End of life care for people with severe mental illness: the MENLOC evidence synthesis

Article history

The research reported in this issue of the journal was funded by the HSDR programme or one of its preceding programmes as project number 17/100/15. The contractual start date was in November 2018. The final report began editorial review in February 2020 and was accepted for publication in October 2020. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HSDR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Copyright statement

Copyright © 2022 Hannigan et al. This work was produced by Hannigan et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This is an Open Access publication distributed under the terms of the Creative Commons Attribution CC BY 4.0 licence, which permits unrestricted use, distribution, reproduction and adaption in any medium and for any purpose provided that it is properly attributed. See: https://creativecommons.org/licenses/by/4.0/. For attribution the title, original author(s), the publication source – NIHR Journals Library, and the DOI of the publication must be cited.

2022 Hannigan et al.

Burden of disease and costs

The overarching background for this project includes what is already known about the burden of SMI, cancer and end-stage lung, heart, renal or liver failure. Mental ill-health is a leading cause of years lived with disability around the world, with major depressive disorder the leading cause of years lived with disability in 56 countries and the second leading cause in a further 56. 5 Specific cancers, along with mental ill-health, neurological conditions, drug use disorders and specific organ diseases, all feature in the leading 20 causes of disability adjusted life-years in England for 2013. 6 The wider economic costs of mental illness in England were recently estimated at £105.2B each year. 7 This figure combines the direct costs of services, lost productivity at work and reduced quality of life, with the annual costs of the same in Wales estimated at £7.2B. 8

Meeting the physical health needs of people living with severe mental illness

The term ‘severe’ (or, often used interchangeably, ‘serious’, ‘serious and enduring’ or ‘serious and persistent’) mental illness, as used throughout this project, has longstanding currency in the fields of mental health policy, services and practice, dating back in the UK at least as far as the publication of Building Bridges: A Guide to Arrangements for Interagency Working for the Care and Protection of Severely Mentally Ill People. 9 It continues to be used in research10 and has currency with the NIHR Dissemination Centre, which published a themed review on SMI in 2018. 11 Building Bridges recognised the imprecision of the term ‘severe mental illness’ and endorsed a multidimensional framework definition encompassing five areas: safety, need for informal or formal care, disability, diagnosis and duration. Diagnosis is, therefore, an important, but not the only, dimension used in the identification of people with SMI and includes the International Classification of Diseases, Tenth Revision, diagnoses of schizophrenia, schizotypal and delusional disorders, bipolar affective disorder, and disorders of adult personality and behaviour,12 along with similar Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, diagnoses, including schizophrenia spectrum and other psychotic disorders, bipolar and related disorders, and major depressive disorder. 13

The particular focus in this project on EoL care can also be set in the context of people with SMI having high comorbidities2 and experiencing higher mortality rates and reduced life expectancy than the general population. 14 Higher mortality and morbidity rates for this group have been found across all age groups,15 with a 10- to 20-year reduction in life expectancy reported. 16 Inequities, not limited to care at the EoL specifically, can be explained with reference to individual- and system-level factors. People with SMI are less likely to attend health screenings and may respond to symptoms differently. 17 They may delay or avoid seeking help, and are more likely to exhibit disruptive behaviours or miss contacts with health professionals,3,18 putting them at risk of delayed disease detection. 2 Inadequate support systems are also common among those with SMI, affecting their ability to access appropriate clinical care and navigate complex health systems. 19 Other factors influencing variations in mortality and morbidity for people living with SMI include poor previous experiences of seeking help from health-care professionals (HCPs), care staff’s incorrect attribution of physical symptoms to psychiatric disorder and the lack of experience by mental health professionals in determining how and when to refer people onwards to other appropriate services. 3,20

End-of-life care for people with severe mental illness

In this study, following the definition used by the General Medical Council,21 EoL care is used to refer to the care of people who have diagnoses of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and are likely to die within the next 12 months. It includes care provided in hospitals, hospices and other institutional settings (e.g. care homes, prisons and hostels) and care provided in the home and via outreach to people who may also be homeless.

Beyond the inequities identified above, commitments to parity of esteem demand that EoL care for people with SMI should be as timely and as high quality as it is for others. However, early evidence gathered in support of this initial project proposal suggests that this group is poorly served, with England’s Cancer Strategy 2015–2022 recognising that people with SMI need to be the focus of improvements diagnosis and care. Although the incidence of cancer among people living with SMI is similar to that of the general population, mortality rates among those with SMI are double. 23,24 This disparity may be related in part to late presentation and reduced use of interventions such as surgery, chemotherapy or radiotherapy. 25,26 The experience of SMI can delay detection and treatment of life-threatening physical disorders as people are less likely to seek treatment, to verbalise pain and to access timely health care. 27 Consequently, this patient cohort is more likely to present with more advanced cancers that are invariably more complex and costly to treat, with patients less likely to undergo invasive treatments and more likely to die. 28 Some cancers, other terminal conditions and/or related treatments may also compound mental illness and precipitate potential ‘problematic’ behaviours. 29 For many patients, therefore, palliative care is often the only meaningful treatment option available.

The original case for this evidence synthesis also built on the observation that, once people with SMI are in touch with EoL services, their symptoms may be poorly recognised and undertreated, with staff working in EoL services lacking knowledge, training and experience in this area. 30 Undetected and hence untreated mental illness can jeopardise treatment outcomes, reduce patient satisfaction and increase health-care costs. 31 Variable adherence can be a complicating factor,19 compounded by comorbid disorders, such as substance misuse, and social factors, such as homelessness, isolation or lack of transportation, all of which can exert an impact on planning for EoL care and treatment. 3,15 Assumptions about the capacity of people with SMI to make EoL decisions, and concerns that EoL discussions would be too distressing or would exacerbate mental health problems, may lead to inadequate consultation and advance care planning. 32 The case for this project included reference to the fact that people with SMI have a higher percentage of ‘do not resuscitate’ orders than other groups and are less likely to have had discussions about their explicit wishes for EoL care. 33

Palliative care

Although not all people at the EoL need palliative care, a review of the evidence34 confirmed inequities in palliative care provision for both cancer and non-cancer patient populations. In Wales, an estimated 24,000 of the 32,000 people who die per year would benefit from palliative care, but over one-quarter do not have access to it. 34,35 Access to palliative care services among people who die from cancer is 46%, compared with 5% among those dying from other conditions,36 including end-stage lung, heart, renal or liver failure.

The case for this project recognised that people with mental illness and advanced incurable cancer and/or lung, heart, renal or liver failure face inequities and discrimination35 and a lack of integrated care. 2 Some may be excluded from EoL care planning3, and even from hospitals or hospices entirely. 29 They may be referred back to mental health services, where staff are often inadequately prepared to provide appropriate EoL care. 3 Difficulties accessing appropriate services mean that for people with mental illness and a life-threatening disease palliative care may be the first line of treatment. 2 The specific provision of palliative care for people with SMI is known to be poorer than for the general population. People with SMI are approximately 50% less likely to access appropriate palliative care, including symptom control and pain relief. 3,20 Palliative care and hospice staff often feel unskilled,20 lacking confidence and training in conducting discussions about EoL care with people SMI. 32 Evidence gathered to inform the proposal for this synthesis also indicates that there is a lack of co-ordinated EoL care, and that access to appropriate psychosocial support is often limited. 2,3 Medical and nursing staff working in hospices have also been shown to be unprepared for working with people with SMI, basing their assessments on instinct rather than using evidence-based approaches. 37 In a survey of psychological services in hospices in the UK and the Republic of Ireland, only 30% of hospices had access to a psychiatrist, 41% had access to a clinical psychologist and 45% had neither. 38 Patient experience data underscore these observations, with England’s National Cancer Patient Experience Survey showing that people with a long-term mental health condition (2% of those surveyed; n = 1184) reported less positive experiences of cancer care. 39 In Wales, the most recent National Cancer Patient Experience Survey40 found that the lowest proportion of respondents reporting positive experiences of their cancer care were those identifying as also having mental health problems.

Evidence explaining why this research was needed

Research at the interface of physical and mental health care is recognised as a UK priority. 41 Against the background presented above, this project has created generalisable knowledge to improve EoL care and services for an underserved group. Policies from the four UK governments focus on improving EoL care, where diagnosis is immaterial. 42–50 These policies require the introduction of palliative and supportive care earlier in the illness trajectory, with patient surveys showing that this is rated very highly by those receiving it. 51 The charity Marie Curie (London, UK) identified triggers that should initiate palliative care for people with diagnoses other than cancer,52 but, apart from dementia, it does not mention those with pre-existing SMI. In national policy the needs of people with SMI who develop advanced incurable cancer and/or end-stage lung, heart, renal or liver failure are acknowledged poorly, or not at all. This group face the prospect of ‘disadvantaged dying’53,54 at a time when quality of care in the last months of life should be uniformly high for all groups.

Uniquely among the cancer, palliative and EoL strategies developed across the four countries of the UK, the Independent Cancer Taskforce’s Achieving World-class Cancer Outcomes: A Strategy for England 2015–202022 makes the specific recommendation that the NIHR commission research in the area of cancer care for people living with SMI. This project has responded to this call, and expanded it to also cover EoL care for people living with SMI and facing end-stage lung, heart, renal or liver failure. It has answered a question that is both timely and relevant: what evidence is there relating to the organisation, provision and receipt of care for people with SMI who have an additional diagnosis of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and who are likely to die within the next 12 months?

