Development and modelling study of an evidence-based manualised intervention: PErsonalised Treatment packages for Adults with Learning disabilities who display aggressive challenging behaviour in community settings (PETAL)

Patient and Public Involvement

There has been Patient and Public Involvement from people with lived experience, including family carers and self-advocates with an intellectual disability, who were recruited through the Challenging Behaviour Foundation and Camden Disability Action respectively. Both groups had experience of participating in research and service development, and have been involved at every stage of the study, from the development of the application throughout the intervention development and the clinical trial. Full details of our PPIE work are provided in a separate document that is currently under review. Regarding the co-production work with our PPIE groups, individual contributions in the development of the intervention are detailed in the manuscript.

The development of the intervention was informed by the framework for developing complex interventions by the Medical Research Council and National Institute for Health Research^19,20 and guidance by O’Cathain et al. (2019)²¹. A dynamic iterative process was used to inform and refine each phase of the project (Figure 1).

Figure 1. Stages of intervention development.

Assembling an intervention development team

An intervention development team was assembled to include experts in the field of intellectual disabilities with a range of specialisms and experience in psychology, psychiatry, intervention development, implementation, and behaviours that challenge. Five members had contributed to NICE guideline development groups for guidelines 11, 54 and 93 relating to the management and services for behaviours that challenge and mental ill-health. Two lived experience advisory groups (5 self-advocates with an intellectual disability and 8 family carers) recruited through Camden Disability Action and the Challenging Behaviour Foundation respectively, were also involved at every stage of the development of the PETAL intervention and attended co-production meetings. We refer to co-production in this context as a way of working with multiple stakeholders, including academics with extensive experience in development of complex interventions, professionals, carers with lived experience and self-advocates with an intellectual disability on an equal basis to design and plan the intervention. The process of co-production for this study followed the guidelines outlined by the Challenging Behaviour Foundation (see https://www.challengingbehaviour.org.uk/what-we-do/sharing-best-practice/co-production/2024).

Overview of the PETAL intervention

The prototype PETAL therapy is a manualised 7 module intervention to be delivered over a maximum of 14 weeks, including up to 2 review sessions (see Table 1). It utilises a multicomponent approach with the aim of reducing and/or preventing aggressive challenging behaviour, incorporating adapted elements from existing evidence-based interventions. There is an accompanying therapist manual which includes session outlines and scripted session plans. Workbooks that map onto the manual are provided to carers and people with an intellectual disability. Workbooks include module content and home practice tasks to support the embedding of skills. Home practice tasks are set as SMART (Specific, Measurable, Achievable, Relevant, and Time-Bound) objectives. The therapy content, setting of delivery, frequency, and duration of sessions (maximum of two hours) can be personalised to the individual, e.g., goal/objective setting, having the person present or not in a session, extending a module to more than one session. Participants are encouraged to set SMART objectives, record these in their workbooks, practice skills between sessions and incorporate realistic and manageable changes into their daily routines.

Intervention delivery and therapist training/supervision

With input from both advisory groups, we developed and ran a two-day online training programme, with an additional 6 hours of preparatory work, to train therapists in essential clinical skills (e.g., risk management, communication, building a therapeutic alliance, managing conflict, delivering behaviour change techniques, etc.) and manual content. Training utilised an interactive approach, including content talks, case scenario discussions, role plays, recordings and input from family carer advisors. Videos were also filmed with both advisory groups (including with self-advocates with an intellectual disability) that were shown during the training focusing on their views of aggressive challenging behaviour and approaches to management. The completion of a case study assessment with three questions at the end of the second training day was required for therapists to receive permission to deliver the intervention. All trained therapists received individualised feedback and a certificate, and they all completed an evaluation form. A webpage was designed to provide therapists with key resources to support therapy delivery. This included downloadable versions of the manual and workbooks, printable module specific resources and guides for the collection of fidelity checklists. Therapists received supervision fortnightly from a senior member of staff within their clinical team to discuss specific participant clinical issues and risk, as well as group supervision every 6 weeks with the PETAL research team to explore challenges with therapy delivery. Separate supervisor meetings were held with the research team every 6 weeks. The research team were also available for 1-1 meetings in-between supervision sessions.

