Trial design and flow chart

The design is a multicentre, parallel group randomised controlled trial with single blind assessment to test whether the automated VR cognitive treatment added to treatment as usual, compared with treatment as usual alone, leads to a post-treatment reduction in real world distress and avoidance for patients with psychosis attending NHS mental health services. Treatment as usual will be measured but remain unchanged in both groups. Assessments will be carried out at 0 (baseline), 6 (post-treatment) and 26 (follow-up) weeks by a researcher blind to treatment allocation. A summary of the trial design can be seen in figure 1. The trial is prospectively registered with the ISRCTN registry: ISRCTN17308399. There is a Data Monitoring and Ethics Committee.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Trial flow diagram. TAU, treatment as usual; VR, virtual reality.

Randomisation, blinding and code-breaking

Participants will be randomised once they have completed the baseline assessment. Participants will be allocated to one of the trial arms using a 1:1 allocation ratio. Randomisation will be carried out by a validated online system, Sortition, designed by the University of Oxford Primary Care Clinical Trials Unit. Randomisation using a permuted blocks algorithm, with randomly varying block size, will be stratified by site (Bristol/Manchester/Newcastle/Nottingham/Oxford) and service type (in-patient/early intervention/community mental health team).

The research assessors will be blinded to group allocation, but the patients and staff member present will not be (they cannot be blinded to whether psychological intervention is delivered or received). If an allocation is revealed between assessment sessions, this is logged by the trial coordinator and re-blinding will occur using another assessor.

Participants

The trial participants will be patients with psychosis and self-reported difficulties going outside among other people due to anxiety. The principal method of recruitment will be via seeking referrals to the trial from the relevant clinical teams (adult community mental health teams; early intervention services, and inpatient units) in the participating mental health trusts. The trial centres will be Bristol, Manchester, Newcastle, Nottingham, Oxford, with recruitment from local NHS mental health trusts. With the approval of the clinical team, patients interested in taking part will then be approached by the research team, given information about the trial, and screening conducted. Our Lived Experience Advisory Panel (LEAP) have also emphasised the importance of patients of the participating trusts self-initiating referral to the trials, in order to minimise the chances that particular patients are overlooked by clinical teams or the clinician was not present at a referral meeting. Hence we will also advertise the study and patients within participating trusts will be able to self-refer for a screening to take part in the study. However, in all instances we will also seek to confirm that a member of the clinical team gives approval for a patient to enter the trial. Informed consent will be obtained from all patients before participation.

Inclusion criteria

• Adults aged 16 years or older;

• Attending an NHS mental health trust for the treatment of psychosis;

• Clinical diagnosis of schizophrenia spectrum psychosis (F20 to F29) or an affective diagnosis with psychotic symptoms (F31.2, 31.5, 32.3, 33.3) (International Statistical Classification of Diseases and Related Health Problems: Tenth Revision);18

• Having self-reported difficulties going outside their home primarily due to anxiety that they would like treated;

• And participant is willing and able to give informed consent for participation in the trial.

Exclusion criteria

• Unable to attempt an Oxford-Behavioural Assessment Task (O-BAT) (the primary outcome measure) at baseline (eg, due to being unpermitted to leave a psychiatric ward);

• Photosensitive epilepsy;

• Significant visual, auditory or balance impairment;

• Current receipt of another intensive psychological therapy (or about to start it within the 6 week trial therapy window);

• Insufficient comprehension of English;

• In forensic settings or psychiatric intensive care unit;

• Organic syndrome;

• Primary diagnosis of alcohol or substance disorder or personality disorder;

• Significant learning disability;

• Or current active suicidal plans.

