Recognising and managing key transitions in end of life care
BMJ 2010; 341 doi: https://doi.org/10.1136/bmj.c4863 (Published 16 September 2010) Cite this as: BMJ 2010;341:c4863
This article has a correction. Please see:
- Kirsty Boyd, consultant in palliative medicine,
- Scott A Murray, St Columba’s Hospice professor of primary palliative care
- 1Primary Palliative Care Research Group, Centre for Population Health Sciences, University of Edinburgh, Edinburgh
- Correspondence to: K Boyd Kirsty.Boyd{at}luht.scot.nhs.uk
- Accepted 15 May 2010
Palliative care is being introduced earlier in the trajectory of illness, often in parallel with disease modifying treatment. A care pathway that starts with the identification of people approaching the end of life and initiating discussions about their preferences is central to the end of life care strategy in England.1 The Scottish government action plan also advocates a person centred approach based not on diagnosis or prognosis, but on the needs of patients and carers in all care settings—home, care home, and hospital.2 These needs include information about the illness and prognosis, symptom control, attention to psychological and spiritual concerns, continuity of care, and practical support. In view of the increasing numbers of people who could benefit, the emphasis of the UK strategies is on improving end of life care delivered by primary care teams, hospital staff, and social care services. Specialist palliative care should be available to people in any care setting who need additional expertise, and it serves most effectively as a resource to support ongoing care by other clinical teams.3
In economically developed countries, most people now die from one or more complex long term conditions.4 End of life care encompasses three overlapping phases of illness (figure)⇓. In this article we offer guidance about recognising end of life transitions. We also consider the challenge of changing the goals of care in patients with slowly progressive or fluctuating long term conditions.
Transition 1: would my patient benefit from supportive and palliative care?
Managing the transition to supportive and palliative care is arguably more of a challenge than identifying people who are in the last days of life.5 Doing so earlier can affect how, and potentially where, people die, but what constitutes “end of life care” is not uniformly understood and opinions vary as to who is a “palliative care” patient. Judging prognosis is particularly difficult for non-cancer patients.6 Identification of people with a life limiting illness when they are starting to need a change in their goals of care contributes to end of life care planning and can aid communication with patients and families. It depends on clinical judgment and weighing up a complex mix of pathology, clinical findings, therapeutic response, co-morbidities, psychosocial factors, and rate of decline.7 UK primary care teams are now expected to decide which patients should be included in their supportive and palliative care registers and when. We have reviewed two types of prognostic tools as the basis for a pragmatic approach to identifying candidates for palliative care needs assessment in primary and secondary care.
Disease specific prognostic tools use statistical models to predict the risks of individuals dying from conditions such as heart failure, chronic obstructive pulmonary disease, or liver disease. These tools tend to be used in clinical trials or when selecting patients for treatments like transplantation, but less often in end of life care. 8 9 10 Prognostic models were not found to be specific or sensitive enough when used to estimate survival of six months or less in older people with a range of non-cancerous illnesses.11 Such survival data have limited meaning for individual patients who are “sick enough to die”. In advanced heart failure, prognostic data suggested that an average patient had a 50% chance of living for six months on the day before their death.4
Performance status is strongly associated with survival time in patients with advanced illness, regardless of the diagnosis. This factor therefore forms the basis of the palliative performance scale, which is used in several countries to aid referral to hospice and specialist palliative care services.12 A similar tool, the palliative prognostic index, adds the symptoms of anorexia, breathlessness, and delirium to functional status.13 Such tools will identify most (though not all) patients who are likely to die within weeks, but are much less reliable for patients with supportive and palliative care needs who may still have 6-12 months to live.12 13
An alternative to prognostic tools is the use of criteria based on the clinical features of different advanced illnesses. The National Hospice and Palliative Care Organisation tool is used to decide eligibility for hospice care in the United States, where many services will only enrol patients with a prognosis of less than six months.4 These US clinical indicators were updated in 2001. They formed the basis of the prognostic indicator guidance tool that is used in the UK Gold Standards Framework for palliative care in the community.14 Both tools have good face validity and are widely used, but formal validation studies have been limited.
