To assess the feasibility of conducting a randomised controlled trial (f-RCT) of BRIDGE, a brief intervention programme for young people (age 14-24) with BPD features (sub- or full-threshold), in a community sample of Glasgow, Scotland.

Our study design is a single-blind, parallel groups f-RCT following the Medical Research Council Complex Interventions Framework (27) over a period of 36 months. After a two-stage screening process (details below), eligible participants meeting diagnostic criteria for BPD features are invited to baseline research assessment, then randomised to BRIDGE and service as usual (SAU) or SAU (alone).Aspects of feasibility are explored using qualitative and quantitative methods.

The study is set in the city of Glasgow, Scotland’s largest and most ethnically diverse city with an estimated population of 635,130, of which almost 12% represent an ethnic minority. Adolescents and young people represent approximately 12% of the Glasgow population; over half live in areas in the two most deprived deciles within Scotland, while only 4% Glaswegian children live in areas in the least deprived decile within Scotland (https://www.gov.scot/collections/scottish-index-ofmultiple-deprivation-2020/).

Recruitment is taking place over 24-months, from October 2021 to September 2023. As per recommended Medical Research Council guidelines for f-RCTs (27), a formal sample size calculation is not appropriate. Our pilot work (29) suggests that we will achieve a sample size of forty adolescents and young people (age 14-24), with subthreshold or full-threshold BPD features, recruited from the community and randomised to each of the two arms of the trial. This sample will allow us to estimate the sample size for a phase III trial and the human resources required to achieve recruitment and retention.

We will also explore the feasibility of different routes of referral to the trial. Participants might or might not be involved with a mental health service provider to meet criteria for trial. However, there will be an expectation that all participants will be registered with a general practitioner (GP).Where a participant has no fixed abode, a GP will be identified within the homelessness GP team. Participants will be recruited through ethically approved channels that might include:

Professionals within each of the above services will be made aware of the study through presentations with the trial team. They will then be regularly asked if any patients on their caseload/service users meet inclusion criteria for the study and can be approached for participation. With potential participants, verbal consent will be sought and noted before referring for further participation with the clinical trial.

Symptomatic remission, occurring over a number of years, is a common feature of BPD (30). Yet, there is robust evidence to support that young people who meet subthreshold criteria for persistent, long-standing difficulties characteristic of BPD features, also have psychiatric comorbidity and functional impairment and are likely to be referred to mental health services for disruptive or suicidal behaviour (31). To address the heterogeneity inherent in psychopathology (32) and within subsyndromal BPD (33), screening for randomisation of participants to BRIDGE intervention (plus SAU) or SAU would include a) enhanced phenotyping at phase 2 (below) which will ensure adequately defining the screened population and b) a robust measure of social function (i.e. KIDSCREEN-10) (34).

Brief, intensive assessment and integrated formulation (BRIDGE) intervention development has been guided by the emerging evidence base for community based comprehensive assessment and brief intervention with young people with longstanding unmet needs characteristic of BPD features.

A fidelity evaluation framework will be optimised to the specific needs of this complex intervention (37) with the focus on a) participant responsiveness as engagement and retention is the primary focus of the trial b) complex intervention adherence (Frequency, duration and nature of engagement) c) quality of delivery (protocol adherence, monitoring assessments and treatment delivery) d) identifying barriers and facilitators (recruitment, engagement, retention) e) outcome measurement of change and sustainability. Trial therapists will have two-weekly supervision with a senior CAT practitioner and supervisor, accredited by the Association of Cognitive Analytic Therapists (ACAT), with extensive experience of working with young people with BPD.

For participants randomised to SAU, a routine letter acknowledging their participation will be shared with their service provider(s), including the GP. SAU, which is likely to range from social services, mental-health services, forensic services, to no intervention, will be documented for each participant. The intervention will be described in detail through the qualitative process evaluation.

There will be no formal measure of treatment fidelity for SAU.

Our three-month pre-trial set-up phase will establish a detailed system for managing referrals, randomisation, and networking with likely referral sources, including any self-referrals detailed in

Figure 1 . All potential participants will receive a ‘Participant Information Leaflet’ and a ‘Trial Consent Form’ prior to Screening 1. The trial research team will then invite young people meeting criteria at screening 1 for an assessment (Phase 2) to determine eligibility for subthreshold or full-threshold BPD features. The assessment will be conducted by a clinically trained researcher, prior to study randomisation.

Young people (age 14-24) or parents (for those aged 14-16 who are judged by a healthcare professional, i.e. GP, as incapable of providing written, informed consent) with child assent will complete the Participant Information sheet (online or paper format) and consent form. Capacity to consent will be confirmed and documented by a healthcare professional at diagnostic screening 2, prior to randomisation.

Randomisation will be managed by the Trial Team at the University of Glasgow. Participants will be randomised (1:1) to either BRIDGE (+SAU) or SAU. A statistician from the Robertson Centre for Biostatistics at the University of Glasgow will generate a randomisation table. Research outcome assessors will be blind to treatment allocation. The randomisation list, containing the unique study ID and participant name, will be stored on a password protected file within the Trial Team’s private.

