How do people who are suicidal experience themselves, their world, and others in it?

In 2009 the research team at SANE embarked upon a programme of qualitative research into the experience of suicide. This research is based on the first-person perspectives of people who have experienced suicidal feelings, and close family and friends of people who have attempted or completed suicide.

The programme consists of two projects:

PROJECT 1 – THE EXPERIENCE OF SUICIDAL FEELINGS

In this study we are hoping to get as close as possible to the personal reality of being suicidal – we want to know what it is like to feel suicidal.

In order to discover this we have been asking those who have experienced suicidal feelings and those who have attempted suicide about their experience

Stage 1 consists of 120 participants filling out questionnaires – this is now completed. Stage 2 we explore 40 participants’ experiences further using interviews – this is currently underway and we hope to complete the project by August 2011.

PROJECT 2 – A NEW FOCUS FOR SUICIDE PREVENTION

In this study we will investigate the process of suicide – in particular, how it appears to those in close relationship with someone who is feeling suicidal. We want to develop an understanding of how suicidal people, and their significant others, experience the process of suicide together. Ultimately we hope to uncover insights into the process of suicide that will help everybody in the community, including professionals, take part in preventing suicide.

We will talk to friends and relatives of around 30 people who have completed suicide, and to 30 people who have attempted suicide themselves, together with members of their family or close friends. We aim to complete this project by June 2013.

DISSEMINATION

We will produce an internet resource (developed with service users) based on the findings of the programme which we hope will help families, friends, colleagues, health care professionals and others to offer more appropriate support and help to those at risk of suicide. We will also: (i) write articles for publication in academic journals based on the research findings; (ii) organise a conference; (iii) present at conferences organised by others and; (iv) publish a leaflet to be distributed through GP surgeries.

In addition our findings will be used to make our own services better, and improve the way our helpline and email volunteers respond to people who disclose suicidal behaviours, feelings or thoughts. The results will also inform our campaigning and media activities, as we work towards better services for and more appropriate representation of people who are suicidal or experiencing suicidal feelings.

Meeting of the Smaller Mental Health Research  Funders  

Monday, 21st February 2011, 13:00 – 16:30 Wellcome Trust 

             Attendance

 The meeting was attended by the following:  

 Apologies were received from Dennis Stevenson, Naomi Garnett, Til Wykes, Nick Rawlins, Ged Flynn, Raghu Gaind, Shirley Smith.

1. CC welcomed everyone to the meeting.  Attendees introduced themselves and their organisations.  

            News from members of the forum.            

1. AM gave an update.  At the last meeting the Judi Meadows Memorial Fund was about to make grant awards.  Awards have now been made to two academic teams:

1. A systematic review of international research literature on risk factors for suicide for those with depression to be undertaken by Keith Hawton of Oxford University.  On the basis of the results they will produce guidance for clinicians to assist in pre-assessing suicide risk.  A two page summary of the research report will be available on the Mental Health Research Funders website.  The project starts in June for eighteen months.  Three six monthly reports will be produced and made available.

Action: AM to ensure reports made available on website.

1. The second project will be undertaken by Ann Adams of Warwick Medical School.  She will be looking at implementation of a GRIST Clinical Support System currently used within secondary care.  The project will look at widening its use to both primary care and self care with an aim to improve detection of suicide risk in depressed patients.  There will be a version designed for primary care and a My-GRIST that will be filled in by patients.  Tools will be web based and will assist doctors and service users to detect suicide risk.  The project already has good links with Improving Access to Psychological Therapies (IAPT) and four Primary Care Trusts have agreed to trial during development phase.

Both projects were selected following an open tender and short-listing.  Each project was peer-reviewed by at least four people.  The two selected were peer reviewed by six individuals.

1. CC explained that IAPT is being rolled out to give faster access to services for people with anxiety and depression.

1. CMH gave an update on the work of the James Wentworth Stanley Memorial Fund.  She reported that a meeting of all charities with an interest in suicide prevention had been held with all invitees attending.  The idea of the meeting was to avoid duplication of funds and efforts by trying to work collaboratively and cross-referencing etc.  There were a number of main points for discussion:

Changing attitudes

Stigma

Co-ordinating resources

Lobbying government

Training of General Practitioners and frontline NHS staff.

Establishment of a national 24 hour suicide helpline.

Coroners and Inquests – Coroners don’t give out details of suicides to protect families but this data would be helpful for future prevention strategies.

Confidentiality and capacity – Is it correct that confidentiality should still apply when someone is a danger to themselves?

