Article Text

Protocol
Observational prospective study of social media, smartphone use and self-harm in a clinical sample of young people: study protocol
  1. Amanda Bye1,
  2. Ben Carter2,
  3. Daniel Leightley1,3,
  4. Kylee Trevillion4,
  5. Maria Liakata5,6,
  6. Stella Branthonne-Foster7,
  7. Grace Williamson1,
  8. Zohra Zenasni2,
  9. Rina Dutta1,8
  1. 1 Psychological Medicine, King’s College London Institute of Psychiatry Psychology and Neuroscience, London, UK
  2. 2 Biostatistics and Health Informatics, King’s College London Institute of Psychiatry Psychology and Neuroscience, London, UK
  3. 3 King’s Centre for Military Health Research, King’s College London Institute of Psychiatry Psychology and Neuroscience, London, UK
  4. 4 Health Service and Population Research, King’s College London Institute of Psychiatry Psychology and Neuroscience, London, UK
  5. 5 School of Electronic Engineering and Computer Science, Queen Mary University of London, London, UK
  6. 6 The Alan Turing Institute, London, UK
  7. 7 No affiliations, London, UK
  8. 8 South London and Maudsley NHS Foundation Trust, London, UK
  1. Correspondence to Dr Rina Dutta; rina.dutta{at}kcl.ac.uk

Abstract

Introduction Young people are the most frequent users of social media and smartphones and there has been an increasing speculation about the potential negative impacts of their use on mental health. This has coincided with a sharp increase in the levels of self-harm in young people. To date, studies researching this potential association are predominantly cross-sectional and reliant on self-report data, which precludes the ability to objectively analyse behaviour over time. This study is one of the first attempts to explore temporal patterns of real-world usage prior to self-harm, to identify whether there are usage patterns associated with an increased risk.

Methods and analysis To study the mechanisms by which social media and smartphone use underpin self-harm in a clinical sample of young people, the Social media, Smartphone use and Self-harm in Young People (3S-YP) study uses a prospective, observational study design. Up to 600 young people aged 13–25 years old from secondary mental health services will be recruited and followed for up to 6 months. Primary analysis will compare real-world data in the 7 days leading up to a participant or clinician recorded self-harm episode, to categorise patterns of problematic usage. Secondary analyses will explore potential mediating effects of anxiety, depression, sleep disturbance, loneliness and bullying.

Ethics and dissemination This study was approved by the National Research Ethics Service, London - Riverside, as well as by the Joint Research and Development Office of the Institute of Psychiatry, Psychology and Neuroscience and South London and Maudsley NHS Foundation Trust (SLaM), and the SLaM Clinical Research Interactive Search (CRIS) Oversight Committee. The findings from this study will be disseminated through peer-reviewed scientific journals, conferences, websites, social media and stakeholder engagement activities.

Trial registration number NCT04601220.

  • Suicide & self-harm
  • Child & adolescent psychiatry
  • MENTAL HEALTH
  • STATISTICS & RESEARCH METHODS
  • QUALITATIVE RESEARCH
https://creativecommons.org/licenses/by/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https://creativecommons.org/licenses/by/4.0/.

Statistics from Altmetric.com

Request Permissions

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

Strengths and limitations of this study

  • The Social media, Smartphone use and Self-harm in Young People (3S-YP) study uses a novel prospective, observational strategy to investigate the patterns of real-world social media and smartphone usage prior to episodes of self-harm in a clinical youth sample.

  • Co-production with youth experts by experience, including coauthor (SB-F), a leading national UK youth mental health charity—YoungMinds, and a steering group comprising of experts in the field of self-harm and suicide prevention will ensure acceptability and relevance, as well as facilitate dissemination and impact.

  • Due to the online nature of the study, participants need to be able to independently engage with the digital tools and study documentation that is only available in the English language.

  • The flexible approach to participation regarding which data young people choose to share, will ensure our methods are acceptable and promote autonomy among participants, though it will increase the risk of missing and inconsistent data which could compromise the validity of the results and interpretations made.

  • Mixed model analyses, state-of-the-art natural language processing methods combined with thematic analysis, will be used to analyse social media metadata, textual and imagery data and smartphone metadata.