This rigorously conducted evidence synthesis has brought together reports of approaches to service organisation, processes and interventions shown to both facilitate and hinder the provision of high-quality, accessible, equitable and acceptable EoL care to people with SMI. The project team has also gathered research and other evidence reporting the views and experiences of service users, families, and health and social care staff. Outputs from the project are intended to have an impact on services and practice, and plans are in place to present findings in accessible ways to NHS and other managers, practitioners and educators. It is anticipated that findings will inform future National Institute for Health and Care Excellence (NICE) guidelines, and thereby help shape the provision of services. Current relevant NICE guidance55–59 lacks standards or recommendations that particularly address EoL care for people with SMI.

A pre-project search of NIHR databases found studies that have investigated the physical health of people with SMI (e.g. Health Technology Assessment 12/28/0560), services for this group across organisations (e.g. HSDR 11/1023/1361) and care for people experiencing mental health difficulties after receiving cancer diagnosis (e.g. Health Technology Assessment 09/33/0262). As no research has been commissioned in the area addressed here, this project has effectively opened a new and important programme of work with value to the NHS and its partners. Having used a methodical, systematic and transparent approach, the MENLOC project team now plans to use this synthesis as the starting point for a programme of research that builds on what is already known, is designed with people who have experience of mental health difficulties and with people who have lived with cancer and have cared for family at the EoL, and generates new evidence of what works with value to the NHS and other relevant organisations.

Defining the project’s parameters: ‘severe mental illness’ and ‘end of life’ (and related)

At the first, key, combined project team and SAG meeting, candidate database search strategies and search terms developed by the project team for both SMI (see Appendix 6) and EoL care (see Appendix 7) were distributed and discussed at length with the purpose of refinement.

Finding grey literature, including non-research material

Websites to search that had already been identified by the project team were circulated (see Appendix 8), and in this first, key, meeting SAG members were invited to identify additional online sites. Suggestions made in the meeting or received thereafter included a Hospice UK (London, UK) report (KC), deaths in custody (SP), palliative care in prisons (NP) and searches on institutional repositories.

Aims and objectives

The aim of this project was to synthesise relevant research and other appropriate evidence relating to the organisation, provision and receipt of EoL care for people with SMI (including schizophrenia, bipolar disorder and other psychoses, major depression and personality disorder). Specifically, it answered the following question: what evidence is there relating to the organisation, provision and receipt of care for people with SMI who have an additional diagnosis of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and who are likely to die within the next 12 months?

The specific objectives were to:

1. use internationally recognised, transparent literature review approaches to locate, appraise and synthesise the relevant research evidence relating to the organisation, provision and receipt of care in the expected last year of life (LYoL) for people with SMI who have additional diagnoses of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and who are likely to die within the next 12 months

2. locate and synthesise policy, guidance, case reports and other grey and non-research literature relating to the organisation, provision and receipt of care in the expected LYoL for people with SMI who have additional diagnoses of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and who are likely to die within the next 12 months

3. produce outputs with clear implications for service commissioning, organisation and provision

4. make recommendations for future research designed to inform service improvements, guidance and policy.

Inclusion criteria

Developing the search strategy

The focus of the search strategy was to achieve high sensitivity and specificity for retrieving studies relevant to the review question. The search strategy was comprehensive and designed to ensure that all relevant literature was obtained, with the precise emphasis on concepts associated with the research question discussed in the first combined project team and SAG meeting, described in Chapter 2. Although some terminology remains equivocal in this area, the search strategy as it became finalised was designed to identify all relevant evidence relating to EoL care (i.e. in the LYoL) in those with SMI. Reflecting the importance of diagnosis in the framework definition of SMI,9 diagnostic terms included in the search strategy included schizophrenia, schizophrenia spectrum and other psychotic disorders, schizotypal and delusional disorders, bipolar affective disorder, bipolar and related disorders, major depressive disorder and disorders of adult personality and behaviour. Reflecting prevailing definitions of SMI (e.g. as used by the NIHR Dissemination Centre11), searches were not made for studies into mental and behavioural disorders due to psychoactive substance use or for studies into dementia or other neurodegenerative diseases, and items in these areas were not included in the review, except where participants’ diagnoses co-existed with the disorders included above.

Preliminary searching

Preliminary database searching using MEDLINE was carried out as part of initial scoping undertaken in preparation of the proposal for funding, with material from this drawn on in Chapter 1. The preliminary keywords that were used to inform these searches included the following:

Palliative care OR Hospice care OR Terminal care OR Terminally ill OR End of life care OR Last year of life

AND

Neoplasms OR Cancer OR heart failure, lung failure, liver failure or renal failure

AND

Mental health OR Depression OR Mental disorders OR Depressive disorder OR Personality disorders OR Bipolar disorder OR Schizophrenia OR Mental illness

An analysis of the text words contained in the titles and abstracts, and of the index terms used to describe each article, was undertaken. As a result of this, the search strategy was further developed, and in July 2018 the opportunity was undertaken to update and extend the preliminary scoping exercise across four databases. These were, on Ovid, MEDLINE^® ALL, MEDLINE Epub Ahead of Print, In-Process and Other Non-Indexed Citations, PsycInfo and EMBASE. The MEDLINE search can be found in Appendix 9.

Comprehensive searching

A comprehensive search strategy was developed by Mala Mann following discussion at the first combined project team and SAG meeting (see Chapter 2). This search strategy was initially developed for MEDLINE (see Appendix 10) and PsycInfo (see Appendix 11). As a means of testing and refining this search strategy before applying it across multiple databases, Deborah Edwards first screened the citations retrieved across MEDLINE and PsycInfo to ensure relevance and to assess that the strategy was neither too broad nor too narrow. Once the project team was satisfied with the search strategy, this was tailored to the remaining eight databases, with searches run from each database’s inception. Databases searched included:

• on the Ovid platform – MEDLINE ALL, EMBASE, HMIC (Health Management Information Consortium), PsycInfo and AMED (Allied and Complementary Medicine Database)

• on the EBSCOhost platform – CINAHL (Cumulative Index to Nursing and Allied Health Literature) and CENTRAL (Cochrane Central Register of Controlled Trials)

• on the ProQuest platform – ASSIA (Applied Social Sciences Index and Abstracts)

• others – DARE (Database of Abstracts of Reviews of Effects) and the Web of Science.

Supplementary searches were undertaken to identify additional papers, information on studies in progress, unpublished research, research reported in the grey literature and personal blogs. Members of the SAG advised the project team as to which relevant websites to search (see Chapter 2), and a full list of websites searched along with the search terms utilised can be found in Appendix 12. Members of the SAG were also asked to inform the research team of any other publications of which they were aware that they thought might be relevant to the review. The principal investigator (BH) and the project manager (DE) followed a number of key EoL and mental health authors and organisations on Twitter (Twitter, Inc., San Francisco, CA, USA) to identify any material that could be potentially relevant to the evidence synthesis.

Searches were also conducted using Google (Google Inc., Mountain View, CA, USA), as described by Mahood et al. 70 The first 10 pages of each Google output were screened using the following terms:

• “palliative care” and “mental illness”

• “end of life” and “mental illness”

• “end of life” and schizophrenia (searching the first five pages of output)

• “end of life” and bipolar (searching the first five pages of output).

To identify published resources that had not yet been catalogued in electronic databases, recent editions of the Journal of Pain and Symptom Management, Cancer, Psycho-Oncology and BMJ Supportive & Palliative Care were hand searched. These journals were selected because a large number of outputs identified in database searches had been published in them. Reference lists of included studies were scanned and forward citation tracking was performed using the Web of Science.

Primary research citations retrieved from database searches

All citations retrieved from the 10 database searches were imported or entered manually into EndNote™ (Thomson Reuters, CA, USA) and duplicates were removed. The total numbers of hits retrieved for each database are displayed in Table 1.

Primary research citations identified from supplementary searching

All primary research citations identified as potentially relevant from across all supplementary sources (Table 2) were entered manually into EndNote. A total of 128 citations were identified.

Removing irrelevant citations

Once all of the primary research citations had been imported into EndNote, irrelevant citations were removed by searching for keywords in the title using the search feature of the EndNote software. The project team had previously agreed which keywords to use to identify papers that did not meet the review’s inclusion criteria. All hits for each keyword were screened by Deborah Edwards to ensure that they were, in fact, irrelevant before they were removed. The following keywords were used:

• Abuse/substance

• Adolescence/adolescent/youth

• AIDS/HIV

• Alcohol

• Anxiety

• Assessment

• Bereav*

• Bipolar diathermy/electrocautery

• Book review

• Child

• Dementia/Alzheimer’s

• Depression

• Diabetes

• Diagnos*

• Fatigue

• Grief

• Incidence

• Instrument/tool

• Letter

• Neonate/neonatal

• Parent

• Predict*

• Prevalence

• Prognosis/prognostic

• Psychometric

• Scale

• Screen*

• Sleep/insomnia

• Smoking/tobacco

• Spiritual/religious

• Survival

• Thesis

• Validity/reliability/validation

• Version.

At the end of this process, the remaining citations were exported as an XML file and then imported into the software package Covidence™ (Melbourne, VIC, Australia).

