Impact of a Distress Brief Intervention on Suicidal Ideation, Suicide Attempts and Self-harm in the immediate, short and longer term: a mixed method evaluation study protocol

Conceptual framework

The conceptual framework for this study is grounded within the Integrated Motivational-Volitional (IMV) Model of Suicidal Behaviour¹². The IMV model is based on the premise that factors associated with the emergence of suicidal thoughts (e.g., early life adversity, feelings of defeat and entrapment) are distinct from those that influence the transition from ideation to behaviour and understanding the process of ideation to action is crucial to preventing suicide. In the IMV model, volitional moderators (e.g., access to means, planning, impulsivity, past suicidal behaviour) govern the transition from suicidal ideation to behaviour. This is a dynamic process that for many is cyclical in nature, moving from suicidal thinking to attempt and back to thoughts over time, with the time between thoughts and action becoming less, reducing opportunity to intervene¹².

Intervention and study setting

To address the need to improve the response to adults (aged 18+) in distress identified in their Suicide Prevention and Mental Health strategies^13,14, the Scottish Government launched the Distress Brief Intervention (DBI) programme in 2016 (https://www.dbi.scot/). The aim of the DBI programme is to enable and improve inter-agency co-ordination and collaboration to deliver an effective response for people in distress to improve their experiences of support and their outcomes.

The DBI programme was tested and developed across Health Boards in four pilot sites in Scotland (Aberdeen, Inverness, Lanarkshire, and the Scottish Borders). A mixed-methods realist evaluation of DBI with adults (>=18 years) in the four pilot sites demonstrated that the intervention was feasible and acceptable to practitioners and to people receiving the service⁶.

The DBI pilot programme brought together local NHS (primary care and emergency departments) and voluntary DBI provider partnerships, national agencies (Police Scotland, Scottish Ambulance Service, and NHS 24), voluntary sector mental health organisations and University of Glasgow with six delivery teams operating in four regions (Aberdeen, Borders, Inverness, and Lanarkshire). In 2019 DBI was extended to those aged 16 years and over and is on course to be embedded in each of Scotland’s 31 Health and Social Care Partnership areas during 2024¹⁵. To enable Scotland-wide access for people in distress during the COVID-19 pandemic, in 2020 the NHS 24 Mental Health Hub was added as a referral route to DBI in all Health Board areas.

DBI comprises two levels. DBI Level 1 is provided by trained front-line staff (Primary Care, Police Scotland, Scottish Ambulance Service, Emergency Departments and NHS 24) who offer a compassionate response to individuals in distress and, where necessary, provide onward referral to a Level 2 service. DBI Level 2 is provided by trained voluntary sector practitioners. Individuals referred from Level 1 are contacted by a DBI practitioner within 24 hours and offered up to 14 consecutive days of community-based, person-centred, solution-focussed support from a DBI practitioner. DBI Level 2 explores the nature and cause of an individual’s distress and how it can be managed to prevent future distress. Service users can co-create a Distress Management Plan (DMaP) with their practitioner (a previous evaluation of DBI suggests many individuals were using their DMaP three months on from DBI)⁶. If appropriate, DBI Level 2 service users can be signposted to and/or receive support to engage with further community or statutory service support.

An evaluation of the implementation and impact of the DBI pilot programme found that pre-post measurements of distress (CORE-OM 5¹⁶) indicated that most participants’ levels of distress reduced following DBI⁶. Qualitative evidence also indicated that without DBI, approximately 10% of study participants felt they would have gone on to die by suicide⁶. There is, therefore, early evidence that the DBI programme appears to be filling a critical support gap between unscheduled care, emergency service response and suicide attempts and may help to prevent or break the escalation of crisis and cycles of suicidal behaviour. This current study aims to understand whether and how DBI can reduce suicidal ideation, suicidal behaviour, and self-harm among those presenting to front line services in distress and crisis.

While little is known about the impact of brief interventions on suicidal ideation, the DBI programme evaluation findings tentatively suggest that the DBI intervention could play a role in the prevention of suicidal behaviour⁶. This study will make a significant contribution to filling this knowledge gap. The findings would support the improvement of the current DBI intervention and its wider roll-out to the UK and beyond and inform future development of integrated mental health care and suicide prevention policy and practice.

The setting for this study is Scotland, the second largest country in the UK with a population of around 5.5 million. Nearly 96% of Scotland’s population report their ethnicity as ‘white’. Scotland has the highest suicide rate in the UK, with 762 probable suicides in 2022². Rates of attempted suicide have risen among adults (aged over 16) in Scotland from 4% of Scottish adults in 2008/2009 to 7% in 2021/2022¹⁷. Rates of self-harm among adults have also increased over time in Scotland, from 2–3% in 2008/2009 to 10% in 2021/2022¹⁷.