Introduction

Self-harm is characterised as any behaviour where an individual causes damage or injury to their body in response to distress. Traditionally, research has focused more on severe cases of self-harm, namely cutting or self-poisoning, rather than a broader spectrum of behaviour, including burning, scratching or hitting oneself. Self-harm in young people is a significant public health concern, with findings indicating that more than 15% of the young people report engaging in self-harm in the community, coupled with high rates of repetition and elevated risk of suicide.1–3 Prevalence figures are likely to underestimate the true scale of the problem, given young people are often reluctant to disclose and seek professional help owing to fear of stigma.4 Furthermore, epidemiological studies tend to rely on hospital presentations and only reflect those cases of self-harm that are brought to medical attention.5 Whereas community and school-based studies collect self-reported retrospective data. Therefore, while capturing a wider range of self-harm behaviours, they lack information about the specific timing and nature of non-help seeking self-harm.

Reasons for self-harm are complex and multifaceted, however there has been an increasing speculation about the potential negative impacts of social media and smartphone use on youth mental health.6 Young people are the most frequent users of social media and smartphones, with 91% of 12–15 year olds using social media and 98% of 13-year olds owning a smartphone in the UK.7 Over the last decade, social media and smartphones have altered how young people spend their time and interact socially with peers.8 This coincides with a sharp increase in the levels of self-harm, particularly among adolescent women.9 Given the widespread use of social media and smartphones, understanding the implications for youth mental health is of paramount importance.

Social media platforms are often used by people with mental health problems to share personal experiences and seek support.10 Feedback received can be perceived as supportive by the individual.11 Thus, social media use can alleviate loneliness and social isolation that is common among young people with mental health problems as it can provide opportunities for strengthening existing friendships and developing new social connections, however, it can also exacerbate loneliness if excessive time is spent online rather than time that would be otherwise spent interacting socially offline.12 13

Online forums have been a cause for particular concern as users may be exposed to harmful self-harm content and online discussions that normalise behaviours and discourage help-seeking, which may be associated with increased suicidal ideation found among forum users compared with other forms of social media user.10 Other studies have found significant associations between cyberbullying victimisation and self-harm in young people, and to a lesser extent, cyberbullying perpetration, likely perpetuated by the anonymous nature of social media.14 Young people engaging in self-harm behaviours may also be more likely to engage in risky online behaviour compared with their peers.15 More recent findings indicate that while for some young people and in certain circumstances, social media can have a positive or neutral effect, for others it can be harmful and so more research is needed to explore the effects of different types of social media use on an individual level.16 17

Research on the effects on mental health of excessive screen time in general is more conflicting, possibly relating to inconsistencies in measurements and definitions employed.18–20 Sohn et al,20 found evidence of an association between problematic smartphone usage and increased anxiety and depression. Emerging evidence indicates that problematic smartphone usage may be positively linked with suicide ideation and the association may be mediated by emotional intelligence.21 22 Furthermore, there is an established literature linking excessive smartphone use and sleep disturbance23 and poor sleep has been linked to depression and suicidal ideation in young people,24 as well as there being an association between sleep problems and suicidal thoughts and behaviours independent of depression.25

Much of the current evidence on the association between social media and smartphone use and mental health and self-harm among young people is limited to associations from cross-sectional studies (eg, Mancinelli et al).26 Where longitudinal studies are reported, they are based on small samples, are reliant on self-report data and have prolonged periods of time between waves of data collection (eg, Arendt et al.)27 These approaches have precluded the ability to explore potential mechanisms underlying self-harm behaviour over time using real-world data, in a prospective way. This study is one of the first to attempt to explore the temporal patterns of social media and smartphone use prior to episodes of self-reported and clinician-reported self-harm events, to identify whether there are usage patterns of social media and smartphone use that are associated with increased risk of self-harm. The findings from this study have the potential for generating a wide-ranging impact as well as filling the gaps in current knowledge. The overarching aim of the Social media, Smartphone use and Self-harm in Young People (3S-YP) study is to investigate the mechanisms by which social media and smartphone use underpin self-harm in a clinical sample of young people. Primary objectives are:

  • To identify characteristic patterns of social media and smartphone use in the 7 days preceding an episode of self-harm.

