End of life care, palliative care, supportive care, and terminal care are terms that many use interchangeably. There are, as yet, no internationally agreed definitions that explicitly describe the differences between each type of care, greatly hindering the accurate use and full understanding of these terms. For the purposes of this handbook, the following definitions are used.
End of life care
• An active and compassionate approach to care that ensures respect for and dignity of the patient and family.
• Partnership in care between patient, family, and health and social care professionals.
• Regular and systematic assessment of patient/carer needs incorporating patient consent at all times.
• Anticipation and management of deterioration in the patient's state of health and well-being.
• Advance care planning in accordance with patient preferences.
• Patient choice about place of care and death.
• Sensitivity to personal, cultural, and spiritual beliefs and practices.
• Effective coordination of care across all teams and providers of care.
There is no consensus as to the definition of the beginning of end of life care. Many consider that the end of life phase has started when a person with advanced, progressive disease could have a prognosis of less than one year. However, others relate the start of end of life care to when a patient develops palliative care needs, for example requiring control of symptoms and/or psychosocial support.
Palliative care and supportive care
Both terms are well represented by the definition of end of life care above. Arguably, the main difference between end of life care and palliative or supportive care is that the latter is provided from the time of diagnosis with incurable, progressive disease, irrespective of the length of prognosis, whereas end of life care has been arbitrarily defined as occurring within approximately one year of the time of death.
• Supportive care is an ‘umbrella term’ that encompasses a broad range of services that can help support patients, without being directed at treating the illness itself. This would include self-help, information giving, social support, and rehabilitation, and does also incorporate palliative care. Although palliative care practitioners would often provide such support, supportive care is arguably a broader construct than palliative care.
• Supportive care is given to any patient with a chronic disease, even if the disease does not impact on length of life. For example, patients with cancer who are treated and cured may still require supportive care.
• Palliative care has evolved into a distinct specialty, with associated specialist training. Supportive care, however, is like generalist palliative care, and its provision is the responsibility of all healthcare professionals.
Generalist palliative care
All healthcare professionals who look after patients with incurable, progressive disease are providing generalist palliative care. The vast majority of palliative care is provided by patients’ own healthcare teams, and does not need to be provided by palliative care specialists. It is a duty and a privilege for all healthcare professionals to be able to provide compassionate and effective care from diagnosis to death.
Specialist palliative care
Specialist palliative care became a medical specialty in the UK in 1987. Care is provided by specialist teams trained in the art and science of palliative care with extensive experience in providing symptom control and psychosocial/spiritual support for patients with complex needs. Referral criteria for specialist palliative care are given in Chapter 10 [link], and include care of patients with uncontrolled, intractable symptoms or complex psychosocial issues.
Specialist palliative care developed in response to the needs of patients with cancer, and in most countries care is still mainly provided for patients with malignant disease. However, there is increasing recognition that specialist palliative care must be provided to patients based on need rather than diagnosis.
Terminal care refers to the care that is provided to patients who only have a few days or hours to live (see Chapter 12).
It is not possible to provide high quality end of life care if it is not yet recognized that the end of life is, indeed, approaching. This section considers factors that can be used to help predict prognosis. Practical measures for determining and communicating prognosis are considered in detail in Chapter 8 [link]–[link].
Prediction of prognosis
Predicting prognosis is notoriously hard. Doctors tend to be over-optimistic in predicting prognosis. According to a large cohort study,1 doctors only make an accurate prognosis 20% of the time, and overestimate the prognosis of patients with incurable disease by a factor of five.
More experienced doctors tend to be more accurate. However, the better the doctor knows the patient and the more recent the last contact with the patient, the less accurate the prediction of prognosis. Anecdotally, patients prognosticate more accurately than healthcare professionals, and those that verbalize suspicion of a short prognosis are often correct.
• The cancer trajectory is shorter with relatively good functioning until the last few months. The COPD trajectory involves a lower level of functioning for a more prolonged time.
• Patients with cancer tend to deteriorate rapidly in the last few weeks or months of life and the prognosis tends to be relatively predictable. The prognosis of those with COPD is hard to predict, as death may occur during any one of the many acute exacerbations.
The prognosis of patients with malignant lung disease is very poor, usually measured in months from the time of diagnosis. The prediction of prognosis is made easier by this, and by the fact that the disease trajectory tends to involve an exponential decline. Initially deterioration in condition is seen month by month, then week by week and, in the terminal phase, day by day.
Prognostication is particularly challenging in patients with advanced non-malignant respiratory disease, because of the relatively lengthy deterioration and the fact that death occurs unpredictably during one of the many acute short-lived exacerbations of disease.
