Article Text

Download PDFPDF
Development of a generic working definition of ‘supportive care’
  1. Fiona Cramp1 and
  2. Michael I Bennett2
  1. 1Faculty of Health & Life Sciences, University of the West of England, Bristol BS16 1DD, UK
  2. 2Leeds Institute of Health Sciences, University of Leeds, Leeds LS2 9LJ, UK
  1. Correspondence to Dr Fiona Cramp, Faculty of Health & Life Sciences, University of the West of England, Glenside Campus, Blackberry Hill, Bristol BS16 1DD, UK; Fiona.cramp{at}


Context The term ‘supportive care’, despite everyday and widespread use, remains an ambiguous concept that lacks clarity. A generic definition would underpin the development of services and provide a basis for disease specific approaches to care.

Objectives To develop a generic working definition of supportive care that could be applied to a range of diseases and chronic illnesses.

Methods An indepth review of the literature was carried out to identify existing generic and disease specific definitions and descriptions of supportive care. Following the review, a modified two-phase Delphi study was performed. Participants included experts in the area of supportive care identified during the review in addition to charities or their representatives. A draft definition was developed based upon the outcome of the Delphi study.

Results The literature review identified one brief generic definition of supportive care with no explanation of its origin. A further 17 disease specific definitions were identified, all relating to cancer. The review led to the development of 100 statements for inclusion in phase I of the modified Delphi study. 26 responses were received to phase I of the Delphi study and 17 responses to phase II. General agreement was received during the second stage and a final draft definition subsequently developed.

Conclusions The resulting definition of supportive care might provide a basis for service development and could be further adapted to disease specific contexts. It would benefit from broader consultation to determine acceptability among a wider range of health professionals and service users.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Supportive care (SC) may be considered a general term that has been most closely associated with cancer patients.1,,3 However, conceptual confusion has arisen, with a number of terms being used interchangeably, such as SC, palliative care, end of life care, terminal care and conservative management.4 Although there is some suggestion that SC and palliative care overlap, there is a wide divergence of views about how and where they overlap.5 There remains a need to differentiate between these concepts to allow their development in clinical practice and research, and to develop an agreed international understanding of what SC entails. The National Council for Hospice and Specialist Palliative Care Services (NCHSPCS)5 proposed definitions for palliative care and SC (see box 1). However, the definition for SC was only intended to relate to cancer services.

Box 1 Definitions for palliative care and supportive care proposed by the National Council for Hospice and Specialist Palliative Care Services (NCHSPCS)

  • ‘Supportive care is that which helps the patient and their family to cope with cancer and treatment of it – from prediagnosis, through the process of diagnosis and treatment, to cure, continuing illness or death and into bereavement. It helps the patient to maximise the benefits of treatment and to live as well as possible with the effects of the disease. It is given equal priority alongside diagnosis and treatment’ (NCHSPCS, 2002).5

  • ‘Supportive Care in cancer is the prevention and management of the adverse effects of cancer and its treatment. This includes management of physical and psychological symptoms and side effects across the continuum of the cancer experience from diagnosis through anticancer treatment to post-treatment care. Enhancing rehabilitation, secondary cancer prevention, survivorship and end of life care are integral to Supportive Care’ (MASCC 2011).6

A further definition of SC specific to cancer has been provided by the Multinational Association of Supportive Care in Cancer (MASCC).

In 2004, the National Institute for Clinical Excellence (NICE) produced a comprehensive manual on ‘Improving Supportive and Palliative Care for Adults with Cancer’.7 The guidance provided a clear overview of SC specific to cancer and drew upon the NCHSPCS definition. Subsequent to this, NICE produced a guidance document on ‘Supportive and Palliative Care for Advanced Heart Failure’.8 It was clearly acknowledged that SC for heart failure had not been clearly defined and this had contributed to the poor provision of services and coordination of care for this patient group. Despite a lack of a clear definition of SC, a number of National Service Framework (NSF) documents have also referred to this concept, including the NSFs for older people, coronary heart disease, renal services and long term conditions.

The term ‘supportive care’, despite its everyday and wide use, remains an ambiguous concept that lacks clarity. It encompasses several disciplines and conditions with the meaning often varying. An agreed definition may help to reduce variations in patient care pathways by identifying those individuals who need support and ensuring they receive high quality, holistic care, regardless of diagnosis. The purpose of this research was therefore to develop an acceptable working definition of SC that could be applied within healthcare systems and by society in general to a range of diseases and chronic illness.


