Secure care (forensic) hospital evaluation of manualised interpersonal art-psychotherapy (SCHEMA): A randomised controlled trial protocol

Primary objective

To assess the effectiveness of interpersonal art psychotherapy in reducing the frequency and severity aggressive behaviour in adult secure care.

Secondary objectives

1. To determine if interpersonal art psychotherapy is cost-effective compared to UC.

2. To explore patient characteristics and psychotherapeutic processes/mechanisms within interpersonal art psychotherapy that are influential to treatment outcomes.

3. To explore the longitudinal changes in aggressive behaviour after receiving art psychotherapy.

4. To evaluate changes in patient distress relating to psychiatric symptoms

Patient and Public Involvement

When and how were the patients/public first involved in the research?

Extensive Patient and Public Involvement (PPI) consultation took place with ‘Research Abilities’ group (Research Abilities, 2023) during the research funding application development process.

How were the research question(s) developed and informed by their priorities, experience, and preferences?

The testing of Art Psychotherapy as an accessible psychological treatment, less reliant upon verbal communication, was viewed as an important research topic by people with lived experience of having a learning disability or borderline intellectual functioning.

How were patients/public involved in:

Individual consultation meetings took place with patients in secure care. As a result, changes were made to interpersonal art psychotherapy worksheets, with greater inclusion of ‘easy read’ and visual materials.

(a) the design and conduct of the study?

PPI has informed the development of this research design, specifically allowing all participants who have been allocated to UC only to access the intervention after leaving the study.

(b) choice of outcome measures?

Measures of personal distress were included as a secondary outcome measure in response to PPI consultation with patients in secure care, this being indicated as a patient priority.

(c) recruitment to the study?

The Plain Language Summary was informed and revised in response to intensive engagement and detailed comments from PPI panel members at the NIHR Research Support Service (RSS). PPI consultation took place to support development of the study Participant Information Sheet (PIS), study consent form, and accessible study recruitment information video.

How were (or will) patients/public be involved in choosing the methods and agreeing plans for dissemination of the study results to participants and linked communities?

Additional PPI activity is planned for the development of accessible and ‘easy read’ study dissemination materials.

Inclusion criteria

Participants are eligible for the trial if they meet the following inclusion criteria and none of the exclusion criteria apply.

•    An inpatient in a secure hospital/unit/service with the presence of learning disability/borderline intellectual functioning indicated by either (a) meeting validated assessment criteria (recognised cognitive testing and adapting functioning assessment), or (b) a score of 57 or below on the Learning Disability Screening Questionnaire (LDSQ) (McKenzie et al., 2012) (a LDSQ copyright license has been obtained for the purposes of the study).

•    Age 18 to 60 years.

•    Able to give informed consent.

•    A current or historic HONOS (Health of the Nation Outcome Scale) (Royal College of Psychiatrists, 2023) score between 1 and 4 for item 1 (i.e. fitting the descriptors- ‘Overactive, aggressive, disruptive, or agitated behaviour’ / ‘Behavioural problems directed at others’).

•    The participant's involvement in the study is supported by their responsible clinician and/or multidisciplinary team (MDT).

Exclusion criteria

•    Unable to give informed consent as assessed by the clinical team.

•    Learning disability/borderline intellectual functioning not indicated based on a validated assessment or screening questionnaire (i.e. not meeting validated assessment criteria or a LDSQ Score >57).

•    A HONOS score of 0 for item 1.

•    Planned discharge within 9 months of the start of the study.

•    Unstable/unmanaged psychotic symptoms requiring active assessment or treatment including medication dose titration (i.e., dose adjustment in the previous 4 weeks or with potential further dose adjustment planned for the following 4 weeks).

