these results with those obtained in a similar questionnaire responded to by 55 nurses, relating to patients who elected to die by Physician-Assisted Suicide. While there was no difference in overall quality of death, suffering was rated lower and peacefulness higher in those patients who had stopped fluid and foods, than in those who had ingested lethal medication. Once again recall bias may have influenced these results.Several aspects of this report are unexpected, and merit special note: the large number of patients involved, the peacefulness of the deaths and the long period of survival without any fluids.
Many of the review and opinion papers discuss the perceived pros and/or cons of giving or withholding fluids6,7-10,21,27-29. It is clear that there is a range of views amongst those working in the field of palliative care, with some clinicians arguing strongly for7-9, and others against28, the use of fluids administered by artificial means.
Those arguing against the use of fluids in the dying patient who is unable or unwilling to drink have proposed that fluid depletion in the dying patient may be beneficial as it may result in
- a reduction in secretions (pulmonary, salivary, gastro-intestinal) with a consequent reduction in certain symptoms (eg cough, nausea) and less need for interventions to manage symptoms (eg suctioning)
- a reduction in urinary output, hence less incontinence and less need for indwelling urinary catheters
- less peri-tumour oedema with possible consequential pain reduction
- less oedema and ascites with fewer associated symptoms
A small observational study (9 patients) supports the suggestion that artificial hydration may exacerbate respiratory symptoms in the terminal stage and may not alleviate intravascular volume depletion30
One author has suggested, based on results from animal experiments, that starvation and dehydration might also have analgesic benefits produced by the generation of endorphins and the action of ketones21,31.
Proponents of withholding artificial fluid therapy suggest that dry mouth and thirst can be adequately managed with sips of fluid and good mouth care. Certainly there is nothing in the literature to refute this view, and some evidence to support it.
Several studies (three of which were included in the 1997 systematic review by Viola et al) suggest that thirst correlates poorly with fluid intake13,14,16,32,33.
A later study found no correlation between thirst in terminally ill cancer patients and traditional biochemical markers of acute dehydration, though low plasma levels of atrial natriuretic peptide (ANP) did correlate with severe thirst. The authors propose that ANP might be a superior marker of hydration status in this population group34
In their review Viola et al also report the suggestion that artificial therapy can result in increased emotional distress in families, divert the attention of the carer from the patient to the infusion, promote denial of the patient’s terminal status and interfere with physical interactions between patient and family26.
Advocates of fluid therapy have concerns about the symptoms that may arise in the dehydrated patient, such as delirium, or opioid toxicity, especially if renal failure develops8,9,35.
There is some evidence to support the view that fluid administration might improve delirium33. However, a more recent prospective randomised trial showed that administration of one litre of subcutaneous fluid daily was no more effective than medications alone in reducing delirium in dying patients, and no more effective than sips of fluid and mouth care in managing thirst32. The authors caution that “the small number of patients with delirium (15/42) included in this trial is insufficient for general conclusions”. One observational study found that while dehydration was associated with delirium reversibility, its association was not independent. This suggested that dehydration acts in association with other reversible factors such as opioid toxicity35.
Other concerns include the increased risk of constipation, pressure sores and dry mouth.
One paper reports on two patients who clearly benefited from rehydration in the terminal phase36. One had oesophageal cancer, the other gastric cancer. Because of the direct interference with swallowing and digestion, patients with gastric and oesophageal malignancies may experience a greater degree of dehydrationthan those with other malignancies, and therefore be more likely to benefit from rehydration during the pre-terminal phase
Several of the review and opinion articles discuss the ethical and/or legal implications of fluid and nutrition management in palliative patients37-40. Attention has been drawn to cultural differences, with a study in Taiwan reporting that there were ethical dilemmas regarding hydration and nutrition in 25% of patients39. In many countries there are legal precedents that regard the administration of artificial food and fluids as medical treatments. This has been the published view of the British Medical Association since 199243.
Some of the reviews point out that, as with all other medical treatments, there are indications and contra-indications to the use of artificial nutrition and hydration, which should be given or withdrawn/withheld in appropriate circumstances6,46,47.
This necessitates weighing the risks against the benefits of using artificial means to nourish or hydrate. As with other medical treatments, patient autonomy should be respected and when patients are unable to express their views, decision-making should take into account the opinions of those best able to represent the patient38,40,42,45.
Rabeneck points out in her discussion of percutaneous gastrostomy (PEG) tubes that the goal of care, as with other medical interventions, must be to benefit the patient. In the case of PEG insertion the expected benefit is to improve nutritional status, and if that goal cannot be achieved then the treatment should not be offered42. Once again, these principles are applicable to all forms of medical treatment.
McInerney suggests, from a sociological perspective, that it is the symbolic nature of food and feeding that is at the heart of the dilemma for the general public, and even for many members of the health professions, when faced with the question of administering food and fluids to patients with terminal cancer45. Denying food and water, even if it has been shown to be futile, would be viewed as cruel and inhuman. “Ritual and symbolism have an important role to play: their value, however, should be in the enrichment of existence, not as in the case of artificial feeding to many with advanced cancer, its impoverishment.”
There is an emerging consensus in the palliative care literature that, if artificial fluids are to be given, the subcutaneous route is preferred and adequate, and one litre per day is sufficient8,46,47.
Although there is a lack of good evidence on which to base decision-making, a number of recommendations have been published and these might help guide clinicians43,48. In 1997 the British National Council for Hospice and Specialist Palliative Care Services published a policy statement on artificial hydration for people who are terminally ill48. It covered the issues that decision-makers should take into account, and noted after review of the literature that the evidence, although limited, suggested “artificial hydration in imminently dying patients influences neither survival nor symptom control”.
The British Medical Association published a booklet that discusses in depth the issues surrounding the withholding or withdrawing of life-prolonging medical treatments, including the use of artificial hydration and nutrition43. Their recommendations take account of the Human Rights Act and European Convention on Human Rights. They state that the goal of medical treatment is to benefit the patient by restoring or maintaining health, and that this goal cannot be realised if the treatment fails to give a net benefit. Where there is doubt about a patient’s comfort the presumption should be in favour of providing relief from symptoms and distress.
Comments