Using clinical indicators to identify patients for supportive and palliative care assessment
Our review of the prognostic models and guidelines leads us to propose a small group of readily identifiable indicators that can be used by professional carers in both primary and secondary care. Instead of seeking to refine prognostic accuracy, we propose that clinical judgment informed by evidence can improve care.
Box 1 describes how to identify patients for a supportive and palliative care assessment. If a patient has an advanced long term condition or a new diagnosis of a progressive, life limiting illness, or both, then ask the question, “Would you be surprised if this patient died in the next 6-12 months?” If the answer is no, look for one or more general clinical indicators that suggest this patient is at risk of dying and should be assessed for unmet needs. Some people who may benefit from supportive and palliative care have slowly progressive or fluctuating long term conditions. Concerns about deciding which of these patients should have additional assessment and structured end of life care planning are common, as are worries about discussing dying “too soon.” We suggest that a shortlist of disease related clinical indictors drawn from prognostic models and existing palliative prognostic guides be used to support clinical decision making.
Box 1 Supportive and palliative care indicators tool
(1) Ask
Does this patient have an advanced long term condition, a new diagnosis of a progressive life limiting illness, or both? (Yes)
Would you be surprised if this patient died in the next 6-12 months? (No)
(2) Look for one or more general clinical indicators
Performance status poor (limited self care; in bed or chair over 50% of the day) or deteriorating
Progressive weight loss (>10%) over the past 6 months
Two or more unplanned admissions in the past 6 months
Patient is in a nursing care home or continuing care unit, or needs more care at home
(3) Now look for two or more disease related indicators
Heart disease
NYHA class IV heart failure, severe valve disease, or extensive coronary artery disease
Breathless or chest pain at rest or on minimal exertion
Persistent symptoms despite optimal tolerated therapy
Renal impairment (eGFR <30 ml/min)
Systolic blood pressure <100 mm Hg and/or pulse rate >100
Cardiac cachexia
Two or more acute episodes needing intravenous treatment in past 6 months
Kidney disease
Stage 5 chronic kidney disease (eGFR<15 ml/min)
Conservative kidney management due to multi-morbidity
Deteriorating on renal replacement therapy; persistent symptoms and/or increasing dependency
Not starting dialysis following failure of a renal transplant
New life limiting condition or kidney failure as a complication of another condition or treatment
Respiratory disease
Severe airways obstruction (FEV1<30%) or restrictive deficit (vital capacity <60%, transfer factor <40%)
Meets criteria for long term oxygen therapy (PaO2 <7.3)
Breathless at rest or on minimal exertion between exacerbations
Persistent severe symptoms despite optimal tolerated therapy
Symptomatic heart failure
Body mass index <21
Increased emergency admissions for infective exacerbations and/or respiratory failure
Liver disease
Advanced cirrhosis with one or more complications: intractable ascites, hepatic encephalopathy, hepatorenal syndrome, bacterial peritonitis, recurrent variceal bleeds
Serum albumin <25 g/l, and prothrombin time raised or INR prolonged
Hepatocellular carcinoma
Cancer
Performance status deteriorating due to metastatic cancer and/or co-morbidities
Persistent symptoms despite optimal palliative oncology treatment or too frail for oncology treatment
Neurological disease
Progressive deterioration in physical and/or cognitive function despite optimal therapy
Symptoms that are complex and difficult to control
Speech problems; increasing difficulty communicating; progressive dysphagia
Recurrent aspiration pneumonia; breathless or respiratory failure
Dementia
Unable to dress, walk, or eat without assistance; unable to communicate meaningfully
Increasing eating problems; receiving pureed/soft diet or supplements or tube feeding
Recurrent febrile episodes or infections; aspiration pneumonia
Urinary and faecal incontinence
NYHA=New York Heart Association. eGFR=estimated glomerular filtration rate. FEV1=forced expiratory volume in 1 second. PaO2=pulmonary artery oxygen content. INR=international normalised ratio.