Our advisory groups were also mindful to ensure that the training made it clear not to ascribe blame on the person under review as the cause of the problem. The terminology used has been discussed at length with the members of the advisory groups who also participated in the training programme, providing their experience and viewpoint on the management of aggressive challenging behaviour.

Ethical approval

Ethical approval for the modelling study was obtained from Wales Research Ethics Committee 7 on 31^st October 2022 (REC reference: 22/WA/0267). The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975 (https://www.wma.net/policies-post/wma-declaration-of-helsinki/), as revised in 2013. All procedures involving human subjects/patients were approved by the Wales Research Ethics Committee 7.

Sampling and participants

Participants were recruited from Community Intellectual Disability Services from 6 NHS trusts in England. We included participants with an intellectual disability with their family or paid carers aged 18 or over living in the community. Participants were required to display incidents of physical aggressive challenging behaviour to people or property for at least 3 months and for it to be a current problem. People with an intellectual disability without capacity to consent required a family or nominated consultee to agree to their participation by signing a declaration form. Carers also needed to speak and understand English. We excluded participants with an intellectual disability if they were currently an inpatient, were alcohol or drug dependent or were currently enrolled in another clinical trial.

Measures

Baseline measures were collected from participants to test the acceptability of the research assessments (see Extended Data Table 4 for list of measures). All measures were collected with the carer, except for the EuroQoL Five Dimensions Learning Disability Scale (EQ-5D-LD, license obtained) which was collected with people with an intellectual disability who had the capacity to consent.

Fidelity

We developed two fidelity checklists adapted from those used for other psychological interventions^27–33 (Royston et al., 2024). The fidelity checklist for therapists comprises of an attendance log and sections for each module to outline whether aspects of the module were delivered (e.g. discussions of the session plan, introducing the session, personalisation and module specific items; response options: yes, partially completed, no) and sections to record adaptations, reasonable adjustments, challenges and further comments. A fidelity checklist for external coders based on observed recordings of therapy sessions was also developed. This checklist included general items to assess therapist competencies (e.g. communicating warmth, concern and caring, pacing the session appropriately, etc) and module specific items corresponding to the delivery of each module (response options: yes, partially, no). Out of the 80 sessions delivered, recordings were available for 44 sessions. The remaining recordings were missing due to technological problems (e.g. corrupt recordings, parts of recording missing or no recordings taken). Inter-rater reliability was calculated for 8 (18%) of audio recorded sessions on this checklist across five independent raters in the research team and reliability between coders was high (general items: α = 0.8385 and module-specific items: α = 0.8230).

Qualitative interviews

One-to-one semi-structured interviews were conducted to explore stakeholder experiences of participating in the PETAL modelling study from the perspective of i) carers (paid/unpaid); ii) therapists, iii) supervisors; iv) service managers of community intellectual disability teams and v) participants with an intellectual disability. The Normalisation Process Theory (May, 2007) provided the lens through which data was deductively interpreted and analysed. Transcripts were imported into NVivo (QSR International Pty Ltd. Version 12.1.90) to organise and assist with data analysis.

Descriptive statistics

To test the acceptability of the intervention and how it is delivered in practice including fidelity, the PETAL intervention was modelled in a pre-pilot study and recruited dyads or triads of ten people with intellectual disability and one or two family or paid carers (total n=21). The carers supported the intervention and worked with the person with intellectual disability in the sessions and with homework. Baseline assessments (see Table 2–Table 4) were completed for 9 participants and 8 participants received PETAL therapy (1 person withdrew before the baseline assessment and 1 person withdrew after the baseline assessment due to other commitments). All carer administered assessments tended to take between 2.5–3 hours to complete.

Information on the Client Service Receipt Inventory (CSRI) about service use was available for 8 individuals. Participants had contact with the following community-based services over the preceding four months: general practitioners (n=8), care manager/social workers (n=6), community learning disability psychiatrists (n=5), dentists (n=3), occupational therapists (n=2), practice nurses (n=2), learning disability psychologists (n=2), learning disability nurses (n=2), speech and language therapists (n=1), podiatrist/chiropodist (n=1). Participants were taking medication including sedatives (n=4), antidepressants (n=3), physical health medication (e.g. reflux medication, n=3), contraceptives (n=2), sleeping tablets (n=2), antipsychotics (n=2), anticonvulsants (n=2) and mood stabilisers (n=1).