Title and abstract screening

All citations were independently assessed for relevance by two members of the review team using the information provided in the title and abstract using Covidence. Disagreements were resolved through discussion with a third reviewer, and, where needed for further clarity, the full text was retrieved to aid this discussion. Particular care was taken with citations referencing people experiencing depression. Depression is a commonly experienced mental health problem, but it is also one that can manifest as a severe mental health problem, requiring an urgent and potentially life-saving intervention. As our focus was on EoL care for people with pre-existing SMI, we excluded citations reporting on prior experiences of mild depression, and excluded items reporting on the identification of depression and its treatment among people who had experienced psychological distress subsequent to their EoL diagnosis. Only citations clearly reporting on EoL care for people who had pre-existing severe depression were included.

For citations that appeared to meet the inclusion criteria, or in cases in which a definite decision could not be made based on the title and/or abstract alone, the full texts of all citations that appeared to meet the review’s inclusion criteria by title and abstract screening were retrieved.

Full-text screening

To achieve a high level of consistency, two reviewers screened each retrieved citation for inclusion using a purposely designed form (see Appendix 13). Disagreements were resolved through discussion with a third reviewer. All English-language items relating to the provision of EoL care to and the receipt of EoL care by people with SMI and an additional diagnosis of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure were included at this stage. The flow of citations through each stage of the review process is displayed in the PRISMA flow chart in Figure 1. 69

FIGURE 1.

The PRISMA flow chart. Reproduced with permission from Edwards et al. 71 © The Author(s) 2021 This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).

Reproduced with permission from Edwards et al. 71 © The Author(s) 2021 This article is distributed under the terms of the Creative Commons Attribution 4.0 License (https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage)

Grey literature identified from supplementary searching

Sixty-eight grey literature citations were identified as potentially relevant from across all supplementary searches (Table 3), and these were all entered manually into EndNote. Policies and guidance, reports and other non-research materials found in the grey literature were also read by two members of the project team and considered against the topic inclusion criteria, with disagreements resolved as above. Forty were excluded (see Appendix 15), leaving a total of 28 that were assessed as relevant to the review (see Appendix 16). A test search for relevant information in renal-specific documents produced no material suitable to be included in the review. The decision was therefore made by the project team not to search within (the large number of) disease-specific policy and guidance documents for material concerned with the management of EoL care for people also with SMI.

Materials included in evidence synthesis

One hundred and four publications were included in the evidence synthesis, which consisted of research publications (n = 34), case studies (n = 42) and non-research policy and guidance material (n = 28).

Terminology and language use

Language is important (and sometimes contentious), and, as described in Chapter 2, Defining the project’s parameters: ‘severe mental illness’ and ‘end of life’ (and related), the variety of terms used in the mental health and EoL fields and in the materials included in this synthesis were recurring items at combined project team and SAG meetings. Across the items located and included in the synthesis, no agreed definitions could be discerned relating to the precise meanings of SMI or the content of EoL care or palliative care. Specialist palliative care could refer, particularly in the UK context, to care provided in any setting (hospice, home, care home or hospital) by a dedicated specialist team with additional education and skills in palliative care that enables them to support those at the EoL with any diagnosis who have complex, high-level needs, whatever the cause. Palliative care, however, could refer to care provided by all HCPs to those who are at the EoL (across all care settings) and do not have the additional complex needs that require additional specialist support. This clinical distinction may be similar across many countries with similarly developed health systems (e.g. the USA, Australia and Canada), but, in each case, the project team, having finalised the materials to include, noted the existence of different funding and organisational arrangements and varying standards for eligibility to access services. Discussion also took place between project team and SAG members on the language to be used in this report to describe people with SMI at the EoL. In recognition of people’s receipt of (or need for) integrated care crossing mental health and palliative care system boundaries, a decision was made at the third combined project team and SAG meeting to use the word ‘patient’.

Quality appraisal

Following searching and screening, information from research and case study publications was assessed independently for methodological quality by two reviewers using one of a number of agreed appraisal checklists. Alternative tools, which reflected the specific design and methods used in individual research outputs, were used as necessary when suitable Critical Appraisal Skills Programme (CASP) tools were not available. Any disagreement on quality was resolved through discussion with a third reviewer. No research items were judged to have been fatally flawed, and all were included. The policy and guidance documents and non-research reports were not subject to quality appraisal.

For qualitative studies, the appropriate checklist available from CASP was used. 72 For cross-sectional designs, the Specialist Unit for Review Evidence 12-item checklist73 (with responses ‘yes’, ‘no’, ‘can’t tell’) was used and an overall score reflecting the number of items answered ‘yes’ was generated.

Retrospective cohort studies were appraised using the Scottish Intercollegiate Guidelines Network’s Methodology Checklist 3: Cohort Studies. 74 This is a 14-item checklist (with responses ‘yes’, ‘no’, ‘can’t say’ and ‘does not apply’). Six items do not apply to this type of study design (statements 1.3, 1,4, 1,5, 1.6, 1.8 and 1.12). The overall assessment reflects how well a study has sought to minimise the risk of bias or confounders. The final rating is high quality (++), acceptable (+) or low quality (–). Reproduced from Scottish Intercollegiate Guidelines Network with permission. 74 These guidelines are licensed under the Creative Commons Attribution-Noncommercial-NoDerivatives 4.0 International Licence. This allows for the copy and redistribution of SIGN guidelines as long as SIGN is fully acknowledged and given credit. The material must not be remixed, transformed or built upon in any way. To view a copy of this licence, visit https://creativecommons.org/licenses/by-nc-nd/4.0/

As retrospective designs are generally regarded as weaker, the authors of the checklist suggest that these should not receive a rating higher than ‘+’.

For case studies, the Joanna Briggs Institute critical appraisal checklist for case reports was used. 75 This is an eight-item checklist (with responses ‘yes’, ‘no’, ‘unclear’, ‘not applicable’) from which an overall score is generated reflecting the number of items answered ‘yes’.

Data extraction

All demographic data from the 76 research and case study publications were extracted directly into tables based on study design (see Appendices 17–20), following the format recommended by the Centre for Reviews and Dissemination. 68 The work was allocated to a member of the research team who extracted the relevant data, and a second reviewer (DE) independently checked the data extraction forms for accuracy and completeness. A record of corrections was kept. Where multiple publications from the same study were identified, data were extracted and reported as if from a single study.

Data analysis and synthesis

Following discussion at the second of the three combined project team and SAG group meetings, the decision was made to conduct a first analysis and synthesis for the combined research studies (n = 34) and policy and guidance documents (n = 28), and a second analysis and synthesis for the case studies (n = 42). The project protocol stated an intention to perform meta-analyses of data for included intervention studies. However, no intervention studies were found. All synthesised findings are therefore presented as narrative syntheses in Chapters 4 and 5.

All research papers were available in full-text form, and each was uploaded to a project created using the software package NVivo 12 (QSR International, Warrington, UK). All policy and guidance documents were available as electronic documents, and the content of each was searched using keywords relevant to the review (‘mental illness’ and ‘end of life’). Data retrieved in this way were extracted and entered verbatim into Microsoft Word (Microsoft 365 Apps for Enterprise; Microsoft Corporation, Redmond, WA, USA) documents, one for each policy and guidance item. Microsoft Word files containing the extracted data were then uploaded to the same NVivo project as the research items in preparation for analysis. All case studies were available in full-text form and were uploaded to a second NVivo project.

A thematic approach was employed to analyse and synthesise both sets of data,76 with work on the combined research and policy data set led jointly by Ben Hannigan and Deborah Edwards and work on the case study data set led by Michael Coffey. Using NVivo, in both cases inductive data-driven codes were generated and attached to segments of material through the line-by-line reading of documents. To aid discussion and refinement, and to support a visual appreciation of the overall character of the developing research and policy synthesis, codes and their relationships to each other were also reproduced on large sheets of display paper. In both thematic syntheses, codes were grouped into meaningful candidate themes (and, in the case of the research and policy synthesis, subthemes) reflecting the material being synthesised and the overarching objectives of the project. 4 These were shared with all members of the project team for review and further refinement, with no major disagreements noted. Themes and subthemes were then tabled at the third combined project team and SAG meeting, where the overarching thematic structure of both syntheses was discussed and approved. Both syntheses aimed to bring together reports of approaches to service organisation, processes and interventions shown to both facilitate and hinder the provision of high-quality, accessible, equitable and acceptable EoL care to people with SMI. The reporting of the separate analysis and synthesis of the combined research, policy and guidance items is contained in Chapter 4, and the reporting of the analysis and synthesis of the case study items is contained in Chapter 5. Figure 6 in Chapter 6 provides a visual overview of both thematic syntheses as a whole.

Assessing confidence

The assessment of the quality of evidence for retrospective cohort studies was conducted using the GRADE (Grading of Recommendations, Assessment, Development and Evaluation) approach,77 and was conducted by Deborah Edwards and checked by Ben Hannigan. This approach rates the quality of a body of evidence as ‘high’, ‘moderate’, ‘low’ or ‘very low’ (Figure 2). 77 Explicit criteria, including study design, risk of bias, impression, inconsistency, indirectiveness and magnitude of effect, are used. As the studies retrieved for this evidence synthesis were observational as opposed to interventional, the initial quality of the body of evidence overall started off as low. 78 When the evidence base was rated specifically on study design (in particular retrospective cohort studies), this led to the ratings for all evidence generated using material from these types of study being downgraded from ‘low quality’ to ‘very low quality’.