  • To determine whether characteristic patterns of social media and smartphone use associated with an episode of self-harm, differ from young people who have not had an episode of self-harm.

Secondary objectives are to investigate whether the association is mediated by sleep disturbance, depression, anxiety, loneliness and bullying.

Methods and analysis

Study design

This study is a prospective 6-month cohort study and has been reported according to the Strengthening the Reporting of Observational Studies in Epidemiology checklist for cohort studies (see online supplemental additional file 1) (protocol V.1.4, 21/07/2022).

Supplemental material

Study setting

Young people (aged 13–25 years) will be recruited from South London and Maudsley NHS Foundation Trust (SLaM). SLaM is one of the largest mental healthcare providers in Europe, providing secondary mental healthcare to a local population of approximately 1.3 million residents as well as national and specialist services.

Participant selection

Young people will be identified via one of the following selection methods:

  1. SLaM’s Consent for Contact (C4C) patient research participation register of patients who have given prior consent to researchers with ethically approved studies to access their health records and contact them for the purposes of recruitment to research for which they are eligible.28 A researcher who is independent of the study team and who works for the National Institute for Health Research Biomedical Research Centre within SLaM, will apply the eligibility criteria to the register using the Clinical Research Interactive Search (CRIS) system to screen for potentially eligible young people. The CRIS system is a case register that contains de-identified information extracted from the Trust’s electronic health record (EHR) system.29 The study team will then be supplied with the Trust-identifying details of potentially eligible young people so they can check their EHRs in detail for clinical eligibility. For eligible young people, the study team will inform the care coordinator of the intention to approach prior to initiating any contact. This is so that care coordinators can advise if it is not appropriate for a young person under their care to be approached for the study.

  2. If the above yields insufficient numbers, we will invite clinicians to refer young people under their care directly to the study team. We will attend team meetings to explain the study and eligibility criteria to clinicians. Clinicians will screen young people under their care for eligibility and will contact eligible young people (and a parent or carer with parental responsibility for young people aged 13–15 years old) to explain the study and seek permission to share their contact details with the study team.

Eligibility criteria

Inclusion and exclusion criteria are reported in table 1. No recruitment quotas will be imposed, with respect to any participant characteristics. To reduce the risk of participant selection bias, all eligible young people will be invited to participate until the recruitment target is met.

Table 1

Eligibility criteria

Sample size calculation

In a recent national community survey, the reported annual age-sex incidence of self-harm for young people aged 16–18 years old was 36.4% (95% CI 22.4 to 52.2) for men and 43.5% (95% CI 35.5 to 51.7) for women.3 Based on uptake of SLaM C4C patient research participation register, we estimate that if 600 young people consent to take part, 480 will be followed up during a 6-month period, resulting in at least 150 with an episode of self-harm. Using the most conservative estimates, if approximately 22.5% of those who do not display characteristic problematic patterns of social media usage (or smartphone usage) have an episode of self-harm and this increases to 35% in those identified with problematic usage in the previous week, we expect 86% power to detect a 12.5% increase in self-harm at 5% significance, assuming 480 young people are followed up.

Recruitment method and study procedures

The study team will contact eligible young people (and/or an adult with parental responsibility for young people aged 13–15 years old) by telephone, text or email to invite the young person to take part. They will be sent a unique link to the study website which will host the participant information and online consent and assent forms as well as being offered the option of a face-to-face or virtual meeting if needed. Young people aged 16–25 years old who want to take part, will be asked to confirm their consent via an online form. This will include separate consent for the following types of data collection: questionnaires, EHRs, social media and smartphone data; choosing not to consent to the latter three will not preclude participation. For young people aged 13–15 years old, an adult with parental responsibility will first be asked to confirm their consent via an online form, following which the young person will be asked to confirm their assent. A maximum of three contact attempts will be made to non-responders. The flow of participants through the study will be documented in a flowchart (see figure 1 - Participant flowchart).

Figure 1

Participant flowchart.