Between October 2022 and September 2023, the DBI programme received an average around 1500 referrals a month, 60% of which were female. Most Level 2 referrals are made by Level 1 trained staff in the NHS 24 MHH and in-hours primary care. Nearly 70% of referrals are for people living in Scotland’s five most deprived deciles (according to the Scottish Index of Multiple Deprivation)¹⁸. DBI data from the year ending September 2023 indicates that the most common presenting problems are stress and anxiety (68% of referrals), depression and low mood (64%), suicidal thoughts (36%), self-harm thoughts or behaviours (7–9%), and suicidal behaviour (5%). The most reported contributory factors are relationship issues (41% of referrals), emotional wellbeing (29%), life coping issues (28%), employment issues (18%), and money worries (17%)¹⁹.

Patient and Public Involvement

The development of the DBI programme was heavily influenced by people with experience of distress (including those who have been at risk of self-harm or suicide and the LBGTQIA community) via a two-year national engagement programme. Previous DBI service users, including those who had experienced suicidality, have reviewed the study design to establish how and when people with experience of DBI could effectively be involved in the study. This helped to refine the study’s PPI plans and reduce barriers to participation (e.g., worries about experiencing distress from reconnecting with issues related to DBI or being asked to participate in PPI at an inappropriate time). It also influenced these changes to the study design:

PPI is integral to this study; we will invite up to 12 people (seeking to be inclusive of BAME and other minority groups) who have used DBI or have experience of distress, suicide and/or self-harm thoughts or behaviour to join a Study Advisory Group (SAG) meeting nine times during the study (usually online but other modes will be offered). Recruitment will be rolling to allow for drop-out and for those with more recent experience to join. The SAG will be coordinated by the study PPI lead and a paid PPI Champion and PPI co-lead role will be created to support new members and assist in SAG coordination.

Study design

This is a mixed method evaluation which incorporates a combination of quantitative and qualitative measures of DBI impact and continuous improvement elements. This 36-month study combines: longitudinal surveys and in-depth interviews with individuals up to a year following their use of the DBI service; interviews with a qualitative comparator group; comparative analysis of nationally available health outcome data for a DBI service user group and retrospective comparator group; focus groups with providers involved in the delivery of DBI; and a survey of GPs to develop an economic model of care pathways available to GPs to support people in distress and at risk of self-harm.

Research questions and objectives

The overall aim of this study is to understand whether and how DBI can reduce suicidal ideation, suicidal behaviour and self-harm among those presenting to front line services in distress and crisis.

In line with the study aim, the research questions are:

Primary and secondary outcomes

Primary outcomes are changes in suicidal ideation, suicidal behaviour and self-harm reduction in the immediate (1 month), short (3–4 months) and longer term (1 year).

Secondary outcomes are changes in resilience, self-stigma, mental and physical health in the immediate (1 month), short (3–4 months) and longer term (1 year) and health service usage one month and one year prior to and following DBI.

Study governance

The study will be overseen by a steering group comprising representatives with academic expertise in complex health evaluations and economic evaluation, and representatives from services that act as DBI Level 1 providers (e.g., the Scottish Ambulance Service, Police Scotland, and NHS 24 MHH).

Data collection

DIMES is a mixed methods study that makes use of data from a variety of sources to address the research questions, as outlined in Figure 1.

Figure 1. DIMES study data collection, linkage and analysis flow diagram.

Quantitative data collection from individuals accessing DBI

Quantitative data from individuals accessing DBI will be collected at up to five time points to explore whether DBI helps those who present in crisis and distress with current or previous suicidal ideation, suicide attempts or self-harm achieve better outcomes in the immediate, short and longer term and whether their outcomes differ for other DBI service users, and which aspects of the DBI intervention contribute to these different outcomes. Data will also be collected on health and social care use and third sector resource use to estimate the impact of DBI on service users care use (RQs 1, 2, 3, 4, 5 and 9).

The data will be collated from three sources: service user surveys, routine service delivery data collected by the DBI service and Public Health Scotland’s Unscheduled Care Data Mart.

DBI service user surveys

Individuals accessing the DBI Level 2 service will be invited to complete four Level 2 DBI service user surveys. These surveys will be administered by DBI Level 2 staff (at the first and final DBI Level 2 sessions) and the ScotCen research team (3–4 months and 1 year) by email, text or paper.