FIGURE 2.

The GRADE quality of evidence and definitions. Reproduced from GRADE: an emerging consensus on rating quality of evidence and strength of recommendations, Guyatt G, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. , volume 336, pp. 924–6, 2008, with permission from BMJ Publishing Group Ltd. 77

Reproduced from ‘GRADE: an emerging consensus on rating quality of evidence and strength of recommendations’, Guyatt G, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. , volume 336, pp. 924–6, 2008, with permission from BMJ Publishing Group Ltd. 77

The strength of the synthesised qualitative and non-intervention findings was assessed using the CERQual (Confidence in the Evidence from Reviews of Qualitative Research) approach,79 and this was again led by Deborah Edwards and Ben Hannigan. The original CERQual approach was designed for qualitative findings, but in this project a process previously used by members of the research team in HSDR 11/1024/0880 and in HSDR 08/1704/21181 was used. This approach involved using CERQual for the additional purpose of assessing confidence in findings synthesised from surveys and other non-intervention quantitative studies. The confidence in individual synthesised review findings is based on the assessment of four components: coherence, methodological limitations, relevance and adequacy. Overall assessments of confidence are then described using four levels – high, moderate, low or very low – as summarised in Figure 3, which draws on and adapts material previously published. 82

FIGURE 3.

The CERQual: applying high, moderate, and low confidence to evidence. Reproduced with permission from Lewin et al. 82 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (https://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. The figure includes minor additions and formatting changes to the original figure.

Reproduced with permission from Lewin et al. 82 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (https://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. The figure includes minor additions and formatting changes to the original figure.

Description of research studies

Thirty-four publications covering 30 research studies were deemed suitable for inclusion in the review: 19 quantitative studies (reported across 21 publications), nine qualitative studies (reported across 11 publications) and two mixed-methods studies. Demographic information on the characteristics of included research studies is given in Appendices 17–20.

Year of publication

Figure 4 displays the year of publication of the 34 research study publications.

FIGURE 4.

Histogram of year of publication of research studies.

Description of case studies

Forty-two publications consisted of 51 case studies of individuals who had an existing mental illness diagnosis and who went on to develop an EoL condition (see Appendix 20). In four case studies the purpose of the paper was to show the application of a particular model of care, such as dynamic system analysis115 or stepwise psychosocial palliative care. 116,117 Most of the case studies were published in peer-reviewed research journals (n = 38), along with two conference abstracts,118,119 one that appeared in a report120 and one that was in a first-person-account blog. 121 The case studies ranged in depth from discursive papers with little direct detail about the individuals involved, to those that focused mainly on the patient’s or patients’ physical illness and a substantive number that discussed issues of mental capacity and the ability of the person to express their choice to accept, refuse or abandon treatments.

Year of publication

The earliest case study was published in 1983133 and outputs appeared regularly up until the end of the search period in 2019. 118,119,150 Figure 5 displays the year of publication of the 42 research study publications.

FIGURE 5.

Histogram of year of publication of case studies.

Results of quality appraisal

Structure of the system

Twenty-six policy and guidance documents22,30,55,120,155–176 and 18 research outputs32,85,88,91–94,97–99,103,105,106,108,110,112–114 contribute to this theme, which addresses the broad shape and structure of the EoL and mental health care systems. Material is organised under four subthemes: policy and guidance; separate commissioning, management and organisation; partnership; and no right place to die.

Professional issues

Ten policy and guidance documents30,156,162–166,171,176,177 and 16 research outputs (reporting 14 studies)32,37,85,93,94,99,102–106,110–114 contribute to this theme, which addresses professional and practitioner issues. Material is organised under four subthemes: relationships between HCPs and people with SMI, mental health professionals doing EoL care, EoL care professionals doing mental health care, and training and education.

Contexts of care

Eleven policy and guidance documents21,30,156,161–166,170,176 and 31 research outputs32,37,83–93,95–102,104–110,112–114 (reporting 28 studies) contribute to this theme, which addresses the organisation, provision and receipt of care. Material is organised under three subthemes: managing the interface between mental health and EoL care; health-care services and treatment utilisation in the LYoL, and meeting individual and family needs.

Living with severe mental illness

Eight policy and guidance documents22,30,156–158,161,162,170 and 11 research outputs32,37,85,99,102,103,106,110,112–114 (reporting 10 studies) contribute to this theme, which addresses the individual and social characteristics of patients with SMI. Material is organised under three subthemes: complexities of EoL care, familiarity and trust, and recognising physical decline.

Diagnostic delay and overshadowing

Although the case studies ranged across many countries, they nevertheless show a consistent picture of delayed or late diagnosis and what might be termed diagnostic overshadowing. Complex and enduring mental health problems are often described as contributing to this. This is signalled in case studies as arising from the mental illness of the individual who fails to recognise the seriousness of their plight, and presents to services only when their condition is advanced and treatment options are limited. 124,132,133,137,144 Kunkel et al. 132 present four case examples of individuals with existing mental health problems, including that of Ms A, a 38-year-old African American woman with an unspecified psychosis presenting with a 20 cm mass in her right breast that was adhering to the chest wall. Investigations revealed a significant cancer with lymph node involvement but, despite the late presentation, there were no metastases. Ms A was treated with chemotherapy and surgery but then failed to attend for radiation treatment and considered the new lesions as nothing more than ‘haematoma’. She later presented to the ED in respiratory distress and with large, presumably malignant, effusions, asked not to be resuscitated and died soon after. The case study paper labels Ms A’s difficulty with accepting her situation and similar behaviours as ‘maladaptive denial’.

This tendency to impute the mental conditions of individuals as a source of late presentation, or indeed of deciding not to continue with treatment, is a recurring feature of this literature. Other case study examples suggest how diagnosis and treatment of serious life-threatening or life-limiting physical conditions are complicated by untreated mental illness and the beliefs and behaviours arising from this. 132,133,137 A number of papers also appear to elide behavioural aspects of presentations with mental ill-health and position these as problematic in the delivery of care and treatment, with terms used such as ‘malingering’, ‘manipulative’, ‘dominance and aggression’ and ‘demanding’,115,122,123,125 or problems for staff who ‘struggled with how best to provide care’. 126

Denial by people with SMI, or seeming to have a lack of insight, is a possible complicating factor. It is not unknown for any persons diagnosed with serious conditions to deny the gravity of their situation, and for people with pre-existing mental health conditions it appears that this is no different. 132 Feldman and Periyakoil126 note that specific issues arise in cases of people with PTSD, where ‘the threat to life inherent in terminal illness may mimic the original trauma’, leading to an exacerbation of the psychological symptoms associated with the condition, including anxiety, anger, denial, avoidance and distrust of authority, impeding medical adherence and resulting in the refusal of treatment.

There are also examples of individuals with long histories of pre-existing mental health problems, often in long-stay facilities, who, despite their access to HCPs, experience delayed or late diagnosis of conditions that placed them on the EoL trajectory. 26,119,127,138,151–153 Terpstra et al. 144 present the case of a 61-year-old man with a longstanding diagnosis of schizophrenia, living in an adult foster home and under legal guardianship. The man was seen every 2 weeks at a mental health clinic and attended a primary care provider to report a new-onset cough. His examination was documented as benign, but 1 week later he complained of hypotension and left-sided weakness. A detailed examination revealed advanced bladder cancer and multiple brain metastases that had not been spotted until this last presentation. The authors locate the problem in the mental condition of the person and suggest that people with conditions such as schizophrenia do not willingly verbalise pain or related symptoms. It is, however, difficult to escape the issue that extended contact with health-care services could have included some monitoring of the physical health of the person.

Decisional capacity and dilemmas

Many of the case studies feature a concern about the decisional capacity of individuals to consent to treatment and/or to refuse treatments offered, and implicate professional dilemmas of determining the value of attempting curative treatments compared with palliation. 118,120,122,127,132,133,146,147,154 Romm et al. 140 argue that mental health staff have a particular duty to ensure that the interests of people with SMI are defended via treatments targeting potentially distorted states of mind, adding that although SMI can impair decisional capacity this should not be assumed and should be thoroughly assessed. 140 In some case study examples, the conclusion arrived at was that the person retained the capacity to determine their treatment choices, and that these choices were then respected. 120

Treatment refusal is highlighted numerous times. 120,124,127,128,133,136,139,140,143,145–147 Past refusals seem to prompt treatment teams to impute future problems. 139 For example, Gonzalez et al. 128 present the case of an 86-year-old with a 20-year history of mastectomy for cancer and refusal of medical care who re-presented with a bleeding and ulcerating mass on the chest wall. The case study reported that the patient was not a case for curative treatment even though there was no evidence of metastases. The reasons for lack of curative efforts were indicated as:

curative treatment was not feasible based on a geriatric evaluation, which considered the patient’s severe cognitive decline, advanced age of the patient’s caregiver, and the expectations of the patient and his family.

Gonzalez et al. 128

In some cases, treatment refusal is reported as arising from psychiatric symptomatology, such as fixed beliefs about damnation or that thoughts could be read by physicians, rather than being related to capacity to understand and make decisions based on available information. 26

Approaching the courts for guidance is reportedly made in a variety of contexts, including where major concerns exist relating to a patient’s capacity or best interests and for declarations on treatment in the case of patients lacking decision-making capacity. 146. Shah et al. 141 report an example of treatment refusal by a man with decisional incapacity where the review of the medical ethics committee concluded that even with full decision-making capacity a person might refuse the radical procedures being proposed because of the risks involved and the deforming nature of the surgery.