Rapid decline in the last weeks or months of life is often associated with progressive cancer, although other diseases sometimes follow this course and cancer can progress more slowly. Patients receiving palliative treatment for cancer may want to focus on fighting their illness, but supportive care, coordinated in primary care, should run in parallel with treatment. It should come to the fore as the patient starts to deteriorate and treatment, except for symptom control, is stopped.15 Advanced cancer at presentation or a poor performance status usually means that the patient would benefit from early supportive and palliative care in line with the general indicators in box 1.
A patient whose illness is associated with acute exacerbations followed by partial recovery may have been receiving health and social care for some time with the emphasis on optimal disease management, personalised care planning, and supported self management (see Resources). This situation is typically seen in those with heart failure, coronary artery disease, chronic lung disease, or end stage liver disease. Too much emphasis on prognostic accuracy in these fluctuating illnesses can hinder a positive focus on reasonable, patient centred goals at the end of life. 16 Variables identified in disease specific prognostic models are particularly useful as additional indicators in this group.
A prolonged, slow decline, sometimes punctuated with more acute episodes, is generally associated with multi-morbidity, advanced dementia, and progressive neurological diseases. . Such patients comprise the largest group in economically developed countries, and they typically need long periods of supportive and palliative care.4 17 The offer of early advance care planning is important because many patients will lose capacity to consent or express preferences about care.18 Many patients in this group can be identified from general clinical indicators, but additional triggers such as recurrent febrile episodes and eating problems suggest advanced cognitive and functional deterioration indicative of a substantial change in an otherwise gradual decline.19
Transition 2: Is my patient reaching the last days of life?
Appropriate use of clinical pathways such as the UK Liverpool Care Pathway for the Dying (see Resources) can help to optimise care in the last days of life, but a timely diagnosis of dying is essential. Patients on such pathways are reviewed regularly, medication is prescribed in line with good practice guidelines, and the holistic needs of the patient and family are addressed. Entry onto an end of life care pathway depends on clinicians being alert to the possibility that the patient may be dying and is based on clinical judgment after careful assessment. 20
Diagnosis of dying can be problematic for a range of reasons including a lack of continuity of care in the community and in hospitals. In the community, anticipatory care planning should ensure that sufficient care and support are in place to enable most patients who are expected to die soon to remain at home or in their care home. However, any potentially reversible causes of deterioration must be excluded in a patient who might still benefit from appropriate treatment. Such treatment should be started on the basis of clear, agreed goals, including a plan for review. Patients in hospital often continue to receive treatment of their underlying illnesses and complications until close to death. The decision to withdraw active treatment at the right time is important but will remain challenging if the outcome is uncertain and if the patient has recovered previously, particularly if earlier discussions about end of life preferences have not been possible.21
Using clinical indicators to identify patients in the last days of life
To improve the transition to terminal care, the care team should ask if a patient’s deterioration was expected, find out if the patient or a healthcare proxy wants further interventions, and exclude all potentially reversible causes (box 2).
Box 2 Clinical indicators for terminal care
Q1 Could this patient be in the last days of life?
Clinical indicators of dying may include:
Confined to bed or chair and unable to self care
Having difficulty taking oral fluids or not tolerating artificial feeding/hydration
No longer able to take oral medication
Increasingly drowsy
Q2 Was this patient’s condition expected to deteriorate in this way?
Q3 Is further life-prolonging treatment inappropriate?
Further treatment is likely to be ineffective or too burdensome.
Patient has refused further treatment.
Patient has made a valid advance decision to refuse treatment.
A healthcare proxy has refused further treatment on the patient’s behalf.
Q4 Have potentially reversible causes of deterioration been excluded?
These may include:
Infection (eg, urine, chest, cholangitis, peritonitis, neutropenia)
Dehydration
Biochemical disorder (calcium, sodium, blood sugar)
Drug toxicity (eg, opioids, sedatives, alcohol)
Intracranial event or head injury
Bleeding or severe anaemia
Hypoxia or respiratory failure
Acute renal impairment
Delirium
Severe constipation
Depression
If the diagnosis of dying is in doubt, give treatment and review within 24 hours.
If the answer to all four questions is “Yes”, plan care for a dying patient.