The research was undertaken in two stages: stage I was an indepth review of the literature and stage II was a modified Delphi study.

Stage I: indepth literature review

A literature search was initially undertaken to identify existing definitions, key concepts and themes related to SC. The search was restricted to the healthcare literature as it was likely that SC would have a different meaning in relation to social care. The search covered peer reviewed publications, textbooks and websites.

A secondary purpose of the literature review was to identify 30 key international experts in SC. An expert was defined as ‘an individual who has demonstrated knowledge about a subject and an ability to articulate a broad perspective on an issue’. Only those individuals able to communicate effectively in English were selected due to potential alterations in meaning during translation.

Search Strategy

The following databases were searched from 1999 to August 2009: The Cochrane Controlled Trials Register (Central/CCTR); MEDLINE; EMBASE; Cumulative Index to Nursing and Allied Health Literature; British Nursing Index; Allied and Complementary Medicine Database; Health Management Information Consortium; and PsycINFO.

Selected search fields were restricted to: title, abstract and subject headings in an attempt to identify papers with a main focus on SC.

The keywords ‘support$’ adj ‘care’ were used. Limits to the search included English language publications only and humans not animals. The reason for excluding non-English language papers was due to potential alterations in meaning during translation. The search was restricted by date to ensure that the findings were relevant to current healthcare policy.

The titles and abstracts were screened for relevance. Initially papers were selected that focused on descriptions of SC, developing SC criteria or definitions of SC. In addition, studies exploring patient and professionals views as well as studies investigating the effects or benefits of SC were selected. Papers that focused upon ‘developmental supportive care’ or ‘best supportive care’ were excluded as these terms tended to only include condition specific descriptions of care with no conceptual description. Only papers that added to the conceptual understanding of SC were ultimately retained for data extraction.

The world wide web was also searched in English using the keyword ‘supportive care’, which yielded several hundred ‘hits’. A rapid onscreen review was undertaken to exclude those that did not define SC or contribute to its conceptual understanding. By the 11th page (hits 101–110), no new information was retrieved. Journals incorporating the term SC within their title were identified through the web search. Editors and/or publishers of these journals were contacted to request an outline of their scope and any SC definition that they used. Additional sources such as publications by WHO, the National Council for Palliative Care, National Cancer Institute and the Department of Health were included to gain further perspective of the concept.

Data analysis

The sources obtained were examined and the content reviewed using the following key headings: supportive care generic definition, condition/disease specific definition of supportive care, supportive care descriptors/frameworks, reference to other potential sources of relevant information and relevant experts for inclusion in phase II.

All data recorded under the definition headings were gathered together to identify the key attributes of SC. Thematic analysis of the data was subsequently undertaken to develop a series of statements relating to SC which were taken forward to the second phase of the study.

Stage II: Modified Delphi Technique

Ethical approval for stage II was obtained from the Faculty Research Ethics Sub-Committee for Health and Life Sciences at the University of the West of England.

The Delphi method is a mixed qualitative–quantitative research tool that enables a panel of experts from a variety of settings to formulate a list of ideas, followed by the development of group consensus concerning the relative importance of those ideas.9 ,10 Delphi is one of the methods more commonly used in medical research to reach consensus and facilitates participation by experts and lay representatives.11 ,12

The objective of assembling the Delphi panel was to include individuals representing a diversity of experiences and backgrounds. Care was taken to ensure that a wide range of health professionals were included representing broad disease/condition profiles as well as professional backgrounds. Experts from a wide international geographical spread were also included. Experts were identified through relevant publications that were included in phase I of the research. Contact information was obtained from journal publications as well as via internet searches. Support groups and UK based charities were also identified for inclusion through the internet search.

The panel was invited to contribute to the study to gain consensus in relation to the development of the working definition of SC. A snowballing technique was also employed where panel members were asked to identify other experts that they felt were qualified to contribute. All communication took place via email. Discussion outside the Delphi process was discouraged and participants were not disclosed to each other as all emails were sent as ‘blind carbon copy’. An email was sent to the participants at the start of each phase. Two weeks later, a reminder email was sent to non-responders.