Informed consent: Site staff will screen potential participants for inclusion into the trial. All participants who match all inclusion criteria and no exclusion criteria will be approached to discuss the trial and provide with the study information sheet. Participants will be given at least 24 hours to consider the trial and informed consent would be taken by an art therapist involved in the trial. As part of the consent procedure participants will be asked if they are happy for intervention sessions to be audio recorded and for their study data to be stored and shared to support future research projects. Participants could decline both of these options and still participate in the trial. Data from participants who have provided consent for their data to be shared for future use will be uploaded to trial record on Open Science Framework data repository.

Interpersonal art psychotherapy

Interpersonal art psychotherapy is a manualised intervention delivered by a trained Health and Care Professions Council (HCPC) registered art psychotherapist. Interpersonal art psychotherapy approach/principles, specific instructions, detailed techniques and intervention delivery approach are described in a standardised therapist manual allowing replication. The intervention is delivered by an art psychotherapist who has completed interpersonal art psychotherapy manual training, treatment fidelity checks, and is receiving clinical supervision (Hackett et al., 2020). The therapy is scheduled for 12 1-hour individual sessions and 3 optional 1-hour sessions can be added at any time point for personalised support and reasonable adjustments required for a participant with specific communication/learning and/or therapeutic need. Interpersonal art psychotherapy includes seven components arranged as follows; (1) personal goals; (2) coping responses and self-management; (3) relationships; (4) life events; (5) interpersonal themes; (6) imagined future; and (7) final review. Participants have an option to write a letter directly to members of their care team to inform them about the things they are trying to achieve during the intervention and to highlight any personal support needs. Participants completing 5 out of the 7 therapy components listed above will be considered as having completed the intervention.

Comparator

Within inpatient secure care UC involves assessment and treatment by specialist professionals. The MDT (multi-disciplinary team) model uses the Care Programme Approach (CPA) (NHS England, 2018a) to coordinate and plan care. MDTs comprise psychiatrists, clinical and forensic psychologists, mental health and ID nursing staff, and Allied Health Professionals (AHPs). MDTs conduct risk assessment/formulation and management, recovery-focused care and/or positive behaviour support (PBS) (NHS England, 2018a). Patients have access to psychotherapy/psycho-educational work and/or specific offence-related treatment and/or pharmacotherapy treatment. We will identify specific characteristics of UC at the study sites using a standardised pro-forma checklist and will use the TiDieR checklist to describe and present this information (Hoffmann et al., 2014). This will inform the cost-effectiveness analysis and identify any cross-site variation.

Compliance and treatment fidelity

Assessing treatment fidelity is important for multi-site studies to ensure that treatments are operationalised and monitored for differentiation, competency and adherence (Borrelli, 2011). Therapist adherence to the interpersonal art psychotherapy manual within a feasibility study were estimated at 82.25% (Hackett et al., 2020). All sessions will be audio recorded and a sample of 3 timepoints (sessions 2,7,12) for 3 participants across 9 therapists (27% of sessions) will be rated by raters independent to treatment and any other trial procedures, using the interpersonal art psychotherapy therapist checklist, incorporating tested methods for assessing treatment fidelity (Hackett et al., 2020).

Prohibited treatments

Once in the trial, participants are not allowed to take part in any other individual art therapy programmes. Participants who have been randomised to UC only can access the interpersonal art therapy intervention once they have completed the study. This option was included in response to feedback during the Patient and Public Involvement (PPI) consultation process.

Primary outcome measure(s)

The primary outcome is the frequency/severity of incidents of aggressive behaviour as measured by the Modified Overt Aggression Scale (MOAS) (Kay et al., 1988). The MOAS will be completed by healthcare staff at baseline, and then again at 19 weeks and 38 weeks post-randomisation. The primary outcome timepoint is at 19 weeks. The MOAS is an observer-rated measure of frequency and severity of aggression (< 10 or > 10 observations over 7 days), (intra-class correlation coefficient (ICC) of 0.94-0.91) (Kay et al., 1988). Previous research has used the MOAS scale to assess incidents of aggressive and challenging behaviour among adults with intellectual disabilities (Cohen et al., 2010). The MOAS was found to be a reliable measure for assessing the effectiveness of psychological interventions in people with ID and has been explored in both research (Oliver et al., 2007) and clinical settings (Ratey & Gutheil, 1991)