Screening and risk assessment

CMH explained that she had visited Maggie’s Cancer Centre.  The centre provides a very good service that is supportive and non-clinical.  CMH felt that this would work well for mental health.

AM informed the group that there is a meeting in July at the University of Guilford where participants will be looking to extend collaboration with clinicians and researchers.  The focus will be on reforming mental health services with the use of care pathways.

CMH explained the suicide research project that is taking place to examine the experience of suicidal feelings – the aim is to get as close as possible to the reality of feeling suicidal.  The method will be via questionnaires and interviews.  Participants were recruited to the research in June 2010.  The project is due for completion in August 2011.  It is hoped that a new focus for suicide prevention may be found by looking at the process of suicide and investigating how people who are suicidal and their families experience suicide.  Results will be disseminated through an internet resource based on the findings, articles produced for publication in academic journals, publication of a leaflet and presentation of the findings at conferences.

1. MW briefly talked about how to intervene best and deal with families that are bereaved.

1. There was a discussion about Health Talk on line.  The web resource contains very detailed patient interviews.  In 2008 they carried out 40 interviews with people to examine bereavement due to suicide.   The web address is www.healthtalkonline.org

1.  CC suggested that lay summaries of the research commissioned should be put onto the mental health funders website.

Action: AM and CMH to provide lay summaries for upload onto website.

 Quantitative Research: What questions can it answer? (Professor Brian Everitt (BE) IoP) 

BE gave a presentation entitled: “Treatments, good, bad or worthless and how do we tell?”

The main focus of the presentation was the importance of evidence and the importance of having a control group to which individuals are randomly selected.  He explained that the double blind, randomised controlled trail (RCT) is seen as the ‘gold standard’ when evaluating treatments.  There was a brief discussion of consent and the amount of information that should be given to patients.  Brian finished by explaining that an RCT is not the only route to the ‘ truth’ but that in psychiatry effects are rarely clear cut so often an RCT is required.

1. JB sits on an ethics committee and a national ethics panel.  He described how most of the RCT’s coming through the committee are historical controlled trails as opposed to RCT’s.  BE explained that these trials are quicker and may be a first step but that there is a high potential for bias.

10.  GS asked if there was an RCT that could detect a rise in suicide rates?  BE stated that in order to detect something like this an estimate of treatment effect would be required so that a power calculation could be applied.  He went on to say that due to meta-analysis even small trials can now contribute to an overall conclusion.

11.  TI enquired as to whether duration of trials could cause problems from a funding perspective?  BE replied that this is very difficult as sometimes trials can continue for a very long time if certain outcomes (e.g. death) are being recorded.  On the other hand, in drug trails there is a pressure to get quick results which also leads to challenges.

12.  VP talked about a ‘cleansed sample’ being used.  Participants in trials are not the same as in ‘real world’ clinics.  In addition, follow ups in trials are not long enough.  BE responded that it is sometimes very difficult to carry out trails that are completely generalisable.  CC put forward the idea of looking at it as a two stage process.  The first stage would be to do the gold standard RCT to demonstrate the treatment.  The next stage would then be to try out the treatment on the totality of patients.

Qualitative Research: what questions can it answer? (Dr Joanna Murray (JM) KCL)

JM explained that the focus of qualitative research is very much on meaning, beliefs and feelings.  It is about lived experience of people most concerned, service-users, their families and clinicians etc. 

JM gave an account of how in the early 1990’s the BMJ published a series of articles describing the benefits of qualitative techniques in medical research.  This helped to put qualitative research on a level footing with quantitative research.

She talked about how qualitative research can be used to identify the variables which can quantitative techniques can then utilise.

13.  MW asked whether it matters if the research is for a person with a neutral agenda? JM responded that a care history stops usually at a descriptive account.  MW questioned where it differs from journalism.  The reply was that it is more systematic than journalism and will involve a research team and more processes.  Others are involved in qualitative research in the process of constant comparison.  Themes are consistently used and coded and there are numerous checks and balances.

14.  MW enquired as to how to replicate in qualitative research.  JM described how it is important to allow people to see you process, methods and coding so that people can see if they agree.

15.  TC talked about how research is a long journey with different methodologies required along the way.  He said that we need to know where we are in the process, what are the questions we are trying to answer and to what effect?  We also need to look at what is fit for purpose?  TC explained that in terms of mental health we are often very early on in the process so qualitative research may be necessary to define the areas upon which quantitative research should focus.