Participants will then instal the study software application on their smartphone device and complete the baseline questionnaires. The study software application has been designed to continuously extract smartphone metadata and to deliver the questionnaires for the follow-up period (up to 6 months, reduced from 12 months following an initial pilot phase). Tables 2 and 3 outline data collection measures over the 6-month follow-up period related to primary and secondary objectives and other measurements. Participants who choose not to provide smartphone metadata (or use the study software application for any other reason) will be provided with a unique link to an online survey platform—Qualtrics (https://www.qualtrics.com)—to complete the questionnaires according to the same schedule. Participants will receive a reminder (an automated push notification and/or a prescheduled standardised text message and/or email) on the day a questionnaire is due, followed by a second reminder 7 days later if incomplete. The final set of questionnaires may also be sent via post. Participants who consent to provide their social media data will be contacted at baseline, months 3 and 6 to ask them to upload data. Participants will be given detailed guidance on how to request their data from the platforms and upload it to the study website via a unique link. Participants will receive a £10 voucher as a thank you for enrolling in the study, followed by £5 per month of continued participation and an additional £10 voucher for uploading their social media data at month 6.

Table 2

Outline of data collection measures over the 6-month follow-up period related to primary objectives

Table 3

Outline of data collection measures over the 6-month follow-up period related to secondary objectives and other measurements

Outcomes, exposures and other measurements

Primary outcome

Participant and clinician recorded self-harm events.

Secondary outcomes

Self-reported symptoms of anxiety, depression, sleep disturbance, loneliness and experiences of bullying victimisation.

Exposures

Social media metadata, textual and imagery data and smartphone metadata.

Self-reported social media and smartphone usage.

Other measurements

Self-reported exposure to COVID-19, sociodemographic and clinical covariates and process evaluation data.

Data analysis plan

Descriptive analysis

Individuals will be grouped into those that have had an episode of self-harm, and those that have not. Frequency distributions for different types of self-harm will be summarised, and individuals with multiple, versus a single episode of self-harm will be compared. Baseline sociodemographic and clinical descriptors will be presented for those that self-harm and those that do not. Patterns of social media and smartphone use will be explored using a mixed methods analysis and presented as a descriptive analysis.

Primary analysis

Patterns of social media and smartphone use will be identified and categorised into specific problematic exposure groups. This will be defined using a mixed methods analysis comparing social media and smartphone data in the 7 days leading up to a self-harm episode. Individuals who do not experience an episode of self-harm will be randomly matched with four cases, and the same relative date and time of exposure will be randomly matched.

Social media and smartphone use will be compared between self-harm cases and non-cases, including social media platform of use, duration of use, time of use and absence of engagement (reduced level of posting but passive presence online). For example, night-time use compared with similar use during the day will be compared by self-harm status.

A binary generalised linear model with a log link will be fitted to derive the relative risk of a prospective self-harm episode in relation to a characteristic pattern of social media and smartphone use in the 7 days prior to the episode. Known social and physical determinants of health associated with self-harm will be included to adjust for baseline effects. Missing data will be explored for patterns of missingness. Secondary analyses will explore potential mediating effects of anxiety, depression, sleep disturbance, loneliness and bullying.

The primary analysis population under investigation will include all participants with a post-baseline assessment of self-harm.

Natural language processing and language analysis

In previous work by coauthor Liakata’s team30 aggregates of both language and smartphone features were obtained at hourly intervals preceding an episode (administration of a mood form). These were used in linear and non-linear regression models for prediction of mood scores. Recent work by Liakata’s team has developed natural language processing methods to identify changes in the mood of individuals over time through sequential analysis of 2-week user timelines on social media,31 and has shown that capturing such changes can be predictive of suicidality risk.32 The current protocol focuses on the 7-day window preceding a self-harm episode, making the above-mentioned research directly applicable to this work. While language features used in earlier work includes word embeddings (sentiment specific and others), n-grams, topics and lexica for sentiment, recent developments have shown the success of pretrained language models (eg, BERT from Devlin et al)33 in capturing information on mental health.31 34 The textual content of individuals posts will be represented using pretrained language models, combined with other smartphone data, to predict both moments of change and episodes of self-harm. Our earlier work on fusing heterogeneous features using multiple kernel learning,30 will provide useful baselines. Measurements of affective change both through our models for moments of change31 as well as more traditional measures of valence and affect35 will feed into the predictive models as well as the descriptive analysis. Models will be evaluated under robust cross validation settings to simulate as much as possible a real-world scenario and avoid overfitting.36

