Table 16, Summary of studies included in the review - Care of Dying Adults in the Last Days of Life

Study	Design	Population (n)	Research aim	Comments
Qualitative studies
Dendaas 2002²⁹	Individual interviews	n=15, nursing staff from hospices and inpatient oncology units who had recent experiencing of caring for multiple oncological patients in the last days of life. USA	To ascertain how experienced oncology nurses described the dying process of people with advanced cancer with relation to its length, recognisability using key signs and symptoms, and whether it is monitored.	A standard interview with open questions was developed but not all questions asked to all applicants. Reliable methods of analysis including external groups to transcribe the interviews and another group of experts to code the interviews were used.
Johnson 2003⁵¹	Individual interviews with set open questions	n=8 junior doctors with limited experience of intensive care medicine discussed their care of a person (the person had to be a current inpatient. If the person had either died or been discharged the discussions were excluded from analysis). USA	To discuss with junior doctors their recent experiences on prognosis and how they approach care decisions when caring for seriously ill hospitalised people.	The people discussed by the junior doctors were not necessarily recognised in the last days of life by the junior doctor, although they were prompted to consider whether they might die. There was no information provided on the time of death of the people discussed.
Van Der Werff 2012⁹⁹	Focus group design	n=18 nursing staff that had recent experiencing of caring for an oncological patient in the last days of life. Netherlands	To assess nurse perspectives on the signs and symptoms that suggest oncological patients are entering the last days of life.	A small study but good methodology on analysis. Study aimed to focus on oncological patients but this was not reflected in the example stem questions the facilitator of the focus group used.
Surveys
Abarshi 2011¹	Retrospective survey design with closed and open questions. Self-reported.	n=251. General practitioners who had looked after people in the last 3 months of their life (includes details of last 1 week of life). Netherlands	To explore the factors that allow primary care physicians to recognise that someone is entering the last days of life, and how this relates to care during this period.	Multivariate analysis undertaken, taking into account the person demographics. The study included an indirect population as all deaths over 1 years were included and were grouped together with younger adults (1-64 years) forming 20% of the study population.
Domeisen 2013³²	Delphi survey	n=252 Nurses, physicians, psycho-social-spiritual professionals, volunteers and carers from 9 participating countries: Switzerland, Italy, Netherlands, Sweden, Germany, UK, Argentina, New Zealand, and Slovenia.	To describe the most pertinent phenomena in identifying whether a person is in the last hours or days of life from any condition.	The population of experts for the Delphi included both the multiprofessional team and lay members but not for all rounds of the Delphi. The different rounds had different populations and it wasns't clear how these were formed.
Brandt 2005¹³	Prospective observational study	n=474 All long term nursing home care people assessed by physicians to be entering the last 6 weeks of life. Other inclusion criteria included admittance to nursing home for long term care or admitted for rehabilitation but during their stay it became obvious that the person would not leave the nursing home. Netherlands	To examine the dying person in nursing home settings, in particular the patient characteristics and signs that lead physicians to recognise entering the last 6 weeks of life. It also aims to look at the relationship between specific underlying disease and these symptoms with categories of cardiovascular disease, mental/behavioural disorders, and malignant neoplasms.	The doctors were asked to enter people into the study when they believed they were in the last 6 weeks of life. Even though this is indirect from the protocol looking at the last days of life, the majority of the people who were included had died by day 9 of the study.
Christakis 2000²²	Prospective survey design	n=468 people were discussed with 343 doctors who estimated their likely time of death. The actual date of death was then later collated from national records. USA	To investigate factors (such as optimism, and pessimistic and medical experience) that affect doctors' prognostication of people in outpatient hospice settings.	Included an indirect population- the median survival was 24 days of the people included.
Kumagai 2012⁵⁴	Delphi survey	n=72 community palliative nurses who had experience in looking after people who had died from either lung or gastrointestinal cancer. Japan	To identify predictors of the last 10 and 3 days of life in people with lung, gastric, or colorectal cancer at home.	The methods were well described with a good use of existing literature to formulate the initial Delphi. There was poor response rate in the study. The study only focuses on symptoms from 2 particular conditions.