Secondary outcomes measure(s)

The secondary outcome measures are as follows:

1. Cost-effectiveness of art psychotherapy will be based on estimating the incremental quality-adjusted life years (QALYs) and costs between trial-arms. QALYs will be based on the EuroQoL EQ-5D-3L (self-reported based on the adapted EQ-5D-3L for learning difficulties, proxy-reported based on the EQ-5D-3L Proxy Version-1) and Recovering Quality of Life 10-item version (ReQoL-10; both self and proxy-reported) for deriving the ReQoL Utility Index (ReQoL-UI) as collected at baseline, 19-week, and 38-weeks post-randomisation. Costs will be based on applying unit costs to resource-use compiled using a bespoke resource use measure (RUM; e.g., collecting data on GP appointments, A&E attendance, medication) informed by the patient’s clinicals records, as completed at 19 weeks (i.e., resource use from baseline to 19-weeks) and 38 weeks (i.e., resource use from 20-weeks to 38-weeks) post-baseline.

2. Psychotherapy processes will be evaluated through observer-analysis of transcribed audio-recorded therapy sessions using the Working Alliance Inventory-Observer Ratings Scale (WAI-O), Dual-role Relationship Inventory-Revised (DRI-R), the ‘interpersonal art psychotherapy therapist checklist’, as well as linguistic analysis of change in patients’ use of anger-related words and relational words/pronouns, using the Linguistic Inquiry and Word Count (LIWC) software.

3. Longitudinal changes in aggressive behaviour will be assessed weekly between week 19 and week 38 post-randomisation using the MOAS.

4. Patient distress attributed to psychiatric symptoms as measured by the Brief Symptom Inventory (BSI) Positive Symptom Distress Index (PSDI) (Derogatis, 1993).

Participant timeline

A schedule and timeline of procedures for enrolments, intervention, and assessments for participants is shown in Table 1.

Data collection

Information regarding participant treatment and data collection is displayed in Figure 1. Trial data will be collected using both electronic data capture (EDC) through REDCap and paper forms. Participant facing questionnaires (quality of life questionnaires, and BSI) will be completed on paper and then entered into the EDC system. The system will include all the necessary validation to ensure a level of data quality. All data management processes, including quality control checks of paper forms entered onto the database, will be fully detailed in the trial data management plan.

Figure 1. Participant treatment and data collection.

Data management

Missing, unused & spurious data. Missing data will be investigated for cause and extent. If required, the Statistical and Health Economic Analysis Plan (SHEAP) will detail the methods to be used to deal with missing data. Any missing data will be queried with the sites using the resolution workflow in REDCap to ensure that this is missing, and a code will be assigned using the database.

Procedures for reporting deviation(s) from the original SHEAP. These will be submitted as substantial amendments where applicable and recorded in subsequent versions of the protocol and SHEAP.

Inclusion in analysis. The primary analysis for the trial will be performed on an intention-to-treat basis therefore all participants who are randomised will be included in the analysis and so this will be conducted under the treatment policy estimate strategy.

As the per the Good Clinical Practice (GCP), the definition of Source Data is “All information in original records and certified copies of original records of clinical findings, observations or other activities in a clinical trial necessary for the reconstruction and evaluation of the trial. Source data are contained in source documents.” For any data element, there can only be one set of source data , and this will be defined in the site source data agreement.

The source data information for SCHEMA will come from a variety of sources (see Table 2). Data will be collected using an electronic Case Report Form system (CRF) with paper CRF back up. There will also be data collected from participants’ medical notes and patient-reported questionnaires. All delegated staff at the sites will receive appropriate training to complete the CRFs.