Microsoft Teams channel (following sponsor representative GDPR guidelines, at the university of Glasgow secure server). Access will be restricted only to those members of the research team who will remain unblinded throughout the study. A log of access will be maintained. Participants will be randomised by a member of the administrative team independent of the trial, using the randomisation procedure provided by the study statistician.

All participants consenting to participate will be invited to complete assessments, post-randomisation, at baseline, 12- and 24- weeks post-randomisation. A minimum data set of care pathways will be described at baseline. Follow-up of all participants will occur for the purpose of future modelling of health economics impacts in relation to quality of life and functional outcomes.

Case-study methodology will be used to understand individual contexts, involving embedded qualitative evaluation with about 10- 15 young people (five to seven from BRIDGE intervention and five to seven from SAU’s) and, where possible, their parents and relevant service providers. Topics will include acceptability of the three elements of BRIDGE, perceived mechanisms of change and exploration of requirements for data collection for a future d-RCT, including for future health economic evaluation.

The only quantitative outcome measures in the f-RCT are recruitment (i.e. proportion of young people referred as eligible who are consented and randomised) and retention (i.e. proportion of those randomised who complete follow-up assessments at six months post randomisation).

We have examined the literature and best practice guidelines to select likely outcome measures for use in a future definitive RCT (d-RCT) (e.g., the International Committee for Harmonization of Outcome Measures for Personality Disorder (38). The f-RCT will investigate the acceptability and appropriateness of these measures for a future d-RCT which include, at screening 1/pre-randomisation: The Structured Clinical Interview for DSM‐V Axis II for Personality disorders (SCID-II PQ-BPD) consists of 15 questions which explore the presence of BPD symptoms using ‘Yes’ or ‘No’ answers. The SCID-II PQ-BPD has been found to have fair-to-good agreement with a SCID-II interview BPD diagnosis and good test–retest reliability in young people (35).

Should an individual score a minimum of 11 of the 15 questions they are invited to the Structured Clinical Interview for DSM‐V Axis II Personality Disorders (SCID‐5-PD) BPD module. This is a semi-structured interviewed which explores these 15 questions in depth measuring on a scale of?-3 to what extent an individual’s symptoms impact on their functioning. (? = inadequate information, 1= absent or false, 2 = subthreshold, 3 = threshold or true. A pervasive pattern of instability in relationships, self-image, mood, impulsivity, anger and the occurrence of dissociation are explored. Although the DSM-5 (39) used the criteria that five of these categories are met, for the purposes of BRIDGE, the presence of three or more BPD traits will be used as diagnostic inclusion criteria, as identifying early signs and symptoms is the objective.

In this second screening further measures are carried out. KIDSCREEN 10 (34) is a health questionnaire for children and young people, exploring how they have perceived their health to be in the previous week. The KIDSCREEN 10 (34) is a standardised assessment tool for health and its impact on quality of life, demonstrating good internal consistency (Cronbach’s alpha =.82), and good test-retest reliability (r=.73; ICC=.72) (40).

Participants will complete the following measures pre-randomisation: Pathways to care and demographic questionnaire, Mini International Neuropsychiatric Interview (MINI) 7.0.2 (18yrs+) and the Mini International Neuropsychiatric Interview (MINI-KID) 7.0.2 (under 18yrs) (41); The MINI is a diagnostic, structured tool for assessing common psychiatric disorders. Studies (42) have shown that the MINI has psychometric reliability and validity and can be delivered with short training input and in a brief period of time. Adverse Childhood Experiences Scale (ACES) (43) asks 10 questions regarding traumatic experiences that occurred before the age of 18. Autism Symptoms adolescents (ASSERT) (44)is a self-report tool with 8-questions pertaining to poor social understanding, rigidity and repetitive behaviours and interests. Adult ADHD Self-Report Scale (ASRS) (45) is an 18-question screening tool used by BRIDGE to screen for attention deficit disorder symptoms such as feelings, conduct and level of impairment in the previous 6 months. It has 2 subscales; 1) Inattention and 2) Hyperactivity. It has been found to be a reliable and valid scale, showing high internal consistency (Cronbach’s alpha 0.88) and rater-administration (Cronbach’s alpha 0.89) (46).

At baseline, 12- and 24- week follow-up, assessments include: Borderline symptoms list (BSL-23)

(ref) This 23-item list explores the impact of BPD symptoms in the course of the previous week (47).