16.  AM emphasised that as funders there is a need to determine the sort of research methods we should be looking at that would be most likely to influence policy.  She posed the question, will policy makers see the robustness of qualitative methods alone?

17.  CC stressed that we are trying to get across the importance of whatever we do being fit for purpose to answer the question.  There was again description of research as a long road with different methods applicable along the road.  Peer review can be helpful in helping us to determine where we are on the road and whether the methodology is fit for purpose.

18.  AM stated that collaboration between researchers is very important. Data can be interrogated in different ways by different academics in a variety of studies.

What would you do with £50k? (chaired by VP)

There were four presentations by the following individuals:

• Clair Chilvers on behalf of Professor Keith Hawton (University of Oxford)
• Professor Robin Murray (IoP)
• Dr Rachel Churchill (University of Bristol)
• Professor Hugh Gurling (University College London)

19.   A good debate followed.  Unfortunately there was not enough time remaining for questions on the last two presentations.  CC suggested questions could be sent via email.

Action: group members to email any questions regarding the presentations.

20.  CC thanked the presenters and the group for attending.  She enquired as to whether people thought the meeting has been useful and to the way individuals wanted to proceed e.g. do they want to meet again? CC then brought the meeting to a close.

Action: Members to email CC with their views on how best to proceed.

Dr Ann Adams – Implementing the GRiST clinical decision support system within primary care and the community, to improve detection and management of suicide risk amongst depressed patients.

Prof Keith Hawton – Systematic review of risk factors for suicide in people with depression.

A lay summary of both projects is available in the Projects section of this website.

Funded Research:

Project with University College London via Big Lottery funding on primary care and depression.

Project in partnership with Rethink being undertaken by the Institute of Psychiatry on evaluating the impact of the Time to Change campaign in reducing stigma and discrimination against people with mental health problems. Funded by the Big Lottery and Comic Relief.

Project to evaluate the impact of ECOMinds.

We also undertake smaller pieces of research into the direct experience of people with mental distress.

· Bring together the skills, talents and co-operation of individuals, groups and organisations from both the mental health and environmental arena to deliver sustainable projects that improve rural and urban environments;

· Create an effective, integrated volunteer workforce between the mental health community and the community at large;

· Raise mental health awareness amongst communities and thereby reduce the stigma of mental distress;

· Enable mental health service users (MHSUs) to gain skills, improve employment prospects and contribute and integrate into their local communities;

· Improve the physical and mental well-being of participants through increased physical activity, healthy eating and increased mental health awareness.

Research relating to the scheme:

We have commissioned the University of Essex to evaluate the scheme. The aim of the research is to:

· To assess to what extent Ecominds achieved what it set out to do

· To determine if it achieved other things either planned or unplanned

· To assess if it was value-for-money

· To determine how the scheme could have been improved

Principles of collaboration: Forum for Mental Health Research Funders

1. The charities agree to work together to promote research into mental illness. Some will want to fund research but all will have an interest in research.

2. Any collaborative venture must be transparent as to which charities are involved.

3. If members of the collaboration make research grants together they will use high standards of grant making.

4. We will explore how we register with AMRC.

5. We will share information as we agree on our web site: http://mentalhealthresearchfundersforum.org.uk/

6. Our name is Forum for Mental Health Research Funders.

How we might work together:

• Raising the profile of Mental Health Research and the lack of specialist funding sources.

• Join MHRN Clinical Research Groups

• Provide stories and personal narratives to support hope for people with mental illness and how research will help to give that hope

• Dissemination of research findings

• Database of grants awarded

• Shared peer review systems and scientific advice

• Work with the Science Media Centre

• Work with the mental health charities (eg MIND, Rethink, MHF)

Clair Chilvers

30 January 2011

13.30 Welcome and introductions, Clair Chilvers

13.35-13.45 News from members of the Forum, All

13.45-14.30 Quantitative Research: what questions can it answer? Professor Brian Everitt (IoP)

14.30-15.15 Qualitative Research: what questions can it answer? Dr Joanna Murray (KCL)

15.15-15.30 Tea

15.30-16.30 What would you do with £50k? Each presenter is given 10 minutes to answer this question with a subsidiary question: And suppose it was £100k?