Protocol/GCP non-compliance. The PI will report any divergencies from the trial protocol or principles of GCP to the central trial team by email as soon as they become aware of it. The central trial team will review the divergence against the protocol, and local standard operating procedures (SOP) to determine the nature and severity. All divergences will be reported to the Sponsor.

Sponsor details. Cumbria, Northumberland, Tyne and Wear NHS Foundation Trust, St. Nicholas Hospital Jubilee Road Gosforth Newcastle upon Tyne NE3 3XT.Contact email: CNTWSponsorManagement@cntw.nhs.uk

The sponsor did/will not have any involvement in study design, data collection, management, analysis and interpretation of data, or writing of the report.

Main analysis

Outcome/effectiveness analysis. The analysis of the primary outcome MOAS will be performed using analysis of covariance, modelling 19 weeks follow-up MOAS score controlling for baseline MOAS score. Reflecting the sample size calculation analyses will be undertaken with 2-level hierarchical models with participants clustered within therapists. Secondary outcomes will be analysed in a similar way. Multiple imputations will be used in case of missing values in scores. The results will be summarised using point estimates, 2-sided 95% confidence intervals and p-values. Data analysis will be in accordance with a pre-specified statistical analysis plan.

Psychotherapy process analysis. It has been recommended that future research on psychological interventions for institutional aggression should include an assessment of patient characteristics and interpersonal styles that facilitate participation and progress during treatment (Daffern et al., 2008). These process analyses will be conducted on samples of treatment sessions for each patient, to reflect the segments of the treatment and change over time. The reading of the session transcripts together with the process coding of the WAI-O and DRI-I are expected to take about 1.5 hours per session.

Working Alliance Inventory-Observer version (WAI-O). The WAI-O (Darchuk et al., 2000) is a 12-item measure of the working alliance as measured by an observer. This brief WAI-O was developed from the original 36 item Working Alliance Inventory (WAI) (Horvath & Greenberg, 1989). The WAI-O contains three subscales: agreement on tasks, agreement on goals, and development of bonds, each with 4 items rated on 7-point Likert-type scales (7 = high). The client and therapist reported WAI was adapted to fit an observer perspective (the WAI-O) by altering the pronouns (Tichenor & Hill, 1989). The WAI-O will be rated by two independent judges for the sessions. The working alliance score used in analyses will be averaged between the two raters across a sample of sessions for each patient.

Dual Role Relationship Inventory (DRI-R-SF). The Dual-Relationship Inventory-Short Form is a 9-item abbreviated versions of the Dual Role-Relationship Inventory (30 items) that assesses the quality of relationships in mandated treatment settings (Gochyyev & Skeem, 2019). The DRI-SF is a nine-item relationship instrument with items rated on a 7-point Likert-type-scale: 1 = never, 2 = rarely, 3 = occasionally, 4 = sometimes, 5 = often, 6 = very often, and 7 = always. A higher DRI-SF total score indicates a better quality of client-supervising officer relationship. The DRI-R-SF assesses three relationship aspects; caring-fairness, toughness, and trust. The DRI-R-SF has shown high internal consistency and predictive validity for probation violations and revocation, and moderate inter-item correlations.

Linguistic Analysis of Word Count (LIWC-2022). We will apply the text analysis software LIWC-2022 to the transcripts of the sessions (LIWC-22, 2022). We specifically will analyse the change in frequency of patients’ anger-related word use over the course of treatment, and the change in affiliation/pronoun use. As with the other two process measures, we will use the transcripts from a sample of sessions per treatment as a proxy of the treatment process as a whole.

Sub-group & interim analysis

A sub-group analysis of differences between therapy responders and non-responders will be completed. Full details of the analysis will be included in the primary trial SHEAP.