In the first round, members of the panel were asked to provide background information relating to their area of expertise. They were subsequently asked to rate their agreement with each of the statements developed in phase I using a 5-point Likert scale (1=strongly disagree to 5=strongly agree). All references to the source of the statement were removed to avoid bias. Next to each statement there was also an optional free text box for participants to provide any explanation for their selected score. A free text box was also provided following the final statement to enable participants to add any information that they felt was missing.

Following round one, the mean score for each statement was calculated. Where data were missing for a statement, the mean score was calculated from the existing responses. Any statement with a mean score <4 (agree) was removed from the second round. Scores for statements that were considered to be very similar were compared, the statement with the higher mean score was subsequently retained and the other statements discarded. Free text responses were also collated and where appropriate additional statements developed for inclusion in phase II.

A report was subsequently developed that summarised the statements that were not proposed for inclusion in the final definition. The retained and new statements were then arranged to form a proposed definition of SC. This was then sent to all participants (including those who did not respond in round one) for comment. Free text responses were then collated and used to inform the final draft definition.


Stage I: Indepth Literature Review

The keyword search resulted in 5867 hits (duplicates removed). This was reduced to 290 full text papers following a review of the titles and abstracts. The majority of excluded studies were randomised controlled trials describing a control arm as ‘best supportive care’. Prior to the review commencing it was repeatedly observed that these types of studies did not define best SC other than a description of current routine care. From the 290 full text articles retrieved, only 53 were found to contain a definition or description of SC. Table 1 indicates the patient populations to which the articles related. Further to this, six of the 53 studies considered paediatric populations with the remaining studies focused upon adults.

Table 1

Patient populations of included studies

No generic definitions of SC were identified from the literature search. Twenty-one papers provided disease specific definitions, all related to cancer. The authors of these papers derived their definitions from 15 different original sources suggesting that even within cancer there is no consensus regarding a definition of SC.

Additional information relating to descriptions but not definitions of SC was obtained from 45 of the 53 papers. The definitions and additional information were subsequently used to inform 90 statements. Some statements were deliberately very similar to determine preference for wording.

The web search revealed a further two definitions of SC relating to people with cancer. In addition, two sources of the same generic definition of SC were identified:

‘Supportive care is treatment given to prevent, control, or relieve complications and side effects and to improve the patient’s comfort and quality of life.'13 ,14

No details were provided regarding the source of this definition. The information obtained from the website search provided an additional 10 statements relating to SC. The final 100 statements were carried forward to stage II of the study (see table 2).

Table 2

SC statements and mean scores (presented in descending order)

The website search also identified the following journals incorporating the term ‘supportive care’ within their title:

  • Supportive Care in Cancer

  • Palliative and Supportive Care

  • Supportive and Palliative Cancer Care.

None of the journal websites provided a clear definition or description of SC and despite several attempts to contact the editors and publishers to obtain this information no details were obtained.

The email addresses were identified for 37 experts categorised as specialists in the following areas: 18 oncology, six nephrology/urology, four neurology, three cardiology, two respiratory, two rheumatology and two palliative care. They were based in the UK (n=19), the USA (n=8), Canada (n=6) and Australia (n=4). In addition, 96 charities were identified representing a wide range of diseases and chronic illnesses.

Stage II: Modified Delphi Technique

Phase I

The 100 statements developed during stage I were presented in a random order although following initial randomisation the list was checked to ensure that similar statements were not positioned adjacent to one another. The statements were then sent to the 37 experts and 96 charities for scoring and additional comment.

The 96 charities contacted resulted in 25 responses. However, 19 of the 25 reported that they were unable to complete the survey. The remaining 71 failed to respond despite an email reminder.

Six charities or their representatives provided completed responses to phase I (including one charity indentified through snowballing). These responses were considered in combination with the responses from the experts. One charity did not complete the survey but provided additional information relating to SC. This information was considered in the development of the documentation for consideration during phase II along with any other additional information that participants provided. Four additional charity contacts were identified for contact through the responses received. As indicated above, only one of the additional contacts subsequently responded to the survey. Eighty-one charity contacts remained for phase II.