Fluctuating mental capacity requiring multiple assessments118,146 is also reported. In some reports, the option to review and return to previous decisions is implied in case examples. 141

The absence of previously declared wishes on life-saving treatments appears to be a recurring issue and suggests one area for future intervention testing. 150 The absence of any declaration on treatment may mean that teams seek agreement of a substitute decision-maker, such as a family member,139,140 sometimes with power of attorney, but the patient may indicate their refusal in their actions, for example repeatedly removing life-saving treatments such as catheters in renal dialysis150 or not agreeing to take medication. 132 Enforcing medical treatment of a person who is actively refusing is not supported in this literature and it is argued that doing so may complicate future alliance building. 139

Medical futility

Some case studies reveal, in the example of people with anorexia nervosa, how experience of this condition can exhaust the optimism of those doing the caring, and the language used, such as ‘refractory’ and ‘incurable’, reflects this. 148 Lopez et al. 134 describe the case of Ms A, a 30-year-old woman with a diagnosis of anorexia nervosa, where ‘neither forcing her into involuntary treatment nor waiting for her to voluntarily engage in treatment appeared likely to resolve her illness, return her to a state of life-sustaining clinical stability or provide her a decent quality of life’. 134 The paper invokes the concept of medical futility, and the case demonstrates an example of the use of palliation and referral to hospice care for this psychiatric condition. In this case, it is also reported that the option of an out-of-state eating disorder treatment programme, which was acceptable, was eventually rejected on cost grounds, implying that perhaps the situation was less futile than indicated. Nevertheless, the treatment team and the ethics committee concluded ‘that her physical and psychiatric impairments were likely to lead to her death, despite any plausible attempts at aggressive intervention’. 134 The palliative care team therefore commenced involvement with the consent of both the patient and the family.

Individuals and their networks

For many people with SMI the availability of support networks in their communities is crucial, and the absence of these supports is implicated in delayed treatment seeking. Some case studies refer to family involvement,115,129,130,135,145,154 whereas others indicate the lack or absence of such support. 122,124,180 The EoL condition in some circumstances appears to have led to the re-emergence of family support,152 or that teams actively support the person to reconnect with distant family. 120 In some instances, tensions in family involvement are reported,123,131 and in one example family dynamics are situated as the source of subsequent mental health-related troubles. 148 Other case studies identify the reciprocal nature of support, and report the concerns of the individual with the EoL condition for their family. 129,130 EoL care places an additional set of demands on families, such as learning to manage symptoms,116 and some conditions appear to result in exhausted or burned-out family members. 134 An additional demand on support networks can arise from the absence of advance decisions or where decisional capacity is in question so that, for instance, families may be involved in discussions on treatment and with do-not-resuscitate decisions for which they may be ill-prepared. 127,132,139,140,150 Families also show some concern that their family member’s mental health problems are overlooked by the palliative care team. 145

Care provision

Where case studies report on the ongoing delivery of care to individuals with pre-existing SMI and subsequent EoL condition, they advocate the benefits of a transdisciplinary approach involving palliative care specialists, psychiatric specialists (preferably the team whose members already know the person) and wider community members (e.g. religious ministers). 15,26,119,120,134,140,148 Stecker142 provides an example of drawing on a specialist in both oncology and psychiatry to advise an inpatient mental health nursing team.

Liaison consultation and collaboration is reported as beneficial for hospice staff. Examples include learning from psychiatric consultation,125 learning from psychologists on how to alter the environment to reduce distress116 and a hospice nurse being supported by a psychiatric nurse. 131 Case reports note the mutual benefits of hospice and mental health nurses working together in EoL care, and the similarities in their work, but they also note a need for hospice staff to be better informed about mental health conditions. 145 Case studies also demonstrate the interest that mental health teams have in learning from their colleagues working in palliative care. 152

Some case study papers present models of care, although these are small scale and may not indicate either transferability or generalisability. Examples include treatment models such as dynamic system analysis,115 stepwise psychosocial palliative care116,117 and the patient, provider, systems model. 26

Case studies reveal issues in the provision of care for treatment teams, such as how to handle psychiatric presentations. 115,123,139,145 Case studies also report examples of what has worked in supporting people with mental health problems at the EoL, including building rapport and trust, especially with people with PTSD;116 using music therapy at the EoL;149 having conversations about death;26,129,130,152 initiating hospice at home;26,129,130 and providing multidisciplinary mental health care and palliative care at home. 120

An array of places for the provision of EoL care is proposed and supported in case studies. In almost every example the place of care and treatment is positioned as one arising from the preferences of the individual. These include palliative care provision in mental health institutions,127,152,153 hospice care for mental health patients,125,135,144,148 acute hospital15 and home care. 120 Repeated transfer between settings related to psychiatric and/or physical symptomatology is also evident,139 suggesting potential challenges to continuity of care.

Challenges reported in this literature include mental health staff being emotionally unprepared for caring for people who are terminally ill,142 an observation that it is suggested supports the need for staff to become aware of feelings that may interfere with their therapeutic functioning. Where people with SMI receive EoL care in hospices, palliative care staff are reported as experiencing strong emotions, such as anger and frustration. 116 Case study papers also occasionally offer psychodynamic interpretations of staff–patient interactions and care. 115,131,142

Treatment challenges are reported in managing some mental health-related medication alongside the provision of chemotherapy. Two case study papers report issues with the prescription of clozapine, as this is implicated in depressing white blood cell counts, which, in consequence, could precipitate a life-threatening infection. 136,138 Treatment challenges also arise from people with SMI at the EoL absconding or being lost to follow-up. 137,147 One paper reports a concern regarding suicide risk assessment in EoL care, although this is not a feature directly relevant in the case presented. 145 A further challenge is in providing palliative care to patients who, despite being provided with careful explanation, continue to believe that they are not going to die. 134

Summary of what has been learned

Figure 6 summarises the main findings from the two thematic syntheses, the first drawing on research, policy and guidance and the second drawing on case studies.

FIGURE 6.

Summary of thematic syntheses.

Overarching summary of first synthesis

This section presents an overarching summary of the first synthesis (see Chapter 4) complete with an assessment of confidence. The qualitative and non-intervention findings were assessed for confidence using the CERQual (Table 8) approach. The quality of evidence derived from retrospective cohort studies also included in this first synthesis was assessed using the GRADE approach. Appendices 21–24 include the detail of how the CERQual statements were generated, along with the detail of the ungraded summary statements derived from policy and guidance material.

A high degree of confidence was judged to apply to 10 of the 52 synthesis summary statements, distributed across all four major themes (Structure of the system, Professional issues, Contexts of care and Living with severe mental illness). All 10 were summary statements synthesised from generally well-conducted, relevant studies in which there was a high degree of coherence. A moderate degree of confidence was judged to apply to 32 of the statements, denoting concerns relating to one or more of the dimensions assessed in the underpinning evidence (methodology, coherence, relevance and adequacy), with the confidence in the remaining 10 summary statements judged as being either low or very low.

Implications for policy, services and practice

The purpose of assessing confidence in summaries of synthesised findings is to better support decisions by policy-makers, managers and senior practitioners, including in the areas of policy and guideline development. 82 The GRADE and CERQual assessments presented above are explicitly drawn on in this section. Particular weight is given to the 10 GRADE and CERQual summary statements, derived uniquely from research studies included in the review, and therefore not from case studies, policy or guidance, in which there is a high degree of confidence.

The key implications are as follows:

1. This evidence synthesis points with a high degree of confidence to the importance of formal and informal partnership opportunities being taken and encouraged. Partnerships can involve the building of coalitions across the whole system, including mental health, EoL, primary care, social care and other services. This is necessary to support the exchange of information in the integrated care of people with SMI at the EoL. In a context in which there is often ‘no right place to die’ and in which mental health staff are often poorly equipped to care for people at the EoL, and where people with SMI at the EoL face frequent moves between services, finding ways of supporting people to die in the location of their choice is a priority.

2. This synthesis also has implications for professional practice. Many people with SMI have limited social networks and little or no family contact, but even though mental health staff value the quality of their relationships with patients, some find it difficult to invest in relationships with people with SMI at the EoL because of the upset caused when patients are transferred to other facilities for EoL care. Mental health staff, many of whom find talking about death and dying with patients difficult, find that when opportunities are found patients are receptive. In the case of HCPs working in EoL care, stigmatising and prejudicial attitudes towards people with SMI and people who are homeless can affect EoL decision-making. All of these summary statements have clear implications for the future education, support and supervision of all care staff working across both the mental health and the EoL care systems.

3. A high-confidence synthesis summary from the contexts of care domain is that programmes and services for people with SMI at the EoL require a comprehensive team approach that incorporates symptom relief, psychological and psychosocial support, and spiritual care. Comprehensive services of this type are exactly as should be expected by all members of the population, but their importance for people with SMI at the EoL is worth restating for the purposes of promoting parity of esteem. Capable advocates are able to increase the numbers of people with SMI at the EoL who are referred to palliative care services, and to help make sure that palliative care is provided and received. Not having an advocate risks people with SMI who are lacking social and family support becoming ‘lost in the system’. Taken together, these synthesis summaries have important implications for the identification of roles for members of the care team in co-ordinating services across boundaries, advocating for and on behalf of patients, and providing direct care.