The questions explored include concentration, emotions, vulnerability and overall daily functioning. A BSL supplement is also used in BRIDGE which assesses behaviours over the course of the previous week. The BSL-23 has excellent psychometric properties, high internal consistency (Cronbach = 0.97) and test-retest reliability (0.82) (48). Sheehan’s Disability Scale (SDS) (49) is a self-reporting tool which assess the impact of symptoms on functional impairment. There are three scales, 1) Work*/school, 2) Social life, 3) Family and home responsibilities. Days lost and days unproductive due to symptoms are also rated. Barratt Impulsiveness Scale (BIS-11) (50) is consists of 30 questions which consider 1) Attentional Facet, 2) Motor Facet, 3) Planning Facet. The Cronbach alphas range from 0.73 to 0.83 indicating good reliability. Difficulties in Emotional Regulation Scale-SF (DERS-SF) (51) consists of 18 questions and measures how an individual relates to their emotions such as understanding, awareness and acceptance. In terms of validity and reliability the DERS-SF had a Cronbach-Alpha range of 0.90-0.98 and 81-96% shared variance. Patient Health Questionnaire-9 (PHQ – 9) (52) is a self-report tool of 9-questions which monitors the severity of depression. It measures areas such as sleep, low mood and appetite. Quality of Life Questionnaire (EQ-5D 5L) (53) is a short self-report tool which asks participants to measure their health on that day on a scale of 5 levels - “no problems, I slight problems, moderate problems, severe problems and unable to carry out activity”.

Quality of Life will be assessed with ICECAP-A (54), with a Cronbach’s Alpha of 0.86. Suicidal

Ideation Scale (SIS) (55) is a 10-question 1-5 scale (1 being “never or none of the time and 5 being “always or a great many times”) which can detect early suicidal ideation among adolescent and young people in the general population. It was found to have excellent internal consistency and test-retest reliability.

The feasibility and acceptability of this battery will be explicitly assessed as one of the trial’s key feasibility parameters: qualitatively, we will examine the acceptability and appropriateness of these outcome measures in terms of whether or not participants think we are asking the right questions and whether the measures are acceptable or too burdensome. Prior to final decisions about use of measures in a future d-RCT, the measurement battery will also be discussed in the expert Scientific Advisory Group and Youth Advisory Group.

All the measures will be incorporated into a user-friendly questionnaire “book” and participants will have the option of completing on paper, in a telephone interview, or a mixture – a technique that has worked well in previous studies. These measures will be conducted by trained researcher assessors, blind to group status, at 12- and 24-weeks post-randomisation.

Statistical analyses will be conducted by a statistician/researchers who are blinded to treatment allocation. Recruitment, retention rates and the BRIDGE sessions attended will be calculated with 95% confidence intervals (CI). A simple descriptive analysis of recruitment/retention relative to eligible/approached population will be conducted. While underpowered, between-group change in each measure after adjustment for baseline will be estimated using analysis of covariance (ANCOVA). The emphasis will not be on reporting of significance but variance in outcome measures between groups by reporting mean differences based on change from baseline with 95% CI and effect sizes reported. The standard deviation of the assessments for the putative outcome measures at baseline and follow-up will be estimated, shaping the future d-RCT. Statistical analysis plan and data sharing agreement will be agreed by the Scientific Advisory Group in consultation with the NHS Scotland’s Mental Health Network PPI group.

Interpretative Phenomenological Analysis will be used to explore participant’s experience of engagement and retention with the BRIDGE project. Process evaluation of the contexts and experiences of the young person presenting to the BRIDGE project will enable us to develop a framework for context‐mechanism‐outcome configurations (CMOCs) (56) that will be unique to the young person’s experiences and the multi-agency engagement (or the lack of) around them (57). This would involve qualitative interviews with ten to fifteen young people and, where possible, their parents and relevant service providers. Topics will include acceptability of the three elements of BRIDGE intervention, perceived mechanisms of change and exploration of requirements for data collection for a future d-RCT, including for future health economic evaluation.

For the economic analysis, data will be collected on cost of delivering the intervention, in addition to participant’s use of health, personal, and social services and broader educational and societal resources in the six months before study baseline data collection. Adolescent and young adult’s quality of life will also be measured within trial using the ICECAP-A (54). Measurement of quality of life is an important input for the economic component of this study and will enable assessment of any short-term change in quality of life for young people between baseline and 6 months, and also between the trial arms. The clinical and service use characteristics of young people referred to the study will be used to build a health economic logic model for potential lifetime impacts of BRIDGE.

This proposal builds on our previous Pathways study (29), (2016 – 2018; funded by NHS GGC): majority of recruits came from Child and Adolescent Mental Health Services (CAMHS); recruitment barriers included client disengagement with specialist services before meeting the research team, or clinicians considering patients as high risk to take part. Yet, once consented to the study, there were no withdrawals and the completion rate of the assessments was high, with incomplete data on only 4% of participants. Our work with Pathways has shown that multi-modal recruitment methods, with excellent service user consultation, can achieve high recruitment rates of young people considered hard-to-reach. In the BRIDGE project, we expect similar barriers to engagement/referral from services. We expect data from eligible participants to demonstrate functional difficulties across different facets of life in those presenting with subthreshold and threshold BPD. The development of workable multi-agency services proposed in the trial would be a major advance in optimising care pathways, regardless of trial outcome.