Clair Chilvers on behalf of Professor Keith Hawton (Univ of Oxford)

Professor Sir Robin Murray (IoP)

Dr Rachel Churchill (Univ of Bristol)

Professor Hugh Gurling (UCL)

16.30 AoB and Close

Apologies: Dennis Stevenson, Naomi Garnett, Til Wykes, Nick Rawlins, Gef Flynn

Ann Adams: Principal Research Fellow, Warwick Medical School

Christopher Buckingham: Senior Lecturer in Computer Science, Aston University

Aims and Objectives

The overall aim of this research is to help save lives threatened by suicide risk. This will be addressed using the GRiST Clinical Decision Support System (CDSS) for detecting risks associated with mental health problems (see www.galassify.org/grist). The proposed research will adapt it for use by service users for self-assessment (my-GRiST) and for clinicians in primary care and Improving Access to Psychological Therapies (IAPT) services (GRiST-PC). The research will provide a preliminary evaluation of the effectiveness of these GRiST versions in helping detect and manage suicide risk.

Research goal

To improve the detection and effective management of suicide risk by extending the GRiST mental-health risk assessment tool for use within primary care and the community.

Research objectives

1. Investigate how GRiST can be developed to:

a) provide a self-assessment tool, my-GRiST, for members of the public to use;

b) provide a version of GRiST, GRiST-PC, for primary care (PC) and IAPT services that can share information with my-GRiST and promote a collaborative clinician/service-user engagement with risk assessment;

2. Investigate the ability of my-GRiST and GRiST-PC to:

a) determine how depressed people and their families use my-GRiST to monitor and self-manage risk in the community;

b) determine how clinicians use GRiST to promote thorough and systematic assessment of suicide risk, risk detection, risk management, and ongoing monitoring of service users within the community;

c) help understand how service users’ health and social circumstances influence suicide risk, with a particular emphasis on depression and co-morbidity with chronic conditions such as tinnitus.

Research outcomes

1. A web-based tool to help clinicians and service users detect, monitor, and manage suicide risk.

2. Increased knowledge about the cues and cue combinations influencing suicide risk, especially the roles of tinnitus and depression.

Information about GRiST

GRiST is a web-based decision support system for assessing the risks of suicide, self-harm, harm to others, self-neglect and vulnerability. It is designed to reflect how mental health experts think about and assess risk, because it is based on the elicited expertise of multi-disciplinary mental health clinicians. The GRiST technology contains software simulations of how these experts assess low-level cues (e.g. lives alone), through higher level concepts (e.g. depression, anxiety, anger), to top-level risk categories such as suicide and harm to others. Based on psychological processes, GRiST can fully explain how a set of service user cues generates specific risk quantifications in a way that is intuitive, comprehensible, and resonates with clinicians’ own understanding of risk.

GRiST is unique, because it combines the best of both established approaches to mental health risk assessment: structured clinical judgment (i.e. how clinicians formulate risks using their own expertise and training) and actuarial approaches (i.e. those based on statistical analysis of population data). Each time the GRiST CDSS is used, the risk profile of the person assessed and the risk judgments attached to it by the assessor are stored in the GRiST database in anonymous form. These data permit on-going analysis of risk assessment decisions, which will inform improved clinical practice by highlighting how risk decisions are made: what information they are based on, influences on them (e.g. gender, age) and how they link to outcomes.

The hierarchical model of risk expertise contained by GRiST provides a precise index and structuring of risk information which makes it easy to store and analyse data as well as link it to other information-gathering systems. In addition there is the capacity for people to add narrative information to any part of the hierarchy. For example, there are questions exploring links between physical and mental health. People can use the “comment” boxes attached to them to put in more precise details about how their mental health is affected by, for example, tinnitus.

Since its original development, different clinician versions of GRiST have been created (e.g. for older, working age, and children and young people), and implemented in a number of NHS MH Trusts, all of which can ‘talk’ to each other. The data gathering interfaces have their own specific question wording and order, but because questions are indexed to the underlying risk model, equivalent information is collected. The same approach will be used to create my-GRiST and GRiST-PC.

Below is a summary list of GRiST’s attributes:

1. It is unique amongst multiple risk assessment tools in explicitly modelling structured clinical judgements:

a) obtained from multidisciplinary mental-health clinicians

b) through using rigorous research methods for eliciting consensual clinical expertise

c) and with a clear audit trail demonstrating the evidence base for the risk model.

2. It integrates empirical evidence with structured clinical judgements

a) within a single system

b) using its accumulating database of risk data and judgements

c) in accordance with Department of Health guidance

3. GRiST’s validated model of hierarchical risk expertise links risk data to top-level risks through intervening concepts and

a) provides a precise formal structure and location for each piece of service-user data

b) acts as an index to risk information held in other patient documentation to facilitate its linkage and collation

c) has the potential to populate information in other patient records and avoid double data entry

d) makes it easy to find any piece of information and format it for reports.