The survey instruments include a combination of validated scales and bespoke closed questions as well as a small number of open questions to measure the following:

Participants will receive a £10 high street shopping voucher as compensation on completion of each of the surveys.

DBI routine dataset

The DBI routine dataset comprises the following: DBI service user socio-demographic characteristics, service usage, including length of support, number of sessions and presenting problems.

Unscheduled Care Datamart

The Unscheduled Care Data Mart (UCD) is a collaboration between Public Health Scotland (PHS), NHS 24 and Scottish Ambulance Service (SAS). The data mart links data from NHS 24, Scottish Ambulance Service, Out of Hours Primary Care, Emergency Department, Acute, Mental Health and Deaths to show a patient journey for all those with a valid Community Health Index number.

The linkage of outcome data with service user characteristics and DBI intervention activity and unscheduled care use will facilitate analysis of service user sub-groups and intervention factors and to examine differences in unscheduled care use one year before and after DBI.

Individual level data linkage between the DIMES participant survey dataset, the routine dataset collected by DBI and the Unscheduled Care Datamart (UCD) will be undertaken in the Scottish National Safe Haven.

Recruitment and sampling: All individuals accessing DBI in the first data collection year at participating DBI provider sites aged 16 years or over will be invited to participate by trained DBI Level 2 staff. DBI service users under the age of 16 or for whom participation in the study is not deemed appropriate by DBI staff, depending on individual circumstances, will be excluded. Based on the previous evaluation (pre-Covid) the estimated Level 2 DBI service users’ survey sample size is a maximum of 2,700 respondents with anticipated attrition at each time point (n=2,700 for 1^st and n=1350 for 2^nd DBI surveys, n=675 for 3–4-month follow-up survey, n=330 for 1 year survey).

Consent: A participant information sheet and privacy statement containing full study details will be provided to participants by DBI L2 staff. Consent to take part in the study and to link data to both the DBI dataset and UCD data is collected and recorded at the start of the first survey.

Analysis: Data analysis of quantitative data collected via surveys, routine data and unscheduled care will include descriptive statistics and crosstabulation analysis with logistic regression if appropriate to track and compare changes in outcomes at the different time points up to one-year post-DBI. Demographic, referral source and presenting problem characteristics of survey respondents and non-respondents will be compared to assess the representativeness of the survey sample and whether any weighting adjustments are required.

Qualitative interviews with individuals accessing DBI and a qualitative comparator group

To gain deeper insights into the outcomes for those who access DBI with current or previous suicidal and/or self-harm ideation or attempts and whether, how and why DBI contributes to these different outcomes (RQs 1, 2, 4 and 5), qualitative interviews will be undertaken with a sample of those who access DBI and a qualitative comparator group.

Qualitative interviews with a sample of DBI service user participants with a history of suicidal risk based on quantitative survey responses will be undertaken at 1 month, 3–4 months and 1 year following DBI L2 intervention. Based on quantitative survey responses, potentially eligible participants will be contacted by the research team to conduct a safe screening process to further assess their eligibility to take part in interviews. The safe screening follows a standard protocol, developed by the Suicidal Behaviour Research Laboratory and has been used in many previous studies with suicidal and vulnerable groups. The safe screening protocol includes:

During an initial Safe Screening telephone call, a suicide risk assessment will be carried out. A risk mitigation strategy proportionate to the level of risk will be made; all participants will be encouraged to develop or maintain a safety plan and provided with a sources of support sheet. During the screening call, eligibility criteria are checked. Risk assessment and mitigation protocols are followed prior to and following any subsequent interview calls.

The eligibility criteria for inclusion in the interviews are:

The semi-structured interviews will be conducted face to face, by video or telephone and last for approximately one hour. Interview guides will be used to explore:

To provide a comparator with those accessing DBI, qualitative insights from users of the NHS 24 MHH between October 2019 and March 2020 will be gathered via Scottish lived experience networks. This group provides a comparison with DBI as the NHS 24 MHH was the most similar national service available for individuals experiencing distress prior to the roll out of DBI via NHS 24. Similar screening, risk assessment and mitigation protocols used with DBI service user interview participants will be employed for this group of participants. Unlike DBI service user interviews, those conducted with the comparator sample will only be conducted once.

The interviews will cover the individual’s story from referral to exit from the NHS 24 MHH and will address a subset of issues explored with the DBI service user participants focused instead on experiences of NHS 24 MHH. Participants will receive a £25 thank you voucher as compensation for their time following each interview.