On the MOSS-PAS (ID), 6/8 individuals scored above the threshold for an anxiety disorder, 2/7 scored for depression and 1/7 scored above the threshold for obsessive compulsive disorder. Two assessments were only partially completed or not completed as carers felt it was taking longer than anticipated. None of the participants scored above the threshold for autism, mania, psychosis or possible organic conditions. However, 6/8 carers reported the person they cared for had a clinical diagnosis of autism.

Six participants did not have access to alternate therapies for aggressive challenging behaviour, two people were receiving or were scheduled to receive Positive Behaviour Support. One person withdrew from the study before completing this measure.

Therapy

Two sets of training sessions were held in February 2023 (n=14 therapists) and April 2023 (n=3). Therapists came from a range of healthcare professions (research assistants: n=3, assistant psychologists: n=2, community learning disability nurses: n=2, psychiatry trainees: n=3, behavioural advisor: n=1, clinical psychologist: n=1, mental health learning disability nurse: n=2, occupational therapist: n=1, senior practitioner: n=1, specialist registrar: n=1). Therapists had on average 8 years of experience working with people with intellectual disabilities (SD: 8.25, range: 1–24). The feedback for the training was positive, with 80% of therapists reported the training programme was useful. Therapists particularly valued the input by carers with lived experience in the training and found case scenarios and group discussions the most helpful.

Therapists were asked to complete fidelity checklists for each participant. Modules 1, 5, 6, 7 and the review sessions each tended to be delivered in one session, whereas Modules 2, 3 and 4 each usually took two sessions to deliver. Therapists reported a particular challenge with discussing with participants how they put their goals/objectives into practice, as this item tended to be rated as ‘no’ or ‘partially completed’ across the modules. Therapists reported it was sometimes challenging to keep participants engaged but most therapists were able to complete all sessions, with adaptations being made in some cases (e.g. delivering some content just to the carer if it was too challenging for the person with an intellectual disability) and therapists commented that some of the materials needed improvements or revisions. There was also variation in whether home practice tasks were completed.

All available audio recordings were coded for fidelity. Fidelity for the general item scores were high across every module (Mean: 19.5, SD: 2.5, range: 0–22, max score=22). There was more variability within the module specific items, with lower fidelity for Modules 3–5 (see Table 5). Some of the main areas where therapists scored lower included reviewing previous sessions, encouraging participants to engage in role plays, delivering the mindfulness activity and setting home practice tasks using SMART objectives. Coders also noted that therapists experienced more challenges when delivering the therapy to triads, and this impacted the ability of the person with an intellectual disability to contribute meaningfully to sessions.

Seven dyads and one triad received the intervention between April 2023 and February 2024. On average, the therapy was delivered in 11.5 weeks (IQR: 1.24, range: 7–22 weeks). For 7/8 participants, the intervention was delivered within the set period of 14 weeks. One participant received the therapy over 22 weeks, and this was due to holidays and limited participant availability. The average session length was around 1 hour (range approximately 20–90 minutes). Five participants received the full intervention (all modules and at least one review session delivered). Two participants did not receive Module 4 ‘Emotions’ as the therapists deemed this as not relevant or appropriate in those specific cases. One participant did not receive the review session and therapy for another participant was terminated before modules 5–7 were delivered due to significant delays with delivery and the modelling stage having been completed. Participants received a mean of 10 sessions (SD:2.27, range: 6–13).

Qualitative findings

Following the delivery, we completed 19 interviews with carers (n = 8); therapists (n = 5); supervisors (n = 3); service managers (n = 2), persons with an intellectual disability (n =1) once the therapy was complete. The participants with an intellectual disability who had the capacity to consent (n =2) were both invited to participate in the interviews, but one person declined. Written feedback was also obtained from 2 supervisors and 2 service managers. The findings are presented in accordance with the Normalisation Process Theory (NPT) constructs. There are various sub-domains of the theory, however only the sub-domains relevant to the analysis are presented.

Coherence

This domain primarily focuses on the purpose and benefits of an intervention, and whether it fits in with the overall goals and activities of an organisation.

Communal specification

Carers highlighted the importance of working collaboratively with the therapist to maximise therapy benefits, including to establish the best method of delivery to best support the needs of the person with an intellectual disability. In some cases, this required additional meetings or phone calls between carers and therapists to discuss the timing and location of sessions, to adapt information and to best support the person’s needs.