In all, 18 (48.6%) of the 37 experts contacted provided valid responses. An additional respondent indicated that they were unable to respond at the time and provided contact details for suitable alternatives. In total, seven additional contacts were identified through snowballing and two of the seven subsequently responded. The final number of completed responses to phase I was 20 experts and six charities (n=26). Two respondents indicated that they specifically worked with paediatric populations. The clinical specialty areas and the professions represented by the 20 experts are presented in figures 1 and 2 respectively.

Figure 1

Clinical areas represented by experts (n=20).

Figure 2

Professions represented by experts (n=20).

Charity responses fell under the broad umbrella terms of oncology (n=2), neurology (n=3) and pain (n=1). Two charities represented paediatric populations.

The mean scores for each statement are presented in table 2. One statement had a mean score <2 (disagree) indicating that SC should not be described as ‘non-medical care’. Three statements scored between 2 (disagree) and 3 (unsure) indicating that the majority of respondents did not view SC to be the same as palliative care, comfort care or symptom management. A further 12 statements scored between 3 (unsure) and 4 (agree) suggesting that although some of the respondents agreed with the statement it was not strong agreement. The remaining 84 statements scored ≥4 suggesting that the majority of respondents were in agreement or strong agreement. It should be noted that this includes agreement to strong agreement with some negative statements indicating that respondents did not view SC to be the same as ‘no treatment’, conservative care or end of life care. However, additional comments indicated that end of life care may be a component of SC. There was also fairly strong agreement among respondents that SC does not aim to be curative. Additional comments indicated that some terminology did not provide clear meaning. For example, the statements ‘SC validates the illness experience’ and ‘SC should enable adaptation’ led to several queries relating to meaning. Several comments relating to financial support indicated that advice relating to financial planning should be provided as a component of SC but not monetary support.

Phase II

In total there were 17 responses in phase II, four charities and 13 experts. The charities represented neurological conditions (n=3) and pain (n=1). The clinical areas and professions represented by the experts are displayed in figures 3 and 4 respectively.

Figure 3

Clinical areas represented by experts (n=13).

Figure 4

Professions represented by experts (n=13).

Thirteen respondents agreed with the items that were proposed for exclusion from the final definition of SC. The remaining four respondents reported that complementary therapies should be included within the definition. One of the four suggested that non-pharmacological therapies included complementary therapies, implying that they would not need to be specifically mentioned. A further respondent suggested that they should be listed under the final section of ‘may include as needed’. The remaining two respondents did not provide any suggestion as to where complementary therapies should be included.

Based upon the research to date and the responses to the modified Delphi study the following was recommended as a generic working definition of SC:

‘Supportive care is the multi-disciplinary holistic care of patients with malignant and non-malignant chronic diseases and serious illness, and those that matter to them, to ensure the best possible quality of life. It extends as a right and necessity for all patients, is available throughout the course of the condition, concurrent to condition management and is given equal priority alongside diagnosis and treatment. It should be individualised, taking into account the patient’s past life experiences, their current situation and personal goals.'

The definition may also be accompanied by the following supplementary information to provide a more detailed account:

SC aims to:

  • Control the symptoms that occur as a result of the condition or its treatment and prevent complications thus allowing the individual to tolerate and benefit from active therapy more easily;

  • Meet a patient's spiritual, practical, physical, social, psychological, sexual and cultural needs;

  • Inform patient decision making and optimise patient understanding in relation to the illness and its treatment;

  • Enhance health professional–patient communication;

  • Improve general physical and mental health;

  • Optimise patient comfort and ease the physical burden of the condition thus in turn improving the ability to function and reducing the impact of disability;

  • Help the patient and their family cope with their illness and the treatment of it;

  • Empower the patient and their family as well as promoting self-help and user involvement thus enabling the individuals to draw upon their own strengths.

SC may include the following, as needed:

  • Issues of survivorship, palliation and bereavement;

  • Support groups;

  • Professional counselling and psychotherapy;

  • Rehabilitation;

  • Practical help;

  • Benefits advice;

  • Pharmacological and non-pharmacological interventions;

  • Nutritional support.