4. A final high-confidence summary statement is derived from the evidence synthesised in the area of living with SMI. This review has found that the timely provision of palliative care can be hindered when people with SMI, and particularly those who are homeless, receive delayed diagnoses of their EoL-related conditions. This has implications for the proactive provision of physical health care to groups of people who might not recognise the seriousness of their condition and who might miss out on routine health assessments.

Recommendations for future research

The case studies synthesised in Chapter 5 demonstrate the complexities and the ethical dilemmas associated with the provision of care to people with SMI at the EoL. Many are stark in revealing the challenges presented by delayed diagnosis, patients’ fluctuating capacity to make decisions and the futility of treatment. Case studies also demonstrate, as does the included research, the unpreparedness of professionals to meet patients’ multiple needs. Synthesised case study findings also show the dangers of ascribing delays in mobilising (or continuing with) palliative care services in response to what are often described, in written reports, as patients’ challenging presentations, or even their wilfully difficult behaviour.

Among the research publications from which data were synthesised in Chapter 4, only three originate from the UK, and across all included research studies very few include either patients or family members as participants. The consequence of this is that information on the experiences of people with SMI at the EoL features most prominently in the synthesis of case studies, making this a wide-open area for future well-designed research. Significantly, none of the included research publications reported on intervention studies, and no research was found in the area of EoL care provided to people with SMI in the prison setting.

In this context of a paucity of research evidence, key recommendations for future studies arising from this evidence synthesis are as follows:

1. Patient- and family-facing studies should be commissioned to examine the experiences of people with SMI at the EoL, and the experiences of their surrounding paid and unpaid carers and those important to them, to establish the factors that help and hinder care in the UK-specific context.

2. A programme of research should be commissioned comprising studies that explicitly co-produce, introduce and evaluate new ways of providing and organising EoL care for people with SMI with both cancer and other life-threatening conditions. This programme should include studies involving support for structurally disadvantaged groups, including people with SMI at the EoL who are also homeless or who are in prison. Candidate interventions include advance care plans, advocacy and improved education for professionals, along with the development of new or enhanced roles for practitioners and the introduction of models of integrated provision spanning the mental health, EoL and related care systems.

Limitations

All 30 research studies reported in 34 publications included in this review were undertaken in high-income countries with developed health systems, but only three were completed in the UK. 37,112,113 All three UK studies used qualitative methods, but only one113 included people who had experience of using services and none evaluated interventions to improve care. Although implications for policy, services and practice arising from this project have been drawn from all 30 studies, it is recognised that contextual differences in global health and social care systems (e.g. relating to the funding of services) need to be borne in mind when the transferability of findings and summary evidence statements are considered. In addition, it is recognised that the search for policy and guidance was limited to the UK only, rather than being extended to other countries around the world, including those with broadly comparable health systems. With only English-language items included in this project, the possibility also exists that important research and other evidence of relevance has been neither identified nor included. Finally, as no intervention studies were located, no meta-analysis could be performed.

Conclusion

This rigorously conducted evidence synthesis has brought together policy, guidance, research and case studies in an important but neglected area: EoL care for people with SMI. Although findings from this review are of immediate and direct relevance in this very specific area, they also have relevance for the EoL care of other disadvantaged groups in which health inequalities persist. Reflecting approaches developed through the EPPI-Centre, a strength of this synthesis is the inclusion of both research and non-research material, coupled with the exclusive use of research evidence to inform the project’s overall implications for policy, services and practice. The project was launched in response to a need identified by people with lived experience of mental health and EoL services, one of whom proceeded to work as a full member of the project team doing work that included reviewing and approving this full report and its Scientific and Plain English summaries. The establishment of a SAG populated by senior professionals, government advisors, and people with direct experience of service use and of caring was key, and this was a further vehicle for a public and patient voice to be heard at strategic points during the lifetime of the project.

Outputs written for internationally leading journals will follow, reporting main and other findings from this full NIHR Journals Library report along with the production of an accessible summary for wide dissemination. Presentations are planned at key conferences, and with help from the project’s SAG key policy-makers, managers, professionals and organisations these will be targeted to make sure that the findings reach people who are able to make full use of them. With this evidence synthesis designed as the starting point for a programme of research, further projects are planned, including studies involving people with SMI at the EoL and their carers and studies evaluating new ways of providing and organising care.

Contributions of authors

Ben Hannigan (https://orcid.org/0000-0002-2512-6721) (Professor of Mental Health Nursing) contributed to the development of the protocol; screened, selected, appraised and synthesised materials; and wrote for and edited the final report.

Deborah Edwards (https://orcid.org/0000-0003-1885-9297) (Research Associate) contributed to the development of the protocol; screened, selected, appraised and synthesised materials; and wrote for and edited the final report.

Sally Anstey (https://orcid.org/0000-0003-2295-3761) (Senior Lecturer in Adult Nursing) contributed to the development of the protocol; screened, selected, appraised and synthesised materials; and wrote for and commented on the final report.

Michael Coffey (https://orcid.org/0000-0002-0380-4704) (Professor of Mental Health Care) contributed to the development of the protocol; screened, selected, appraised and synthesised materials; and wrote for and commented on the final report.

Paul Gill (https://orcid.org/0000-0003-4056-3230) (Senior Lecturer in Adult Nursing) contributed to the development of the protocol; screened, selected, appraised and synthesised materials; and wrote for and commented on the final report.

Mala Mann (https://orcid.org/0000-0002-2554-9265) (Information Specialist) contributed to the development of the protocol, and searched for materials and wrote for and commented on the final report.

Alan Meudell (https://orcid.org/0000-0001-8138-4744) (Independent Service User Researcher) contributed to the development of the protocol, screened materials, and wrote for and edited the final report.

Publications

Edwards D, Anstey S, Coffey M, Gill P, Mann M, Meudell A, Hannigan B. End of life care for people with severe mental illness: mixed methods systematic review and thematic synthesis (the MENLOC study). Palliat Med 2021;35:1747–60.

Coffey M, Edwards D, Anstey S, Gill P, Mann M, Meudell A, Hannigan B. End of life care for people with severe mental illness: mixed methods systematic review and thematic synthesis of published case studies (the MENLOC study). BMJ Open 2022;12:e053223.

Data-sharing statement

All available data can be obtained from the corresponding author.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HSDR programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HSDR programme or the Department of Health and Social Care.

Background

MENLOC aims to synthesise relevant research and other appropriate evidence relating to the organisation, provision and receipt of end-of-life care for people with SMI who have an additional diagnosis of advanced incurable cancer and/or end-stage lung, heart, renal or liver failure and who are likely to die within the next 12 months.

Project team and stakeholder advisory group: membership

The MENLOC project team comprises people with backgrounds in mental health services and end-of-life care research and practice, health services research, evidence review and information sciences, and lived experience of mental health difficulties and of surviving cancer.

The MENLOC SAG comprises people drawn from communities with interests in the project and its findings, including HCPs, mental health and end-of-life care researchers, service managers from the NHS and charitable sector, and people with experience of using services and of caring for people using services.

Stakeholder advisory group: terms of reference

The SAG will be independently chaired and will provide guidance to the project team over the lifetime of the study. The SAG will specifically:

1. receive information on the aims and objectives of MENLOC, and provide guidance on key definitions (e.g. ‘severe mental illness’), the corresponding refinement of search terms and strategies for evidence searching and review

2. review MENLOC’s progress and receive preliminary findings and provide a final round of guidance on evidence searching and review

3. advise the project team on the dissemination of findings, the promotion of impact and the identification of future research questions.

Working together

It is anticipated that members of the MENLOC project team and the SAG will meet on three occasions in Cardiff. In addition to participating in these meetings, members of the SAG can expect to spend approximately 1–2 hours in preparation. Members of the SAG will also be welcomed to the MENLOC end-of-project launch event.

Funding acknowledgement

MENLOC is funded by the NIHR Health and Social Care Delivery Research programme (project number 17/100/15).

Department of Health and Social Care disclaimer

The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.