Qualitative analysis

Thematic analysis,37 which has been successfully applied in the analysis of social media data of patients and self-harm online communities,38 39 will be employed to explore characteristic patterns of social media use in the 7 days leading up to a self-harm episode from a qualitative perspective, complimentary to the main analyses. Thematic analysis comprises six stages: (1) familiarisation with the data; (2) generation of initial codes; (3) development of themes; (4) review of themes; and (5) defining themes; and (6) writing up the findings.37

Data from a purposive sample of cases will be selected for this analysis, based on age, sex, ethnicity, frequency of social media usage around the time of a self-harm episode and type of self-harm behaviour. Consistent with related work in this field (eg, Williams et al)38 39 around 500–600 social media postings will be analysed. Following removal of any personally identifiable/geographical material and duplicate material (for example, retweets on Twitter), the textual and imagery data of social media posts in the 7 days leading up to and including an episode of self-harm will be analysed using inductive and deductive methods to validate and refine the emerging themes.40 Inductive methods comprise repeated reading of the data to identify salient themes and deductive methods involve testing out predefined themes based on the research question and existing theories, for example, cry of pain theory/integrated motivational-volitional model.41 42 Analysis of visual data will include descriptive summaries of the visual content as well as analysis of thematic content. Coding frames for the analysis of textual and visual data will be developed by the study team, informed by relevant published literature (eg, Shanahan et al)39 the research question and predefined themes. The coding frames will be refined deductively throughout the analytical process.

Process evaluation

Data from participant contact will be collected by the study team to evaluate the study processes for acceptability, feasibility and value. Reasons for decline and withdrawal will be coded and categorised where appropriate and summarised using descriptive statistics. A purposive sample of up to 20 participants (who provided consent to be contacted for further research when they enrolled in the 3S-YP study) will be invited to take part in a brief telephone or in person interview. Participants will complete a separate consent form and interviews will take place following study completion or drop out from the study (including participants who withdraw or stop providing data to the study). Interviews will be audio recorded and transcribed verbatim. Participants will receive a £20 voucher as a thank you for taking part. Interview data will be analysed using Braun and Clark’s37 thematic analysis approach, as outlined above.

Ethics and dissemination

Ethics and safety considerations

This study is being conducted in compliance with the principles of the Declaration of Helsinki43 and Good Clinical Practice as outlined in the UK Policy Framework for Health and Social Care Research.44

This study was approved by the National Research Ethics Service, London – Riverside (ethics ref 20/LO/1187; IRAS ref 269104), as well as by the Joint Research and Development Office of the Institute of Psychiatry, Psychology and Neuroscience and SLaM and the SLaM CRIS Oversight Committee (ref 20–074 and 21–039). The SLaM C4C patient research participation register was approved by the National Research Ethics Service, London – South East (ref 10/H0807/88), as well as by the SLaM Caldicott Guardian and the SLaM Trust Executive. The CRIS system was approved as a data set for secondary data analysis by the National Research Ethics Service, South Central – Oxford C (08/H06060/71). This study is registered on ClinicalTrials.gov.

Given the observational nature of the study, participation in this study is not anticipated to contribute to any specific adverse events. Drawing on our previous research and through careful planning with youth experts by experience, we have developed detailed procedures for conducting the study in a manner that is sensitive to the needs of young people, while maintaining high scientific standards and addressing any arising risks. Young people will provide their informed consent (or assent for young people aged 13–15 years old, along with consent from an adult with parental responsibility) prior to recruitment to the study. Participants can choose what data they want to share. Participants will be informed of the limits of confidentiality and their right to withdraw. Participants will be able to skip questions, take a break or terminate a questionnaire at any time. Participants will be provided with details of resources for accessing support and the study team’s contact information in the event of problems with any aspect of participation. The study team will have received training in the principles of Good Clinical Practice, informed consent, identifying and managing distress and risk (along with receiving regular supervision from the study chief investigator, RD) and the technical aspects of participation.

Data will be handled in accordance with the General Medical Council guidance on confidentiality,45 the principles of Good Clinical Practice, European General Data Protection Regulations46 and UK Data Protection Act.47

Dissemination plan

Data collected in the study will not be made publicly available as it is not possible to guarantee that an individual could not be identified as a participant from their publicly available social media data. The findings will be disseminated through peer-reviewed scientific journals, conferences, websites, social media and stakeholder engagement activities.