Assessments

Basic demographic and clinical data will be collected (eg, age, gender, ethnicity, clinical diagnosis, medication use). The primary outcome, avoidance and distress of everyday situations as measured by the O-BAT (adapted from 14), will be measured at baseline, 6 weeks and 26 weeks. The O-BAT comprises a personalised hierarchy of five real world tasks that the patient finds difficult due to anxiety. The person then tries to carry out the hierarchy, rating anxiety at each step achieved, and stopping when they decide the anxiety is too great. This therefore produces an avoidance score (0 to 5, with higher scores indicating lower avoidance) and a distress score (0 to 10, with higher scores indicating greater distress) for each level. All assessors receive training in administering the O-BAT and the manual. The initial O-BAT for each participant is reviewed by a clinician. A detailed assessment of social avoidance is first carried out using both a semi-structured interview and a self-report measure of social avoidance, the self-report O-BAT.19 This identifies the everyday situations and tasks that are anxiety-provoking for the participant, and provides a predicted level of distress for each. Based on this, the five-step personalised hierarchy is developed. A hierarchy can be constructed within one or a number of feared situations (eg, standing on the front door step for 3 min, standing outside the front gate for 3 min, walking down the local street, walking to the local shop, buying something in the shop). The hierarchy is set up so that it is likely that the patient may only complete a small number of steps at baseline. Secondary outcomes will also be assessed at all three time-points. Anxious avoidance (AMI-A20 and self-report version of the O-BAT19), suicidal ideation (Columbia Suicide Severity Rating Scale),21 overall paranoia (R-GPTS),22 23 paranoia worries (Paranoia Worries Questionnaire)24 and levels of depression (PHQ-9)25 will be assessed. Activity levels will be assessed using actigraphy (over 7 days), complemented with a time-budget assessing meaningful activity.26 The EQ-5D-5L27 and ReQol28 will assess quality of life. Additionally, quality of life will be assessed using the Questionnaire about the Process of Recovery (QPR).29 For mediation, we will assess, at all time-points, threat cognitions and use of defence behaviours (CDBQ)30 and strength of safety, vulnerability and threat anticipation beliefs.31 Moderators will be assessed at baseline only by a brief assessment of negative hallucinations when outside,32 the Beck Hopelessness Scale,33 the Body-Esteem Scale for Adolescents and Adults34 and the Cognition and Defence Behaviours Questionnaire.30 We will record service use, and other relevant health economic data, using the Client Service Receipt Inventory.35 A summary of the measures is provided in table 1.

The VR psychological treatment

The gameChange VR treatment is a virtual-reality application recommended for adults (16+) who have anxieties when outside in everyday social situations. This software is intended to reduce anxieties around other people and therefore to help participants feel safer and more comfortable around people. The aim for the outcome is that patients feel more able to go outside into everyday situations. The treatment was programmed by the University of Oxford spin-out company Oxford VR (www.oxfordvr.org). The treatment is a CE marked Class I Active Medical Device- Z301 (Standalone Software), in conformity with the essential requirements and provisions of the EC Directive 93/42/EEC (Medical Devices).

A mental health professional, most likely a peer support worker or psychology assistant, will be in the room when the treatment is given. This person will help the patient put on the VR headset and start the programme. The staff member will also encourage the person to apply the learning from VR into the real world through the setting up of homework tasks to be carried out between sessions. The applications will run through the Steam software application on a laptop computer connected to a head-mounted display and accessories. All hardware is already commercially available and has not been modified for the trial. Satisfaction will be assessed after completion of the last treatment session using a modified version of the Client Satisfaction Questionnaire.36

The VR Cognitive Therapy (VRCT/gameChange treatment) aims for patients to test their fear expectations around other people in order to relearn safety. The treatment is not designed as exposure therapy (participants are not asked to remain in situations until anxiety reduces) but as repeated behavioural experiment tests (to learn that they are safer than they had thought). The treatment is designed to be delivered in approximately six sessions of 30 minutes. Three sessions will be considered the minimum (adherent) dose of therapy. However participants can proceed at their own pace, meaning that a fewer or greater number of sessions is allowed. The participant typically stands, and is able to walk a few paces in the scenarios. A virtual coach guides the person through the treatment, including encouraging the dropping of defence behaviours, and elicits feedback to tailor the progression of the treatment. When first entering VR, the patient goes into the coach’s virtual office and is guided in how to use VR (ie, the basic functions). At the beginning of the first session, the virtual coach explains the rationale behind the treatment, and the participant selects which one of six virtual reality situations that they would like to begin in. The six virtual reality scenarios are a: café, general practitioner waiting room, pub, bus, street scene and shop. Each scenario has five levels of difficulty (eg, the number and proximity of people in the social situation increases) and participants work their way through each level of difficulty. There are (therapeutic) game type tasks within a number of the levels (that are designed to help the person drop defence behaviours and make new learning). The participant can choose a different scenario in each session or repeat a previous situation (and level). Throughout the sessions, participants’ responses to questions from the virtual coach are given by means of gripping a virtual globe. Belief ratings are repeated within VR at the beginning and end of each treatment session. Figure 2 provides a summary of the treatment design. A video about the gameChange treatment can be viewed here: https://www.youtube.com/watch?v=D31wodNAMZA.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