Thirteen respondents indicated that they agreed with the definition. The remaining four agreed in general but suggested minor changes. One individual suggested that spiritual issues were not relevant to healthcare. This had been highlighted by several respondents in phase I as having high importance and was therefore retained. Another respondent suggested that psychotherapy should be added alongside professional counselling; this was subsequently added. Another suggested that sexuality should be added to the definition; this was subsequently added. This respondent also suggested a change in wording, indicating that it may be impossible to meet patients' needs but to ‘assess and optimise needs’ would be more realistic. However, the stem for this statement is ‘Supportive care aims to…’ indicating that the aim is to meet patients needs even if they cannot always be met in full. The original wording was therefore retained. The final respondent agreed with the definition but queried the appropriateness of giving SC equal priority to diagnosis and treatment. This latter issue requires further input from service users before a definitive decision can be made.


This study has led to the development of a generic working definition of SC that is applicable to a wide range of chronic diseases and severe illnesses. There was general agreement with the definition and it now needs to be subject to further consultation prior to final acceptance.

The majority of SC research to date has been in the area of oncology. This was highlighted by the high proportion of papers relating to oncology that were retrieved as well as the range of SC definitions identified specific to cancer. It is clearly not appropriate to draw upon philosophical perspectives of SC in cancer and apply them directly to other patient groups. Further research is required to establish the SC needs of individuals with non-malignant disease and chronic illness. It is important to also determine what patients and their carers understand by the term SC.

The majority of respondents did not view SC to be the same as ‘palliative care’ ‘terminal care’, ‘conservative management’ or ‘end of life care’. These terms have previously been used interchangeably (eg,15), which is likely to have led to confusion among researchers, healthcare professionals and service users.4 Further work is recommended to ensure that the differences between these concepts are clearly articulated to inform the future design of services and the adequate provision of holistic care. The development of an accepted definition will help to underpin future research in this area.

There were a lot of similarities between the new definition and the cancer specific definition developed by NCHSPCS5 such as the holistic approach to patient care and the equal priority that SC should receive alongside diagnosis and treatment. The key differences were the lack of emphasis on cancer in the new definition as well as some broader perspectives such as the individualisation of SC and interdisciplinary links with health as well as social care professionals. In comparison with the generic definition of palliative care of the NCHSPCS5, there are again similarities including the holistic approach to care that extends to the patient's family. The main difference between the palliative care definition and the new SC definition was the lack of emphasis on the dying process.

Consistency in the definition is essential to facilitate further development of clinical services and interventions. It will also allow comparison of research findings and facilitate the generalisation of results to different populations. An accepted definition could also underpin future workforce planning including the identification of those aspects that should be the responsibility of the healthcare systems and those that are the responsibility of society in general. An agreed definition is also necessary for the development of psychometrically sound patient reported outcome measures. Consideration should be given to disease and condition specific definitions of SC based upon this generic definition.

There are several limitations of the study that should be acknowledged including the sampling method for the literature review. The exclusion of papers not published in English may have resulted in important work published in another language being missed. Further to this, the included charities were all UK based and may not have provided an international context. Other limitations include the high attrition and low response rates to the modified Delphi study, particularly the low response from charities and their representatives. The charities that provided reasons for non-response did not appear to understand the purpose of the research. Low response rates and high attrition are recognised as a potential risk of bias when using Delphi methods and could risk study validity.12 In hindsight, a telephone call to explain the research prior to sending the email may have enhanced the response rate from the charities. There was also a bias towards responses from individuals working in the clinical areas of oncology and nephrology. While there were attempts to ensure that all major clinical areas were represented, there was limited response from some areas including respiratory medicine and primary care which may limit generalisability. Further to this, the majority of respondents were medical or nursing professionals. Therefore, the final definition should be sent for wider consultation among a broad range of healthcare professionals working with a range of patient groups.

In summary, this research has provided clarity regarding the meaning and general understanding of SC which could be used as the basis for the development of disease specific definitions. It is recommended that the proposed definition is also subject to further consultation with a broad range of service users, healthcare professionals and publishers prior to international acceptance.


Funding for this work was provided by The Cochrane Pain, Palliative and Supportive Care Review Group. Philip O’Shaughnessy, Librarian, University of the West of England assisted with the retrieval of full text publications and Shan Aguilar-Stone proof read the questionnaires prior to distribution.


View Abstract


  • Funding Cochrane Pain, Palliative and Supportive Care Review Group.

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the Faculty Research Ethics Sub-Committee for Health and Life Sciences at the University of the West of England.

  • Provenance and peer review Not commissioned; externally peer reviewed.