Project oversight

Project oversight is provided by our steering group, comprised of researchers and clinicians from across the fields of self-harm and suicide prevention, Child and Adolescent Mental Health Services(CAMHS), data science, public health, sleep medicine, cyberbullying and addictions, as well as young people with lived experience, representatives from the leading national UK youth mental health charity—YoungMinds, and communications and funding advisors. The group will meet at regular intervals throughout the study. The role of the group is to monitor study conduct and progress and contribute to decision-making and solution generation. The group will not have access to participants’ personal data.

Patient and public involvement

Coauthor (SB-F), a senior service user consultant, and representatives from YoungMinds are facilitating an embedded participatory research approach to promote youth engagement and representation. Together we have developed safe working practices to ensure young people with lived experience are supported and able to contribute to the various engagement activities. Research priorities and consideration of key ethical issues were identified at the prefunding stage, through a series of consultations and workshops with youth service users from SLaM. When the study was being set-up, with support from YoungMinds, we partnered with youth advisors who joined the steering group and are contributing to co-production for the study duration. Study procedures and participant-facing digital tools and documentation were codesigned through consultations with the youth advisors and a workshop facilitated by YoungMinds with other young people to ensure the procedures and materials are acceptable and inclusive from a range of perspectives. Towards the end of data collection, we will conduct further engagement activities with YoungMinds to inform the analysis and dissemination plans.

Study status

The study commenced recruitment in June 2021 and aims to conclude data collection in 2023.

Ethics statements

Patient consent for publication

Acknowledgments

We would like to thank all the members of the steering group, including Louise Arsenault, Janis Baird, Johnny Downs, Tamsin Ford, Nuala Flewett, Paul Gringras, Matthew Hotopf, Hazel Inskip, Nicola Kalk, Navneet Kapur, Fiona Lacey, Dennis Ougrin, Louise Pratt, Angus Roberts, Edmund Sonuga-Barke, Robert Stewart, Jack Stonebridge, Alastair Sutcliffe, Sumithra Velupillai and Tony Wood, and the young people who have been involved in co-production activities for their invaluable contributions to the development of the 3S-YP study. We would like to thank YoungMinds for partnering with us. We would also like to thank local NIHR CRN South London staff for support in recruiting participants.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Twitter @bye_amanda, @drbencarter, @rina_dutta

  • Contributors RD, BC, KT, SB-F and ML contributed to the conception of the study and obtaining the funding. RD is the study chief investigator. RD, BC, KT, SB-F, ML, DL and AB contributed to the design of the study. BC performed the power analysis. RD, BC, KT, SB-F, ML, DL, AB and ZZ contributed to the analysis plans. DL, AB, SB-F, RD and GW contributed to the development or testing of the systems and software application used for data collection and management. DL, AB and RD contributed to the development of the data protection and storage systems. AB prepared the protocol and other study documentation for ethical approval and the first draft of the manuscript for publication, and is responsible for the study management, with oversight by RD. All authors contributed to revising the manuscript and approved the final version to be published. All authors have agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

  • Funding This paper represents independent research supported by the Medical Research Council (MRC) and Medical Research Foundation (MRF) (grant number MR/S020365/1). This research was also part supported by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. RD was also funded by a Clinician Scientist Fellowship from the Health Foundation in partnership with the Academy of Medical Sciences and her work is supported by the NIHR Biomedical Research Centre at South London and Maudsley National Health Service (NHS) Foundation Trust and King’s College London. BC is also supported by the Nuffield Trust. ML is also supported by the Engineering and Physical Sciences Research Council (grant number EP/V030302/1) and The Alan Turing Institute (grant number EP/N510129/1). DL is also supported by the Forces in Mind Trust (Project number: FIMT/0323KCL), a funding scheme run by the Forces in Mind Trust using an endowment awarded by the National Lottery Community Fund. The views expressed are those of the author(s) and not necessarily those of the MRC, the MRF, the NHS, the NIHR or the Department of Health and Social Care.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; peer reviewed for ethical and funding approval prior to submission.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.