The structure of the virtual reality treatment.

Control condition

Participants who are allocated to the control arm will continue to receive their usual care. No additional interventions will be offered by the research team. Treatment as usual for the participants within this trial will typically consist of long-term prescription of psychiatric medications, and meetings with a mental health practitioner. Treatment as usual will vary across individuals and mental health trusts. We will collect detailed data on treatment as usual (which will also inform the health economic evaluation).

Adverse events

A trial standard operational procedure has been written for adverse events. We will record the occurrence of any serious adverse events reported to us and also check each patient’s medical notes at the end of their participation in the trial. An adverse event is defined by the ISO14155:2011 guidelines for medical device trials as serious if it: (a) results in death or, (b) is a life-threatening illness or injury or, (c) requires hospitalisation or prolongation of existing hospitalisation or, (d) results in persistent or significant disability or incapacity or, (e) medical or surgical intervention is required to prevent any of the above, (f) leads to foetal distress, foetal death or consists of a congenital anomaly or birth defect or (g) is otherwise considered medically significant by the investigator.

Life threatening in the definition of a serious adverse event refers to an event in which the subject was at risk of death at the time of the event; it does not refer to an event that hypothetically might have caused death if it were more severe. A planned hospitalisation for a pre-existing condition, without a serious deterioration in health, is not considered to be a serious adverse event. The sorts of serious adverse events that can typically happen to this participant group include: deaths, suicide attempts, serious violent incidents and admissions to hospital.

We will also record any adverse device effects from the VR treatment, which includes adverse events resulting from insufficient or inadequate instructions for use, deployment, installation or operation, or any malfunction of the software. It also includes any event resulting from user error or intentional misuse.

Analysis

A full statistical analysis plan will be drafted prior to recruitment beginning and approved before any analysis. We will report data in line with the Consolidated Standards of Reporting Trials 2010 Statement37 showing attrition rates and loss to follow-up. The primary analyses will be carried out using the intention-to-treat principle. That is, after randomisation, participants will be analysed according to their allocated intervention arm irrespective of what intervention they actually receive, and with data available from all participants included in the analysis including those who do not complete therapy. The outcome analyses will be conducted by statisticians in the University of Oxford Primary Care Clinical Trials Unit.

We will test the primary hypothesis for between-group difference in the primary outcome (O-BAT at 6 weeks) using a linear mixed effects model which models the response at 6 weeks and 26 weeks, with baseline outcome measure, stratification variables and treatment assignment as fixed effects, with a patient specific random intercept. An interaction between time and randomised group will be fitted as a fixed effect to allow estimation of treatment effect at all time points. The linear mixed effects model will account for missing data assuming data are missing-at-random. Standard residual diagnostics will be assessed for the appropriateness of the model. P value <0.05 will be used as the level of statistical significance. Similar mixed effect models will be used to analyse secondary outcomes. We will recruit around 432 participants into this trial, with 216 in each arm. This sample size takes into consideration a maximum attrition rate of 20%, and provide 90% power to detect a difference of around 8 (SD=23) in O-BAT anxiety score (using the 0 to 100 scaling from Freeman et al, 201614), from randomisation to 6 weeks (ie, standardised effect size of 0.35) at 5% level of significance (two-sided).