Study

Design

Population (n)

Research aim

Comments

Qualitative studies

Dendaas 2002²⁹

Individual interviews

n=15, nursing staff from hospices and inpatient oncology units who had recent experiencing of caring for multiple oncological patients in the last days of life.

USA

To ascertain how experienced oncology nurses described the dying process of people with advanced cancer with relation to its length, recognisability using key signs and symptoms, and whether it is monitored.

A standard interview with open questions was developed but not all questions asked to all applicants. Reliable methods of analysis including external groups to transcribe the interviews and another group of experts to code the interviews were used.

Johnson 2003⁵¹

Individual interviews with set open questions

n=8 junior doctors with limited experience of intensive care medicine discussed their care of a person (the person had to be a current inpatient. If the person had either died or been discharged the discussions were excluded from analysis).
USA

To discuss with junior doctors their recent experiences on prognosis and how they approach care decisions when caring for seriously ill hospitalised people.

The people discussed by the junior doctors were not necessarily recognised in the last days of life by the junior doctor, although they were prompted to consider whether they might die. There was no information provided on the time of death of the people discussed.

Van Der Werff 2012⁹⁹

Focus group design

n=18 nursing staff that had recent experiencing of caring for an oncological patient in the last days of life.

Netherlands

To assess nurse perspectives on the signs and symptoms that suggest oncological patients are entering the last days of life.

A small study but good methodology on analysis. Study aimed to focus on oncological patients but this was not reflected in the example stem questions the facilitator of the focus group used.

Surveys

Abarshi 2011¹

Retrospective survey design with closed and open questions. Self-reported.

n=251. General practitioners who had looked after people in the last 3 months of their life (includes details of last 1 week of life).

Netherlands

To explore the factors that allow primary care physicians to recognise that someone is entering the last days of life, and how this relates to care during this period.

Multivariate analysis undertaken, taking into account the person demographics. The study included an indirect population as all deaths over 1 years were included and were grouped together with younger adults (1-64 years) forming 20% of the study population.

Domeisen 2013³²

Delphi survey

n=252 Nurses, physicians, psycho-social-spiritual professionals, volunteers and carers from 9 participating countries:
Switzerland, Italy, Netherlands, Sweden, Germany, UK, Argentina, New Zealand, and Slovenia.

To describe the most pertinent phenomena in identifying whether a person is in the last hours or days of life from any condition.

The population of experts for the Delphi included both the multiprofessional team and lay members but not for all rounds of the Delphi. The different rounds had different populations and it wasns't clear how these were formed.

Brandt 2005¹³

Prospective observational study

n=474 All long term nursing home care people assessed by physicians to be entering the last 6 weeks of life. Other inclusion criteria included admittance to nursing home for long term care or admitted for rehabilitation but during their stay it became obvious that the person would not leave the nursing home.

Netherlands

To examine the dying person in nursing home settings, in particular the patient characteristics and signs that lead physicians to recognise entering the last 6 weeks of life. It also aims to look at the relationship between specific underlying disease and these symptoms with categories of cardiovascular disease, mental/behavioural disorders, and malignant neoplasms.

The doctors were asked to enter people into the study when they believed they were in the last 6 weeks of life. Even though this is indirect from the protocol looking at the last days of life, the majority of the people who were included had died by day 9 of the study.

Christakis 2000²²

Prospective survey design

n=468 people were discussed with 343 doctors who estimated their likely time of death. The actual date of death was then later collated from national records.

USA

To investigate factors (such as optimism, and pessimistic and medical experience) that affect doctors' prognostication of people in outpatient hospice settings.

Included an indirect population- the median survival was 24 days of the people included.

Kumagai 2012⁵⁴

Delphi survey

n=72 community palliative nurses who had experience in looking after people who had died from either lung or gastrointestinal cancer.

Japan

To identify predictors of the last 10 and 3 days of life in people with lung, gastric, or colorectal cancer at home.

The methods were well described with a good use of existing literature to formulate the initial Delphi. There was poor response rate in the study. The study only focuses on symptoms from 2 particular conditions.

From: 5, Recognising when a person may be in the last days of life

Care of Dying Adults in the Last Days of Life.

NICE Guideline, No. 31.

National Clinical Guideline Centre (UK).

London: National Institute for Health and Care Excellence (NICE); 2015 Dec 16.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Table 16Summary of studies included in the review