Recruitment and sampling

Individuals accessing DBI: Evaluation participants aged 16 or over who have agreed in their second survey (issued at their final DBI L2 session) to be contacted for interview, and who have experienced suicidal thoughts, behaviour or self-harm at any time in the last 12 months are eligible. Full eligibility criteria are listed above. A maximum 75 people will be recruited.

Qualitative comparator group: Individuals who contacted the NHS 24 MHH between Oct 2019 and March 2020 for reasons of mental health or distress, who have experienced suicidal thoughts, behaviour or self-harm and who have been assessed as eligible during the Safe Screening process will be included. The sample of 15 interviews is based on the premise that the open invitation to participate through mental health lived experience networks will yield small numbers, however 15 interviews should be sufficient to provide the insights required for this study.

Consent: A participant information sheet and privacy statement containing full study details will be provided to participants in advance of their interview and verbal consent will be collected and recorded.

Analysis: Qualitative interviews will be summarised, charted and coded using QSR NVivo 12 and analysed with reference to techniques of framework analysis²⁶. We will explore any differences or similarities in the views of the DBI and qualitative comparator participants to assess any perceived impact of the DBI programme, while paying attention to other contextual influences.

Comparator group data collection

A comparator group of those who accessed NHS24 for mental health distress reasons prior to the introduction of DBI will be identified. The use of unscheduled care (as determined by UCD data) in the year before and after the selected year for the comparator group and the data collection year for DBI service users will be compared to assess whether use of unscheduled care in the year after is different for DBI service users (RQ 8).

Following approval from the NHS Public Benefits and Privacy Panel (PBPP), people who received DBI between Jan 2022 and Jan 2023 will be identified through the DBI database held by PHS which will be linked to the UCD by eDRIS. As noted above, a retrospective comparator group will be identified using the UCD to undertake a comparison with the DBI service user group. We will select the comparator group from those accessing the NHS 24 Mental Health Hub (MHH) who could not be referred to DBI because they accessed the MHH prior to DBI’s national roll out throughout Scotland via NHS 24. We will request UCD records from 2018–2021 for people who contacted NHS 24 in October 2019 to March 2020 for mental health reasons. The NHS 24 data indicates whether a call went to the NHS 24 MHH. A call related to mental health will be used as the comparator event. The comparator event(s) is/are a good proxy for presenting to unscheduled care experiencing mental distress.

Information will be collected on:

Sampling: UCD records for 2018-2021 for people who contacted NHS 24 in October 2019 to March 2020 for mental health reasons will be requested. For the comparator group quantitative analysis, the sample size is based on 2500 people accessing the DBI L2 during the study period.

Assuming the model: generalised linear model using binomial family and a logit link with dependent variable ‘UCD use in the year after’ and independent variables ‘UCD use in the preceding year’ and ‘group’ (DBI versus comparator).

Estimating the following odds ratios for the model: intercept 0.80, group 0.70, UCD use in preceding year 2.0 and assuming the DBI group represents 10% of the total analysis sample and that the probability of UCD use in the preceding year is about 70% (estimate based on evaluation data) with alpha set to 0.05 we would have over 90% power to test that the odds ratio for ‘group’ is different to 1.

Analysis: We will use doubly robust estimation which combines a regression model of the outcome with a model for the exposure (i.e., the propensity score) to estimate the effect of an exposure on an outcome. The use of unscheduled care in the year before and after the selected year for the comparator group and the data collection year for DBI service users will be compared to assess whether use of unscheduled care in the year after is lower for DBI service users.

The comparative analysis will have value as an approximation of the impact of DBI on future use of unscheduled care. We would expect unscheduled care use to be lower in the year after the index event than in the year before due to regression to the mean. If DBI L2 reduces emotional distress and increases capacity for self-management then we would expect the reduction in unscheduled care use (particularly for mental health, self-harm, suicide attempt and completed suicide) to be greater for those receiving DBI intervention than in the historical comparator group. In addition, we will compare, within people who are referred to DBI, whether changes in unscheduled care use are different for those referred for reasons of suicidality or self-harm compared to those who are not. This analysis will include adjustment for potentially confounding sociodemographic variables. Secondary analyses and likely zero inflation will be assessed.

Focus Groups with DBI staff

To explore from the DBI provider perspective ways in which DBI might improve its contribution to positive outcomes for people who present with suicidal ideation/ suicidal behaviour/self-harm (RQs 6 and 7), and whether those accessing DBI with suicidal ideation/suicidal behaviour/self-harm would benefit from a ‘DBI Level 3’ to provide follow-up over a longer period and how this might be implemented.