“I think it was, there was a little bit of a warming period as well, where we got to figure out what is the best time for [the participant] as well in terms of his mood, is it best to do it in the morning before he, when he has the whole day ahead, or is it better to do it at the end of the day when he's got nothing else to worry about, trying to figure out what the best time is for that, and over the course of a couple of sessions, we managed to figure it out.” (Carer 2)

Differentiation

It was acknowledged that some of the PETAL therapy content was similar to methods that were already being used by individuals or recommended in clinical practice. Whilst some carers felt that therapy activities such as ABC charts were less useful because they were already using these methods, others reflected that covering these techniques aided their learning and understanding.

“It was very educative because there are a lot of things that we have been doing with him that we never knew had names, or people had done them before. But when we are doing this study, we find out that these are things that we are doing, these are the names of the things that we are doing. So, it was more educative, and it was also very enlightening.” (Carer 7)

It was also felt by supervisors and therapists that PETAL is more structured than other commonly used interventions and takes a more holistic approach, which focuses on building on the person’s strengths to help address aggressive challenging behaviour.

“I really liked that PETAL therapy drew on many aspects of somebody, so their communication, what they enjoy, their relationships, activities, staying healthy. And that the behaviours of concern and kind of focusing on what is triggering those or what might be reinforcing those was actually just one module or one aspect of the therapy. So yeah, I think it's similar, but I think it actually incorporates more than we would normally do so if we're just doing a psychological intervention.” (Supervisor 1)

“I think because PETAL kind of covers such a wide range of topics that we wouldn't normally spend so much time considering in just like a standard piece of PBS work. I guess like we wouldn't normally spend a whole module or like 2 hours talking about healthy habits, for example, or communication from a psychology perspective.” (Therapist 2)

Individual specification

In terms of the individual understanding around PETAL tasks and activities, both carers and supervisors raised that there were some difficulties with understanding the PETAL content. At times, the format or language used was too complex and this made it challenging for the person with a learning disability to understand and engage. It was also noted that some of the content was not relevant to every person.

“The main things that were coming up were bits being not accessible to some service users, so some bits just feeling more relevant, feeling less relevant, which I know the team kind of said you can leave things that don't feel so relevant, but just in the format or the content or the language maybe being less accessible to people with more moderate learning disabilities or that the service users just weren't like tolerating some of those discussions.” (Supervisor 1)

However, some carers found that although not every technique and skill covered was beneficial to the person they care for, they still placed value on the opportunity to test out a range of different tools to see which worked best for the person and those which may continue to be useful moving forwards.

“It was, I mean, it was not repeated, but it was using different tools, you know same sort of thing. But it was just seeing which tool would work. So, I don’t think they were unuseful. I mean what he finds useful afterwards would be different, you know, I mean that would be for him to find is that, you know. What would he like to use, you know? Would he like to use smiley faces? Would he like to use colours? Would he like to use words or whatever? So, I think, I think they were all useful really.” (Carer 6)

The person with an intellectual disability found the resources and content of the therapy useful, particularly the use of the workbooks during sessions.

“The book, I liked the book, and we did pages from the book so only the feelings that I have and the things that I have and so if I’m in a good mood I`m in a good mood now, in a good mood, I don’t feel sad, when I`m happy I don’t feel, I’m smiling.” (Person with an intellectual disability)

Cognitive participation

Cognitive participation addresses whether stakeholders buy-in to the intervention and are willing to engage and commit to it.

Enrolment

Carers and the person with an intellectual disability reported they mostly took part to support behavioural improvement for themselves/the person they care for and to encourage positive change. They felt that participation may help them to further understand the behaviour and to learn strategies to address it.

“So, I want to sort of, I want to see the change. And since they said the aim of the study was to see how we can positively change the behaviours of challenging learning disability patients or persons. I really want to see that change. That's, why, that's what motivated me.” (Carer 7)

“Well, I want to change my anger issues, and want to stay calm and quiet and stop arguing and stop damaging the walls” (Person with an intellectual disability)

Service managers, supervisors and therapists were keen to try out new psychological therapies to add to their existing portfolio. There was an acknowledgement that current practices and interventions do not work for all patients and there was a motivation to find new and effective methods. There was optimism that the PETAL therapy would be relevant to their existing patients and had the potential to be a beneficial approach to address aggressive challenging behaviour.