Date range searched: 1946 to 5 July 2018. Last date searched: 6 December 2019

1. exp Palliative care/ (48,615)

2. exp Hospice care/ (5807)

3. exp Terminal Care/ (47,347)

4. exp Terminally ill/ (6150)

5. (“palliative care” or “hospice care” or “end of life care” or end-of-life).tw. (38,005)

6. ((hospice or terminal*) adj3 (care or caring or ill*)).tw. (12,527)

7. (“last year of life” or LYOL or “end of life” or “end of their lives”).tw. (19,101)

8. (end-stage disease* or end stage disease* or end-stage ill* or end stage ill*).tw. (1112)

9. or/1-8 (104,642)

10. exp Neoplasms/ (3,055,654)

11. (cancer* or tumo?r* or neoplas* or malignan* or carcinoma* or adenocarcinoma* or adeno?carcinoma* or choriocrcinoma* or leukemia* or leukaemia* or metastat* or sarcoma* or teratoma* or lymphoma).tw. (3,168,623)

12. exp Multiple Organ Failure/ (10,043)

13. Organ failure.tw. (16,597)

14. Pulmonary Disease, Chronic Obstructive/ (32,992)

15. (Chronic obstructive pulmonary disease or COPD).tw. (53,007)

16. Heart Failure/ (105,052)

17. Renal Insufficiency/ (14,550)

18. Liver Failure/ (6735)

19. (“cardiac failure” or “liver failure” or “kidney failure” or “renal failure”).tw. (114,515)

20. Chronic Disease/ (248,675)

21. Health service utilization.mp. (1502)

22. lifestyle-related factor*.mp. (337)

23. or/10-22 (4,430,236)

24. exp Bipolar Disorder/ (37,229)

25. exp Schizophrenia/ (97,818)

26. (bipolar or mania or Schizophrenia).tw. (148,564)

27. exp Depression/ (102,372)

28. exp Mental Disorders/ (1,125,241)

29. (“mental illness” or “mental disorder*” or depression).tw. (333,320)

30. mental health condition*.tw. (1794)

31. exp Depressive Disorder/ (99,383)

32. exp Mental Health/ (31,108)

33. exp Personality Disorders/ (39,084)

34. (severe and persistent mental illness).mp. (304)

35. severe mental illness.mp. (3558)

36. or/24-35 (1,3902,77)

37. 9 and 23 and 36 (2648)

38. (dementia or Alzheimer).tw. (106,344)

39. (Algeria$ or Egypt$ or Liby$ or Morocc$ or Tunisia$ or Western Sahara$ or Angola$ or Benin or Botswana$ or Burkina Faso or Burundi or Cameroon or Cape Verde or Central African Republic or Chad or Comoros or Congo or Djibouti or Eritrea or Ethiopia$ or Gabon or Gambia$ or Ghana or Guinea or Keny$ or Lesotho or Liberia or Madagasca$ or Malawi or Mali or Mauritania or Mauritius or Mayotte or Mozambiq$ or Namibia$ or Niger or Nigeria$ or Reunion or Rwand$ or Saint Helena or Senegal or Seychelles or Sierra Leone or Somalia or South Africa$ or Sudan or Swaziland or Tanzania or Togo or Ugand$ or Zambia$ or Zimbabw$ or China or Chinese or Hong Kong or Macao or Mongolia$ or Taiwan$ or Belarus or Moldov$ or Russia$ or Ukraine or Afghanistan or Armenia$ or Azerbaijan or Bahrain or Cyprus or Cypriot or Georgia$ or Iran$ or Iraq$ or Jordan$ or Kazakhstan or Kuwait or Kyrgyzstan or Leban$ or Oman or Pakistan$ or Palestin$ or Qatar or Saudi Arabia or Syria$ or Tajikistan or Turkmenistan or United Arab Emirates or Uzbekistan or Yemen or Bangladesh$ or Bhutan or British Indian Ocean Territory or Brunei Darussalam or Cambodia$ or India$ or Indonesia$ or Lao or People’s Democratic Republic or Malaysia$ or Maldives or Myanmar or Nepal or Philippin$ or Singapore or Sri Lanka or Thai$ or Timor Leste or Vietnam or Albania$ or Andorra or Bosnia$ or Herzegovina$ or Bulgaria$ or Croatia$ or Faroe Islands or Greenland or Liechtenstein or Lithuani$ or Macedonia or Malta or maltese or Romania or Serbia$ or Montenegro or Svalbard or Argentina$ or Belize or Bolivia$ or Brazil$ or Chilean or Colombia$ or Costa Rica$ or Cuba or Ecuador or El Salvador or French Guiana or Guatemala$ or Guyana or Haiti or Honduras or Jamaica$ or Nicaragua$ or Panama or Paraguay or Peru or Puerto Rico or Suriname or Uruguay or Venezuela or developing countr$ or south America$).ti,sh. (1,190,707)

40. Academic Dissertations/ (0)

41. thesis.tw. (7038)

42. book.pt. (0)

43. Books/ (3122)

44. or/38-43 (1,303,105)

45. 37 not 44 (2373)

46. limit 45 to (English language and humans and “all adult (19 plus years)”) (1387)

Date range searched: 1946 to 6 December 2018. Last date searched: 6 December 2019

1. exp Palliative care/ (49,655)

2. exp Hospice care/ (5949)

3. exp Terminal Care/ (48,040)

4. exp Terminally ill/ (6227)

5. (“palliative care” or hospice or “end of life care” or end-of-life).tw. (43,896)

6. ((hospice or terminal*) adj3 (care or caring or ill*)).tw. (12,766)

7. (“Irreversible condition” or “terminal condition” or fatal illness).tw. (791)

8. (“last year of life” or LYOL or “end of life” or “end of their lives” or “last six months of life” or “last 6 months of life”).tw. (19,924)

9. (end-stage disease* or end stage disease* or end-stage ill* or end stage ill* or end-stage or end stage).tw. (60,633)

10. (expected adj3 die).tw. (215)

11. (imminent adj3 death).tw. (561)

12. (“Dying soon” or “expected death” or “imminently dying” or Moribund).tw. (2493)

13. conservative treatment/ (1401)

14. (conservative adj2 (treatment or management)).tw. (42,831)

15. Withholding Treatment/ (10,909)

16. Treatment Refusal/ (11,477)

17. (Refus* adj3 (treat* or care or intervention or dialysis)).tw. (4855)

18. ((withdrew or withdraw* or withhold*) adj3 (treat* or car* or intervene* or therap* or dialysis or transplant*)).tw. (15,455)

19. or/1-18 (244,771)

20. exp Neoplasms/ (3,107,101)

21. (cancer* or tumo?r* or neoplas* or malignan* or carcinoma* or adenocarcinoma* or adeno?carcinoma* or choriocrcinoma* or leukemia* or leukaemia* or metastat* or sarcoma* or teratoma* or lymphoma).tw. (3,243,976)

22. (“Enzymatic disease*” or “enzyme disease*”).tw. (25)

23. Genetic Disease, Inborn/ (13,207)

24. (“duchenne muscular dystrophy” or “amyotrophic lateral sclerosis” or ALS).tw. (40,207)

25. amyotrophic lateral sclerosis/ (16,921)

26. Muscular dystrophy, Duchenne/ (4706)

27. (“Genetic disease” or “genetic condition”).tw. (7305)

28. Cystic Fibrosis/ (33,150)

29. (“cystis fibrosis” or CF).tw. (36,968)

30. exp Multiple Organ Failure/ (10,209)

31. (“Organ failure” or “chronic organ failure”).tw. (17,137)

32. Pulmonary Disease, Chronic Obstructive/ (34,379)

33. (“renal insufficiency” or “serious physical illness”).tw. (21,134)

34. (Chronic obstructive pulmonary disease or COPD).tw. (54,881)

35. (chronic adj3 (illness or condition or disease*)).tw. (250,692)

36. “chronic medical condition”.tw. (382)

37. Heart Failure/ (107,446)

38. Renal Insufficiency/ (14,803)

39. Liver Failure/ (6828)

40. (chronic adj2 (“cardiac failure” or “liver failure” or “kidney failure” or “end-stage renal disease” or ESRD or “renal failure” or “heart failure”)).tw. (44,143)

41. Chronic Disease/ (251,517)

42. Health service utilization.mp. (1569)

43. lifestyle-related factor*.mp. (354)

44. or/20-43 (4,717,135)

45. exp Bipolar Disorder/ (37,764)

46. exp Schizophrenia/ (99,011)

47. (schizo* or “mood disorder*” or “personality disorder*” or psychotic* or psychosis or psychoses).tw. (197,049)

48. (bipolar or mania or Schizophrenia).tw. (151,754)

49. (Depression adj2 (psychosis or psychotic or severe or major)).tw. (28,840)

50. exp Mental Disorders/ (1,143,274)

51. mental health condition*.tw. (1954)

52. exp Personality Disorders/ (39,433)

53. exp psychotic disorders/ (48,805)

54. exp schizoaffective disorder/ (48,805)

55. exp psychosis/ (48,805)

56. Paranoid Disorders/ (3987)

57. (“severe mental illness” or “persistent mental illness”).tw. (3969)

58. ((chronic* or sever* or serious or persistent* or enduring or debilitating) adj2 (mental* or psychological*) adj2 (ill* or disorder* or health)).tw. (12,556)

59. exp Mentally Ill Persons/ (5926)

60. (SPMI or SMI).tw. (3830)

61. or/45-60 (1,239,759)

62. (dementia or Alzheimer).ti. (52,380)

63. (bipolar electrocoagulation or bipolar radiofrequency or bipolar tumour probe or bipolar diathermy).tw. (1123)

64. (“respiratory depression” or “marrow depression” or “hematologic* depression”).tw. (6399)

65. (child* or adoles* or pediatric or paediatric).tw. (1,520,041)