Qualitative analysis

The qualitative analysis will include a minimum of 20 audio-recorded and transcribed interviews integrated into the RCT (Davis et al., 2019). A five-step framework analysis will be completed, including data familiarisation, identifying a framework, indexing, charting, and mapping and interpretation (Ritchie & Spencer, 1994) from a minimum sample of n=20 qualitative interviews. Based upon guidance on the development of complex interventions to improve health, qualitative interviews and analysis will contribute to a process evaluation of study retention, experiences of the intervention, fidelity, dose, and reach (O'Cathain et al., 2019). We will invite a minimum of to 10 participants (recruited from both arms) and 10 study therapists to engage in semi-structured qualitative interviews. A minimum of 10 interview participants can be considered sufficient where the aim of the research is to reach an in-depth understanding of participant experience, rather than to seek generalizability of findings. Data saturation can often be achieved with a sample size of 6 to 12 participants (Guest et al., 2006), and furthermore, evidence suggesting that insightful and meaningful analysis can be achieved from relatively small sample sizes within health research where samples have been purposively selected (Pons-Vigués et al., 2019). There will be no cut-off for sample size of participants included for interview, all study participants and therapists who express an interest in taking part in an interview will be given the opportunity to participate. Full details of analysis will be included in the qualitative analysis plan.

Interview schedules for study therapists will focus on their experiences of intervention delivery, their experiences of training and supervision, seeking examples from therapists about their ‘in-therapy’ responses from the participants, and implications for clinical practice implementation. Participant interview schedules will focus on the participants experience of taking part in research, their experiences and reflections on art making, therapy and their relationship with the therapist within sessions.

Cost-effectiveness analysis

A within-trial cost-effectiveness analysis will be conducted using the same statistical methods as the primary outcome (i.e., 2-level hierarchical model) (Glick et al., 2014). Costs will focus on the incremental costs/resource-use of providing art psychotherapy relative to UC and downstream resource-use collected via a bespoke resource-use measure (RUM) proforma based on patients’ care records (e.g. wider staff engagement and medications). Completion of the RUM involves research staff checking the patient notes for what resources the patient has used based on the proforma, which they then confirm with ward staff. Costs will be estimated based on attaching unit costs to resource-use data, with unit costs sourced from commonly used UK sources (BMJ, 2024; Jones et al., 2024; NHS England, 2023). Incremental effectiveness will be based on two preference-based measures: EQ-5D-3L (EuroQol, 2024), and ReQoL-UI derived from the ReQoL-10 (Keetharuth et al., 2024). These measures will include both a self-reported (i.e., by the patient) and proxy-completed (i.e., by ward staff) version. For proxy-completion, we have specified that at each time point both the EQ-5D-3L and ReQoL-10 should be completed by the same person (i.e., same ward staff member) and where possible it should be the same person across time points, but we recognise that might not always be possible. We will record who has completed the proxy-response. For the EQ-5D-3L, the self-reported version is based on the adapted EQ-5D-3L for adults with mild to moderate learning disabilities (O’Dwyer et al., 2024); whereas the proxy-version is the EQ-5D-3L Proxy Version-1 (EuroQol, 2024). There are no adapted or proxy versions of the ReQoL-10; therefore, the self-reported and proxy-reported versions are both the traditional ReQoL-10. The EQ-5D-3L and ReQoL-UI both have a utility index (UI) which can be used to elicit quality-adjusted life years (QALYs). The UI for the EQ-5D-3L and ReQoL-UI will be based on the relevant UK value set, respectively (Dolan, 1997; Keetharuth et al., 2021).

These measures have underlying conceptual and methodological considerations relevant to the patient population and intended outcomes from art psychotherapy improvements. As such, the choice of preference-based measure as the ‘primary’ outcome for the economic evaluation will be based on a within-study psychometric analysis. We are unaware of any previous psychometric results that aid with understanding the appropriateness of the EQ-5D-3L and ReQoL measures in our trial’s patient population of interest; thus, our proposed psychometric analyses can aid such measure justification in our trial and future trials in a similar patient population where similar outcomes are of interest. This post-hoc psychometric analysis will be conducted based on construct validity (i.e., correlation and effect sizes relative to clinical outcomes) and responsiveness (i.e., standardised response means, and floor and ceiling effects) relative to the other primary and secondary outcome measures, as has been done to judge the relative appropriate use of the EQ-5D (-5L) and ReQoL-UI for cost-effectiveness analyses previously (Franklin et al., 2021; Franklin et al., 2022). The cost-effectiveness results using all measures will be reported and discussed; the psychometric analyses will be conducted and the choice of primary measure will be detailed in our SHEAP before conducting the cost-effectiveness analyses.