“I work in a Community learning disability team and a lot of the work we do is with people with a learning disability who use kind of behaviours of concern to communicate an unmet need. And I kind of from my work, I can see that sometimes the approaches that we're using at the moment, like PBS, don't kind of always work for everyone. So I was kind of interested to see the kind of a different approach and see kind of how that would work out” (Therapist 2)

Collective action

This domain explores what is needed to implement an intervention and how compatible it is within existing practices (Murray et al., 2010).

Contextual integration

Carers and the person with an intellectual disability shared that elements of the PETAL therapy were now being implemented outside of sessions and were becoming embedded in the person’s routine and in response to aggressive challenging behaviour. This was perceived as having a positive impact on the person’s behaviour.

“When I get angry sometimes, I go outside and cool off instead.” (Person with an intellectual disability)

“So out of that, we put systems in place which is doing the breathing technique or somewhere or whatever that we put in place, which we've done that and we also introduced kind of like morning yoga sessions with the staff team and stuff like that using the breathing techniques that we spoke about in PETAL as well as kind of stretching exercises.” (Carer 5)

However, from a carer’s perspective, it was not sufficient to just engage with PETAL themselves, it was essential to pass on the tools and techniques to other staff to improve the standard of care for the person and to ensure therapy tools were becoming embedded into routine practice.

“Yeah so everything that I learned from, obviously everything that I learned from the study, obviously I implemented and give them the tools from that as well, so it's a waste of time me keeping it all to myself because they're (other staff) not going to learn and sharing is the best practise so.” (Carer 8)

Service managers felt further information was needed to provide clarity on who the PETAL therapy is targeted for and the potential benefits it has for people with an intellectual disability, although they felt it could be embedded within services similarly to other interventions.

“Yeah, it's an interesting point - why would we offer this instead of something else? And I guess as a team, we would need to think about “Who? Who might this not work for?”. Because I imagine there are a lot of people it could work for, but actually, who might it not work for. What are the things that would make us think “This is not for PETAL therapy”. But I think if we had, yeah, a clearer idea of that and we may do already, as I said, unfortunately I didn't deliver it, or it may be they're already in the manual who it's for who it's not for. Then I think it can be an offer”. (Service manager 1)

Interactional workability

Several logistical challenges of the PETAL therapy were noted, including when scheduling appropriate times for sessions when the person would be most likely to engage and choosing an appropriate location for therapy delivery.

“It was a struggle to get in the exact same time every week, yeah. Afternoons seem to be better, so around 2:00 o'clock or 3:00 o'clock is when we had the sessions and he seems to be much more, much quicker to engage, much less likely to try and change his mind at the last minute and deciding not wanting the session, so that worked pretty well in the end.” (Carer 2)

“The first one tried to do it at the home address, but that was just a nightmare. So decided to do it in a clinic environment” (Therapist 1)

Having therapists who were adaptable, flexible and willing to tailor the delivery of the therapy to meet the needs of the person helped to provide a more positive overall experience for the person and their carer.

“No, I can't sing [therapist]`s praises enough, she was fantastic. She very much knew what she was doing, she was really willing to engage and be flexible, that was so key, so huge, nothing was concrete, and everything could be moved around and what was best for him. It's fantastic, I couldn’t sing her praise enough, she is brilliant.” (Carer 2)

There was mixed feedback on whether the content was engaging for the person with a learning disability. In some instances, the person did not engage well with the therapy, the workbooks or with the home practice tasks. Carers reported that they found this challenging to address and drive forward.

“We worked on the communication plan, you know, some of the emotional things. Some of them, the service user didn't want to do. So, we had to try and persuade him to do it. And he did, some of them, but not a lot of them. So, I mean, maybe did every other week, or something like that. You know, he didn't do every single one when it was supposed to be done and nor did we.” (Carer 6)

Equally, there were times where therapists felt it challenging to engage the carer, when it was perceived that the therapy was mainly for the person with an intellectual disability. It was not always acknowledged that carers may have as a big a role to play in the therapy.

“It was as though she’d switched off and I’d try to get across it’s for both of you but it I think she sort of felt like ‘well I'm here to bring me daughter’ and much as I was saying that this is for… what do you reckon?” (Therapist 1)

Skill set workability

Carers felt the person supporting the person with an intellectual disability through the therapy should know the person well, support them regularly and understand their needs. They perceived this as helpful to support with the tailoring of information and to best support the person to get the most out of therapy sessions.