Many prognostication tools have been developed to try to integrate the various factors that impact on prognosis. In advanced COPD, the most commonly used is the BODE index, which generates a score out of 10, based on body mass index, degree of airflow obstruction, dyspnoea score, and exercise capacity. A score of 7 or more gives a 4-year survival of only 20%. However, in practice this index is not particularly useful in determining which patients are likely to die in the next year (see Table 4.1).
Table 4.1 BODE Index Score
Body mass index
FEV1 % predicted
MRC dyspnoea scale
6-minute walk distance
MRC dyspnoea scale
Breathless only with strenuous exercise
Short of breath when hurrying on the level or up a slight hill
Slower than most people of the same age on a level surface, or
Have to stop when walking at my own pace on the level
Stop for breath walking 100 metres, or
After walking a few minutes at my own pace on the level
Too breathless to leave the house
A number of other clinical indicators can alert professionals to a prognosis of less than a year. The degree of dyspnoea has been shown to be a more reliable indicator of both a poor prognosis and a willingness to discuss end of life issues than disease severity or pulmonary function:
• MRC grade 4/5 i.e. shortness of breath after 100m on the level or confined to the house because of breathlessness.
• More than three admissions in 12 months with COPD exacerbations.
• Disease assessed to be severe (e.g. FEV1 <30% predicted in patients with COPD).
• Fulfillment of long-term oxygen therapy (LTOT) criteria.
• Symptoms and signs of right heart failure.
• More than 6 weeks of corticosteroids in the preceding 12 months.
• Combination of other factors e.g. anorexia, greater than 10% weight loss over 6 months, depression, previous NIV/ITU, poor quality of life, increasingly housebound.
It appears that the intuitive integration of the many variables that affect prognosis by an experienced clinician is more accurate than scores from prognostic tools.
1. Christakis N, Lamond E (2000) Extent and determinants of error in doctors’ prognoses in terminally ill patients: prospective cohort study. Brit Med J; 320 (7233): 469–472.
‘Early recognition of people nearing the end of their life leads to earlier planning and better care.’
Gold Standards Framework, UK
End of life care has been defined on [link]. There is no consensus, however, on the definition of the beginning of end of life care. A definition that is too tight may mean that patients requiring end of life care will not receive it. Conversely, lack of a clear definition may be used as an excuse to ignore the fact that a patient may be approaching the end of their life.
Triggers for commencing end of life care
Any of the following scenarios suggest that a patient with incurable disease might need end of life care:
• It would not be surprising if the patient were to die in the next year.
• The patient has developed symptoms that cannot be controlled by disease-modifying treatment.
• The patient needs psychological or spiritual support because of the impact of the disease or limited prognosis.
• The patient has had to make a significant change in social circumstances because of progressive deterioration, such as moving to a care home or sheltered accommodation.
• The patient has had multiple recent admissions to hospital because of deteriorating health.
• The responsible health or social care team judges that the patient requires end of life care.
• The patient requests more intensive care because he/she perceives that the end of life is approaching.
• The patient's carers are experiencing high levels of distress and exhaustion.
A pragmatic approach would be as follows:
• Ask the ‘surprise question’ (‘Would I be surprised if this patient died in the next year?’) for all potential patients.
• If the answer is ‘no’, most patients should be receiving end of life care. The implications of this include the likely need for advance care planning, focusing on enhancing quality of life, basing care in the community, and so on.
• If the answer is ‘yes’, most patients may not yet be requiring end of life care. However, if a number of the other triggers listed above are true, they are likely to be in need of palliative or supportive care.
The terminal phase is defined as the last few days or hours of a patient's life. A patient in the terminal phase is in the process of dying. The terminal phase can be thought of as the period where even treatment of reversible causes of deterioration cannot change the outcome or its timing.
As the patient gets near to death the following features of dying may be seen:
• Increasing weakness and periods of sleeping, becoming semiconscious.
• Less keen or able to get out of bed.
• Less interested in or aware of things happening around him/her.
• Only able to take sips of fluid; unable to take tablets.
• Sometimes confused, occasionally agitated.
Some of these features will have been present before the terminal phase. Entry into the terminal phase is suggested when most of the above become evident at the same time.
The key challenge of recognizing entry into the terminal phase is that the same clinical features may be seen when a patient has deteriorated from a reversible cause, such as a chest infection. It is vital to do a clinical assessment, taking a history and examining the patient, to exclude reversible causes of decline, before coming to the conclusion that a patient is dying. It may even be necessary to undertake simple investigations. On occasion, a decision may be made not to treat a reversible cause of deterioration. If it is apparent in advance that such a decision will be made, there is clearly no need to search for the underlying cause.
The care of patients in the terminal phase is described in chapter 12.