Conclusions
Primary care teams are well placed to use computerised disease registers and multidisciplinary review meetings to identify patients using pragmatic clinical criteria. Many more patients stand to benefit from better identification, assessment, and structured end of life care planning. Such improvements will enable professionals to address morbidity related to progressive disease and offer patients and their families opportunities to talk about living well with advanced illness.
Hospital specialists see many patients in the last year of life, often on multiple occasions, so can make an important contribution to identifying the need for additional supportive care, as well as optimising disease modifying treatments that will contribute to quality of life. Specialists can suggest when these patients may be suitable for supportive and palliative care in the community in discharge and outpatient letters, and primary care teams can ensure that such patients going to hospital are clearly identified.
The ability to make an accurate and timely diagnosis of dying is a core clinical skill based on careful assessment that could be done better in all care settings. Education and training of staff are central to the success of end of life policies in the UK.1 2
Resources
Palliative and end of life care
NHS Department of Health. National End of Life Care Programme. www.endoflifecareforadults.nhs.uk/eolc
Scottish Government Health Department: Living and Dying Well: A National Action Plan for Palliative and End of Life Care in Scotland. www.scotland.gov.uk/Topics/Health/NHS-Scotland/LivingandDyingWell
Gold Standards Framework. www.goldstandardsframework.nhs.uk
Gold Standards Framework Scotland. www.gsfs.scot.nhs.uk
Liverpool Care Pathway for the Dying Patient. www.endoflifecareforadults.nhs.uk/eolc/lcp.htm
Long term conditions
NHS Department of Health Long Term Conditions website. www.dh.gov.uk/en/Healthcare/Longtermconditions/index.htm
Cancer
National Institute for Clinical Excellence: Improving Supportive and Palliative Care for Adults with Cancer. 2003 www.nice.org.uk/guidance/index.jsp?action=download&r=true&o=28800
Heart Failure
NHS Heart Improvement Programme. Supportive and Palliative Care in Heart Failure—A Resource Kit for Cardiac Networks. 2006. www.heart.nhs.uk/endoflifecare
Scottish Partnership for Palliative Care. Living and Dying with Advanced Heart Failure: aPalliative Care Approach. 2008. www.palliativecarescotland.org.uk/publications/sppc-publications/living-and-dying-with-advanced-heart-failure
Chronic lung disease
National Institute for Clinical Excellence. Chronic Obstructive Pulmonary Disease. National Clinical Guideline on Management of Chronic Obstructive Pulmonary Disease in Adults in Primary and Secondary Care. 2004 http://guidance.nice.org.uk/CG12
Chronic kidney disease
NHS Department of Health. End of Life Care in Advanced Kidney Disease: A Framework for Implementation. www.endoflifecareforadults.nhs.uk/eolc/kidney.htm.
Dementia
NHS Department of Health. Living Well with Dementia: A National Dementia Strategy. 2009. www.endoflifecareforadults.nhs.uk/eolc/nds.htm
Discussing end of life transitions
Clayton JM, Hancock KM, Butow PN, Tattersall MHN, Currow DC. Clinical practice guidelines for communicating prognosis and end-of-life issues with adults and their caregivers. Med J Aust 2007;186:S77-108.
Notes
Cite this as: BMJ 2010;341:c4863
Footnotes
Part of the costs of producing the BMJ supplement in which this article appeared were met by the British Heart Foundation. The article was commissioned and peer reviewed according to the BMJ’s usual process.
Contributors and sources: This review was written by KB, a consultant in palliative medicine who has worked in hospital, community and hospice settings, in collaboration with SAM, leader of an international primary palliative care research group. We reviewed key policy documents, prognostic tools, and papers from international experts in care planning drawn from a 10 year Medline search and sought the views of colleagues in primary and secondary care. We are grateful for the opinions and papers contributed by W MacNee, P Reid (respiratory medicine); M Denvir (cardiology); P Cantley (geriatric medicine); M Young (general medicine); A Sheikh, E Paterson (general practice); F Downs, J Welsh (palliative medicine). KB is guarantor.
Competing interests: All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from commercial entities for the submitted work; no financial relationships with commercial entities that might have an interest in the submitted work in the past 3 years; no other relationships or activities that could appear to have influenced the submitted work.