“I think we have a bond of trust, that every time I said to him trust me, he will trust. So yeah it took me, I think it took me a few months to do that but after I think, even with the therapy I was more confident to support him when he had the challenging behaviour than he was in a good mood. So for me it was easier to support him when he was upset because I already knew how to do it” (Carer 3)

Therapists felt the training they received and their previous experiences working with this population supported them to deliver the therapy well.

“Yes, I think my previous experience was helpful, but I think also the training was that we had online was really helpful as well” (Therapist 2)

Reflexive monitoring

Reflexive monitoring addresses how others perceive the intervention and how it can be improved.

Reconfiguration

Both carers and therapists felt some PETAL materials could be further personalised to individuals, such as including more visual resources or additional communication aids for those who need more support to contextualise information and engage with the therapy. This could also include adapting and personalising resources to make them more relevant to a person’s specific interests.

“I wondered if there was a way that it could be almost a little bit more interactive. Like I know that the PETAL therapy’s like for anyone with a like a the whole range of learning disabilities or any level and I wondered if that could be any way to kind of have it more as like a talking mat exercise or like a have just like picture cards which they kind of choose and maybe the therapist can like take a photo of the way it was laid out. And so it's recorded that way rather than expecting them to write everything down. (Therapist 2)

Individual appraisal

The person with an intellectual disability reflected that the PETAL therapy has made a difference to them and has helped them to make positive changes beyond the intervention.

“It helped to change the stuff I used to do, like damage walls things around the place and set my mind on things to do and set my mind on more stuff. And my anger issue, I don’t have no anger issues no more.” (Person with an intellectual disability)

Many carers also reported the positive impacts of the therapy in improving their ability to support the person they care for, knowing when and how to intervene, and feeling better equipped to address specific situations. Carers placed value on the skills and information they obtained during the therapy and particularly valued the section on carer wellbeing.

“It was truly fantastic, the concept was brilliant, like I mentioned that having something in for the support workers, having it focused around the staff that was supporting him as well, and how he communicates with staff as well as how the staff communicate with him, all of those brilliant.” (Carer 2)

“I was amazed that there was a session specifically for carers and support workers. I thought that was brilliant, that really took me by surprise.” (Carer 2)

However, one carer felt that the PETAL therapy was not appropriate for the person they care for and felt the therapy may be better suited for a younger demographic, as behaviours and routines for some adults become embedded and may not be easily changed.

“I can understand this more for people perhaps in becoming teenagers and the family still carers but I'm thinking of the family that they lived with all the time, certain behaviours changes as they become teenagers and that, and that I can understand but I felt for us in our situation there really wasn’t a lot to get out of it to be perfectly truthful.” (Carer 4)

Finally, after a very challenging beginning to delivery, this therapist felt like they weren't making any positive impact, but as they continued through the programme and began reflecting, they realised that they and the participant had achieved quite a lot, and the work was making a significant difference.

“At first I was like oh this is really challenging, we are not getting anywhere… but as I kind of got half way, near the end, I started to reflect a bit more. And when I did reflect, it was like no actually we did manage to get lots of things done, we completed all these things, I did it, he did it, we did it, we did achieve… and the achievements do count, and so when you count the achievements at the end it does make you happy and proud, so I did get a taste of that at the end.”(Therapist 5).

How these findings have informed our cluster randomised-controlled trial

The modelling study was beneficial to highlight some areas of refinement for the PETAL therapy and the plan for delivery. The changes made for the main trial are summarised in Extended Data Table 5.

Strengths and limitations

Our main trial will be run across England, Scotland and Northern Ireland; however, we were only able to run the modelling study in England due to capacity issues and delays at the other sites. Some modelling has since been conducted in Scotland, although this was outside of the time frame for the completion of the modelling phase. Additionally, we were only able to interview participants who completed the intervention, although we did receive information from the two participants who dropped out of the study. Only one participant with an intellectual disability completed the interviews, as most participants lacked capacity, and one person declined. In the full trial, we will conduct a process evaluation and will aim to also capture the opinions of those who are allocated to therapy but do not complete it, and those who are allocated to receive treatment as usual.