66. (Algeria$ or Egypt$ or Liby$ or Morocc$ or Tunisia$ or Western Sahara$ or Angola$ or Benin or Botswana$ or Burkina Faso or Burundi or Cameroon or Cape Verde or Central African Republic or Chad or Comoros or Congo or Djibouti or Eritrea or Ethiopia$ or Gabon or Gambia$ or Ghana or Guinea or Keny$ or Lesotho or Liberia or Madagasca$ or Malawi or Mali or Mauritania or Mauritius or Mayotte or Mozambiq$ or Namibia$ or Niger or Nigeria$ or Reunion or Rwand$ or Saint Helena or Senegal or Seychelles or Sierra Leone or Somalia or South Africa$ or Sudan or Swaziland or Tanzania or Togo or Ugand$ or Zambia$ or Zimbabw$ or China or Chinese or Hong Kong or Macao or Mongolia$ or Taiwan$ or Belarus or Moldov$ or Russia$ or Ukraine or Afghanistan or Armenia$ or Azerbaijan or Bahrain or Cyprus or Cypriot or Georgia$ or Iran$ or Iraq$ or Jordan$ or Kazakhstan or Kuwait or Kyrgyzstan or Leban$ or Oman or Pakistan$ or Palestin$ or Qatar or Saudi Arabia or Syria$ or Tajikistan or Turkmenistan or United Arab Emirates or Uzbekistan or Yemen or Bangladesh$ or Bhutan or British Indian Ocean Territory or Brunei Darussalam or Cambodia$ or India$ or Indonesia$ or Lao or People’s Democratic Republic or Malaysia$ or Maldives or Myanmar or Nepal or Philippin$ or Singapore or Sri Lanka or Thai$ or Timor Leste or Vietnam or Albania$ or Andorra or Bosnia$ or Herzegovina$ or Bulgaria$ or Croatia$ or Faroe Islands or Greenland or Liechtenstein or Lithuani$ or Macedonia or Malta or maltese or Romania or Serbia$ or Montenegro or Svalbard or Argentina$ or Belize or Bolivia$ or Brazil$ or Chilean or Colombia$ or Costa Rica$ or Cuba or Ecuador or El Salvador or French Guiana or Guatemala$ or Guyana or Haiti or Honduras or Jamaica$ or Nicaragua$ or Panama or Paraguay or Peru or Puerto Rico or Suriname or Uruguay or Venezuela or developing countr$ or south America$).ti,sh. (1,225,437)

67. Academic Dissertations/ (0)

68. thesis.tw. (7167)

69. book.pt. (0)

70. Books/ (3174)

71. or/62-70 (2,653,676)

72. 19 and 44 and 61 (2434)

73. 72 not 71 (2126)

74. limit 73 to (english language and humans and “all adult (19 plus years)”) (1217)

Date range searched: 1806 to December week 1 2018. Last date searched: 6 December 2019

1. exp Palliative Care/ (11,016)

2. Hospice/ (3062)

3. exp Terminally Ill Patients/ (4562)

4. exp HOSPICE/ (3062)

5. (“palliative care” or “hospice care” or “end of life care” or end-of-life).tw. (15,589)

6. ((hospice or terminal*) adj3 (care or caring or ill*)).tw. (6333)

7. (“Irreversible condition” or “terminal condition” or fatal illness).tw. (216)

8. (“last year of life” or LYOL or “end of life” or “end of their lives” or “last six months of life” or “last 6 months of life”).tw. (9038)

9. (end-stage disease* or end stage disease* or end-stage ill* or end stage ill* or end-stage or end stage).tw. (2002)

10. (expected adj3 die).tw. (44)

11. (imminent adj3 death).tw. (272)

12. (“Dying soon” or “expected death” or “imminently dying” or Moribund).tw. (235)

13. (conservative adj2 (treatment or management)).tw. (460)

14. Treatment Withholding/ (462)

15. Treatment Refusal/ (730)

16. (Refus* adj3 (treat* or care or intervention or dialysis)).tw. (1811)

17. ((withdrew or withdraw* or withhold*) adj3 (treat* or car* or intervene* or therap* or dialysis or transplant*)).tw. (3579)

18. or/1-17 (28,620)

19. exp NEOPLASMS/ (47,829)

20. (cancer* or tumo?r* or neoplas* or malignan* or carcinoma* or adenocarcinoma* or adeno?carcinoma* or choriocrcinoma* or leukemia* or leukaemia* or metastat* or sarcoma* or teratoma* or lymphoma).tw. (76,722)

21. (“Enzymatic disease*” or “enzyme disease*”).tw. (0)

22. Genetic Disorders/ (4048)

23. (“duchenne muscular dystrophy” or “amyotrophic lateral sclerosis” or ALS).tw. (13,477)

24. Amyotrophic Lateral Sclerosis/ (3551)

25. Muscular Dystrophy/ (1293)

26. (“Genetic disease” or “genetic condition”).tw. (611)

27. Cystic Fibrosis/ (854)

28. (“cystis fibrosis” or CF).tw. (3217)

29. “Multi* Organ Failure*”.tw. (71)

30. (“Organ failure” or “chronic organ failure”).tw. (203)

31. exp Chronic Obstructive Pulmonary Disease/ (1272)

32. (“renal insufficiency” or “serious physical illness”).tw. (319)

33. (Chronic obstructive pulmonary disease or COPD).tw. (2277)

34. (chronic adj3 (illness or condition or disease*)).tw. (28,394)

35. “chronic medical condition”.tw. (196)

36. Heart Disorders/ (8989)

37. Kidney Diseases/ (2001)

38. Liver Disorders/ (1145)

39. (chronic adj2 (“cardiac failure” or “liver failure” or “kidney failure” or “end-stage renal disease” or ESRD or “renal failure” or “heart failure”)).tw. (866)

40. Chronic Illness/ (10,804)

41. lifestyle-related factor*.mp. (51)

42. or/19-41 (139,563)

43. exp Bipolar Disorder/ (25,413)

44. exp SCHIZOPHRENIA/ (85,758)

45. (schizo* or “mood disorder*” or “personality disorder*” or psychotic* or psychosis or psychoses).tw. (208,100)

46. (bipolar or mania or Schizophrenia).tw. (137,089)

47. (Depression adj2 (psychosis or psychotic or severe or major)).tw. (34,922)

48. exp Mental Disorders/ (566,262)

49. mental health condition*.tw. (1698)

50. exp Personality Disorders/ (33,638)

51. exp Schizoaffective Disorder/ (2961)

52. exp Psychosis/ (109,879)

53. “Paranoia (Psychosis)”/ (1215)

54. “Severity (Disorders)”/ (15,885)

55. “severe mental illness”.tw. (4681)

56. “persistent mental illness”.tw. (799)

57. ((chronic* or sever* or serious or persistent* or enduring or debilitating) adj2 (mental* or psychological*) adj2 (ill* or disorder* or health)).tw. (16,725)

58. Mentally Ill Persons.mp. (805)

59. (SPMI or SMI).tw. (1991)

60. or/43-59 (637,058)

61. (dementia or Alzheimer).ti. (30,336)

62. (bipolar electrocoagulation or bipolar radiofrequency or bipolar tumour probe or bipolar diathermy).tw. (5)

63. (“respiratory depression” or “marrow depression” or “hematologic* depression”).tw. (407)

64. (child* or adoles* or pediatric or paediatric).tw. (811,926)

65. (Algeria$ or Egypt$ or Liby$ or Morocc$ or Tunisia$ or Western Sahara$ or Angola$ or Benin or Botswana$ or Burkina Faso or Burundi or Cameroon or Cape Verde or Central African Republic or Chad or Comoros or Congo or Djibouti or Eritrea or Ethiopia$ or Gabon or Gambia$ or Ghana or Guinea or Keny$ or Lesotho or Liberia or Madagasca$ or Malawi or Mali or Mauritania or Mauritius or Mayotte or Mozambiq$ or Namibia$ or Niger or Nigeria$ or Reunion or Rwand$ or Saint Helena or Senegal or Seychelles or Sierra Leone or Somalia or South Africa$ or Sudan or Swaziland or Tanzania or Togo or Ugand$ or Zambia$ or Zimbabw$ or China or Chinese or Hong Kong or Macao or Mongolia$ or Taiwan$ or Belarus or Moldov$ or Russia$ or Ukraine or Afghanistan or Armenia$ or Azerbaijan or Bahrain or Cyprus or Cypriot or Georgia$ or Iran$ or Iraq$ or Jordan$ or Kazakhstan or Kuwait or Kyrgyzstan or Leban$ or Oman or Pakistan$ or Palestin$ or Qatar or Saudi Arabia or Syria$ or Tajikistan or Turkmenistan or United Arab Emirates or Uzbekistan or Yemen or Bangladesh$ or Bhutan or British Indian Ocean Territory or Brunei Darussalam or Cambodia$ or India$ or Indonesia$ or Lao or People’s Democratic Republic or Malaysia$ or Maldives or Myanmar or Nepal or Philippin$ or Singapore or Sri Lanka or Thai$ or Timor Leste or Vietnam or Albania$ or Andorra or Bosnia$ or Herzegovina$ or Bulgaria$ or Croatia$ or Faroe Islands or Greenland or Liechtenstein or Lithuani$ or Macedonia or Malta or maltese or Romania or Serbia$ or Montenegro or Svalbard or Argentina$ or Belize or Bolivia$ or Brazil$ or Chilean or Colombia$ or Costa Rica$ or Cuba or Ecuador or El Salvador or French Guiana or Guatemala$ or Guyana or Haiti or Honduras or Jamaica$ or Nicaragua$ or Panama or Paraguay or Peru or Puerto Rico or Suriname or Uruguay or Venezuela or developing countr$ or south America$).ti,sh. (128,851)

66. thesis.tw. (24,537)

67. book.pt. (476,469)

68. exp BOOKS/ (6110)

69. or/61-68 (1,332,511)

70. 18 and 42 and 60 (838)

71. 70 not 69 (609)

Trial registers

Charity websites

Organisation websites