Seven focus groups will be held with DBI provider staff representatives. Six focus groups will be held across DBI provider sites in Scotland with local DBI Level 2 staff members, DBI Level 1 service representatives (police, ambulance and A&E) and management and one will be held with representatives of the DBI Central Team leadership and DBI Programme Board. These focus group discussions will explore staff and programme leadership and board perceptions of:

Recruitment and Sampling: Those working in or with DBI Level 1 and Level 2 would be eligible to take part in the focus groups as well as DBI local provider management, DBI Central team leadership and DBI Programme Board members. Sample size=42 – 56 (6–8 per focus group), one focus group per site.

Consent: Full study details will be provided to participants in advance of their interview and verbal consent will be collected and recorded.

Analysis: Focus groups will be analysed in NVivo using a framework matrix to explore themes across the data²⁷.

Economic modelling of care pathways

To support the future commissioning of DBI, more evidence is needed on the care pathways of people who receive a DBI L2 intervention and their resource use within the health care, social care and third sector. For people who present in distress to GPs with suicidal ideation, suicidal behaviour and self-harm, there is no clear pathway for how they are currently managed within the health care system (RQ 10). This information is important if we seek to consider the impact the role of the DBI L2 route could have in the stages before unplanned access to care/support as well as following such care/support.

A survey will be undertaken with a sample of up to 20 GPs in NHS Boards who have been trained to provide Level 1 DBI to determine the typical care pathway for individuals in distress and at risk of self-harm. The GP route has been selected as the previous evaluation of DBI pilot sites indicated that at that time 40% of referrals were from primary care in-hours services⁶. This indicates that DBI presents a possible complement to GP care and therefore, understanding the different resources required for the range of care pathway options open to GPs would be important in the future delivery of the DBI service.

Analysis: Data from the GP survey will be used to model typical care pathways options, including the role of DBI Level 2 in the pathway. Resource use and costs of pathways will be presented. Units of each item for all care pathways and for participant resource use will be collated and presented. These pathways will not consider individual level resource use.

Data synthesis

A summative data synthesis will be undertaken to draw together the above through a process of triangulation to develop evidence-based insights and conclusions. Throughout the study, the study team and a study advisory group will meet to share thematic and theoretical insights from data collection and analysis and develop interpretive connections and points of synthesis. Policy and practice recommendations will be made.

Challenges

The study was designed before Covid-19 and began when the final social distancing measures were being lifted. Lockdown had a knock-on effect on DBI in terms of a switch in the mode of service delivery to remote (at the time expected to be temporary); rapid expansion of the service to meet growing demand as the stresses of Covid-19 affected increasing numbers of people; more people presenting to frontline services with higher severity of distress than pre-Covid; and DBI workforce challenges. While these circumstances highlighted the vital need for this study to go ahead, they also presented significant challenges for the study team to ensure ethical and safe support for DBI staff, as well challenges for the recruitment of people presenting to DBI in distress in an unscheduled way.

Limitations

It was not possible for scientific, practical and ethical reasons to conduct a randomised controlled study which is a limitation of this study. Prior to the introduction of DBI, there was no standard response or care pathway for people who were in distress or suicidal, but many would present to ambulance, police or A&E, meaning that there were no existing approaches in place which could provide an effective comparison to DBI. However, our mixed-methods design which includes robust qualitative impact combined with QED elements brings multiple benefits. For example, by using different approaches to determining impact, we can analyse not just whether intended outcomes were realised (or not) but how and why, and such analyses will support the interpretation of the quantitative findings.

DBI makes use of a wide network of voluntary and statutory organisations to sign-post DBI service users to but do not follow-up with these services. It would not be possible to comprehensively track post-DBI service use of these services and therefore the study will be reliant on DBI routine data on referrals made and self-report by service users of uptake and perceived outcome.

Strengths

This study is focused on generating high-quality and evidence-based knowledge about the impact of brief interventions on those experiencing suicide and/or self-harm thoughts and/or behaviours through the study of the highly innovative and unique DBI Programme. The study design, like DBI itself, is grounded in lived experience and takes a real-world approach to understanding whether and how DBI makes a difference, using a range of methods. It considers whether, how and why DBI works for people in distress with different needs and circumstances. The collection of longitudinal data over a 1-year period and analysis of in-depth qualitative impact evidence is a key strength of this study, supporting its aim to understand the mechanisms of change involved in DBI. Another strength is our ability to explore the influence of contextual factors on the delivery and potential effectiveness of DBI. To maximise impact, utilisation of interim and end of study findings will be encouraged through participatory learning events with service users, DBI L1 and L2 staff, national and local policy makers and academics.