4. GRiST is underpinned by a database with sophisticated statistical and pattern recognition tools. On-going analyses will contribute to the research evidence base about:

a) how clinicians assess risk

b) social patterns and inequalities associated with risk assessment

c) risk prediction

d) how cue clusters influence risk.

5. GRiST was developed from the start to exploit the semantic web to give:

a) flexible formatting of information

b) multiple delivery modes and web-based interfaces

c) easy and universal access

d) ongoing resources for rapid adaptation in response to changing clinical and patient needs, government directives, and IT requirements.

6. GRiST was designed as an interactive tool with sophisticated interface functionality to give:

a) streamlined data entry with questions displayed only when relevant

b) no need to re-enter data that will not change (i.e. persistent/historical data) for subsequent patient and self-assessments (e.g. date of first suicide attempt)

c) automatic output of reports from the data entered that can be individually customised for the particular clinical service or patient

d) the facility to add comments or narrative to any piece of risk data or to contextualise risk judgments and self-assessment.

7. There are versions for different populations:

a) with the same validated underlying knowledge structure

b) data easily passed between population groups, currently including:

(1) working-age adults

(2) CAMHS

(3) older adults

8. GRiST is based on psychological knowledge structuring and reasoning processes, which enables risk advice to be explained in ways that:

a) are easy to understand

b) easy to validate

c) support judgements, not blind faith.

9. It is intended that GRiST will explicitly support shared decision making and self-assessment through:

a) the service-user version (my-GRiST)

b) accessibility via IAPT, primary care, and other front-line services or public facilities such as libraries and mental-health charities

10. GRiST represents a common risk language with multiple interfaces for collecting information and providing advice:

a) to reflect the needs of different assessors in different contexts

b) all linked to the same underlying model of risk.

11. GRiST provides a whole (health and social care) system approach to risk assessment that aids risk communication across the entire care pathway, from the community to specialist secondary mental-health services in:

a) different services

b) different care sectors

c) between clinicians, service users and their carers.

12. Finally, GRiST creates opportunities for NHS and other organisations to collaborate on research projects with the GRiST team

a) GRiST is based in two research-intensive universities (Aston and Warwick)

b) GRiST is founded on the philosophy of generating rigorous research evidence for its development coupled with extensive clinical testing at the point of care.

Principal investigator: Professor Keith Hawton, Centre for Suicide Research, Department of Psychiatry, University of Oxford.

Co-investigators: Dr Camilla Haw, Consultant Psychiatrist, St Andrews Hospital, Northampton and Dr Kate Saunders, OHSRC Junior Fellow, University Department of Psychiatry, Oxford.

LAY SUMMARY

The aims of the proposed research will be:

(i) To conduct a systematic review of the international research literature on factors associated with risk of suicide in people with depression;

(ii) To produce a guide for clinicians (especially general practitioners) that will assist them in assessing suicide risk in patients with depression.

The methods used will include;

(i) Searching relevant electronic databases for research reports using the methodology previously employed at the Centre for Suicide Research for similar reviews. Screening of reports for suitability will be done independently by two researchers. The main focus will be on longitudinal and case control studies. Data on potential risk factors will be extracted and entered into an electronic database for subsequent analyses.

(ii) Review of list of papers by international expert group to identify any missing reports.

Risk factors from the following domains will be included: Psychiatric disorders – severity, symptoms, number of episodes, co-morbidity with other disorders, alcohol and substance misuse; Family history of psychiatric disorder and suicidal behaviour; Previous suicidal behaviour (including self-harm); Psychological symptoms (e.g. hopelessness, self-esteem, aggression, irritability); Physical disorders – including current and previous disorders (these will be sub-grouped by specific types of disorders); Previous and current psychiatric and psychological treatment; Help-seeking and compliance with care.

The review will not include studies of genetic and other biological mechanisms, although papers on these topics identified through the search will be collected for possible future reviews.

The data on individual risk factors will be subject to meta-analysis. Statistical advice will be provided by members of the Centre for Statistics in Medicine at Oxford University (which already collaborates with our Centre). It is recognised that some risk factors may be relatively rare, yet of importance to clinicians. Tinnitus would be a case in point. Where reports on such factors are identified but data are insufficient for any sort of meta-analysis these findings will be collated and reported descriptively.