The mediation analysis will investigate putative mediational factors using modern causal inference methods. This involves using parametric regression models to test for mediation of VRCT on outcome through the putative mediators. Analyses will adjust for baseline measures of the mediator, outcomes and possible measured confounders. We will include repeated measurement of mediators and outcomes to account for classical measurement error and baseline confounding. The identified moderator variables (negative auditory hallucinations, hopelessness, appearance concerns and social phobia) will be considered for moderation of the intervention effect on the primary outcome.

A microcosting approach will be used to inform the cost per patient of the VR treatment. The within-trial health economic analysis will describe and compare the costs and outcomes of the two trial arms. Incremental cost per activity gained (primary outcome) will be estimated and the costs and remaining outcomes (utilities, psychiatric symptoms and well-being) assessed separately. This will be informed by a health economics statistical plan written prior to the economic analysis. The health economics will use an NHS and social care services perspective with resource utilisation valued using national cost data sets and EQ-5D-5L data converted into utilities using the UK tariffs. A broader perspective including lost earnings, patient out-of-pocket costs and criminal justice costs will also be considered. A state-transition model will be developed to extrapolate the within-trial analysis and estimate the incremental costs per quality-adjusted life years (QALYs) gained from using the VR treatment, supported by the trial data, literature reviews and discussions with clinical experts. Uncertainty around the incremental cost-effectiveness ratio will be reported using the cost-effectiveness plane and the cost-effectiveness acceptability curve. The maximum reimbursable price of the VR treatment conditional on the willingness to pay per QALY will be determined. We will then estimate the affordability to the NHS of a decision to implement the VR treatment. This will take the form of budget impact analysis using a time horizon of 3 years to be consistent with National Institute for Health and Care Excellence (NICE), informed by the results of the trial health economics analysis.

Patient and public involvement

The project has had extensive patient and public involvement (PPI). Principally this has occurred via The McPin Foundation, a charity that exists to ‘transform mental health research by putting the lived experience of people affected by mental health problems at the heart of research methods and the research agenda’. A grant-holder is from The McPin Foundation. Three other people with lived experience commented on the grant application and a focus group of people with lived experience was convened so that they could try VR and comment on the application.

Following the award of the grant there has been considerable PPI. A LEAP has been formed to advise and shape the development of the treatment, the trial protocol and implementation into services. The LEAP comprises 10 individuals with lived experience of psychosis drawn from each of the study sites (Bristol, Manchester, Newcastle, Nottingham, Oxford). For the protocol they have advised on: the choice of outcome measures, recruitment methods, the format of recruitment materials and the content and wording of study materials. The LEAP have also reviewed and commented on the trial protocol document. In addition to the LEAP, we have also worked with people with lived experience from each of the trial sites to develop the VR treatment. A number of workshops were held. Through these workshops, people have contributed to the selection of the VR scenarios, the therapeutic tasks within the scenarios and style of VR coach. These workshops entailed people with lived experience sharing their ideas, reviewing design concepts, and testing these out within VR. In addition to these workshops, there has been weekly input from a smaller group of individuals with lived experience to gain prompt feedback on details of design. There has been detailed user testing of the VR treatment during software development.

PPI will continue throughout the trial. First, LEAP meetings will occur over the course of the trial. The LEAP will advise on any difficulties that occur in the trial. The LEAP will also contribute to the dissemination strategy. Second, there will be a qualitative evaluation of the VR treatment, with the interviews carried out by researchers with lived experience. This work will be run by The McPin Foundation. Third, a McPin staff member sits on fortnightly gameChange review meetings and on the Research Steering Committee comprised of senior team members.

Ethics and dissemination

The trial has received Health Research Authority and Health and Care Research Wales approval (IRAS 256895, The gameChange trial). The trial received ethical approval from the NHS South Central - Oxford B Research Ethics Committee (19/SC/0075). The results of the trial will be published in a peer-reviewed journal and made open access. An anonymised version of the main outcome data will be available from the trial team on reasonable request after publication of the main results paper.

Trial status

The trial is due to start patient recruitment in July 2019. Recruitment will be for a year until July 2020, with final outcome data collected by January 2021. A trial paper with the outcome results should be submitted for publication around April 2021.