Point-estimate cost-effectiveness will be presented using incremental cost-effectiveness ratios. Bootstrapping will be used to report bootstrapped standard errors, and to estimate the probability of cost-effectiveness relative to a range of cost-effectiveness thresholds (including NICE’s £20,000 to £30,000 per QALY threshold) to be presented using cost-effectiveness planes and cost-effectiveness acceptability curves (Glick et al., 2014). Sensitivity analyses will be used to assess point estimate uncertainty.

Trial management

TMG (Trial Management Group) and Advisory Group (AG). The TMG will have regular meetings (every 4–6 weeks) prior to and during the study (Y1–4) chaired by the Chief Investigator (CI) with CTR support. TMG members will sign up to the remit and conditions set out in the TMG Charter.

TSC (Trial Steering Committee). This is an low-risk trial therefore a Independent Data Monitoring Committee (IDMC) will not be convened, unless recommended by the Trial Steering Committee (TSC). TSC membership will comprise of members who are independent of the trial taking up positions as chair, statistician, expert therapist, and a public member having signed-up to the remit and conditions as set out in the TSC Charter.

Quality control and assurance

Monitoring. Trial specific monitoring activity is determined by the risks outlined within the clinical trial risk assessment and the overall determination of risk. All activities will be fully documented within the trial monitoring plan and deviations appropriately recorded by the trial and sponsorship team.

Harms

Safety reporting. It is expected that all Serious Adverse Events (SAEs) be reported no later than 24 hours after knowledge of the event by the site PI or delegated site staff, unless an exemption has been specified within the protocol. Safety reporting terms for the trial are shown in Table 3 and guidance for determining intervention causality is shown in Table 4.

Trial Specific SAE Reporting requirements. There is an additional event that is being classed as an SAE for the purposes of this trial and therefore must follow the SAE reporting guidelines:

Auditing

The trial is participant to inspection by Research Ethics Committee (REC) as the regulatory body. The trial may also be participant to inspection and audit by Cumbria, Northumberland, Tyne, and Wear NHS Foundation Trust, under their remit as Sponsor.

Ethics and dissemination

This protocol and related trial documents were reviewed by the London-City & East REC and received full approval on 13/01/2023 (REC ID: 23/LO/0026; IRAS project ID: 319325). All participants will be approached to provide written consent prior to being entered into the trial.

Confidentiality

The CTR will act to preserve participant confidentiality and will not disclose or reproduce any information by which participants could be identified, except where specific consent is obtained. Data will be stored securely and will be registered in accordance with the General Data Protection Regulation 2016 and Data Protection Act 2018. The data custodian is Cumbria, Northumberland, Tyne and Wear NHS Foundation Trust. Select trial data will be available on the Open Science Framework data repository.

Dissemination policy

Publication policy. All publications and presentations relating to the trial will be detailed in the publication policy which will be drafted and authorised by the TMG. It will state principles for publication, describe a process for developing output, contain a map of intended outputs and specify a timeline for delivery. The publication policy will respect the rights of all contributors to be adequately represented in outputs (e.g. authorship and acknowledgements) and for the trial to be appropriately acknowledged. Authorship of parallel studies initiated outside of the TMG will be according to the individuals involved in the project but must acknowledge the contribution of the TMG and the CTR. We will include PPI consultees in the development of accessible ‘easy read’ summaries of results and accessible materials.