Using the Medical Research Council Framework for the development of a communication partner training intervention for people with primary progressive aphasia (PPA): Better Conversations with PPA

Background: Primary Progressive Aphasia (PPA) is a language-led dementia resulting in a gradual dissolution of language. PPA has a signicant psychosocial impact on both the person and their families. Speech and language therapy (SLT) is one of the only available management options, and communication partner training (CPT) interventions offer a practical approach to identify strategies to support conversation. The aim of this study was to dene and rene a manual and an online training resource for speech and language therapists to deliver CPT to people with PPA and their communication partners (CPs) called Better Conversations with PPA (BCPPA). Methods: The BCPPA manual and training program were developed using the Medical Research Council framework for developing complex interventions. The six-stage development process included 1. Review of existing literature, 2. Consultation and co-production 3. Development of an initial draft 4. Review by a group of speech and language therapists, using consensus methods 5. Two focus groups to gather opinions of people with PPA and their families 6. Renement. Results: The BCPPA manual and training program comprises seven modules, and four CPT intervention sessions. Eight important components of CPT were identied in the consensus work undertaken with speech and language therapists, including use of video feedback to focus on strengths as well as areas of conversation breakdown. Three themes arose from the focus groups with people with PPA and their families 1) Timing of intervention, 2) speech and language therapists’ understanding of types of dementia, and 3) Knowing what helps. Renements to the manual included additional practice activities and useful strategies for the future. Using MRC theoretical rationale based Co-development Gathering

Results: The BCPPA manual and training program comprises seven modules, and four CPT intervention sessions. Eight important components of CPT were identi ed in the consensus work undertaken with speech and language therapists, including use of video feedback to focus on strengths as well as areas of conversation breakdown. Three themes arose from the focus groups with people with PPA and their families 1) Timing of intervention, 2) speech and language therapists' understanding of types of dementia, and 3) Knowing what helps. Re nements to the manual included additional practice activities and useful strategies for the future.
Conclusions: Using the MRC framework to develop an intervention that is underpinned by a theoretical rationale based on a review of the literature, increases the likelihood of behaviour change. Codevelopment of the manual and training materials ensures the intervention will meet the needs of people with PPA and their CPs. Gathering further data from speech and language therapists and people living with PPA and their families to re ne the manual and the training materials enhances the feasibility of delivering this in preparation for a phase II NHS-based randomised controlled pilot-feasibility study, currently underway.

Background
The number of people living with dementia worldwide continues to rise, estimated at around 50 million at present with nearly 10 million new cases each year 1 . Of these, perhaps a half a million people worldwide and several thousand in the United Kingdom have primary progressive aphasia (PPA): a group of language-led dementias associated with Frontotemporal Dementia and Alzheimer's disease 2 . PPA presents as an insidious dissolution of language skills with relative sparing of other cognitive functions 2 .
At present there are three internationally recognised PPA variants; people with semantic variant (svPPA) experience a gradual loss of word meanings affecting both comprehension and naming, people with Many speech and language therapists report delivering CPT to people with PPA and describe using resources developed for stroke aphasia or brain injury related communication di culties 12 . CPT has a growing evidence base in stroke aphasia 16,18 and delivers positive changes in the conversation skills of people with aphasia as well as their CPs 19,20, . However, CPT approaches in stroke aphasia are not designed to meet the needs of people with progressive communication di culties. Currently there are only case study reports of CPT for people with PPA 21,22 . There is some suggestion of increased communicative effectiveness as a result, however, it is di cult to attribute these gains to CPT due to the fact that individuals were concurrently participating in additional interventions. Thus, there is a clinical need to develop a CPT intervention designed to meet the needs of people with PPA and their families 6,23,24 .
To our knowledge there has been no speci c research undertaken asking people with PPA and their families what interventions are important or need to be developed. People with PPA have written about their general experiences of SLT and the value of developing "a wide range of personalized strategies that continually evolve as the disease progresses" 25 . Spouses report a need to develop practical approaches to deal with communication di culties and maintain a close bond with their loved ones 6 . These issues are more likely to be met by tailored interventions, that build capacity by helping them to adjust and reframe their communication over time 6 . Speech and language therapists themselves have identi ed a need to engage family who are motivated to understand how they can best support their loved ones 26 Therefore, gathering ideas and contributions of people living with PPA, often described as Public Involvement (PI), is important to ensuring an intervention will meet their needs. PI is de ned by the UK Standards for PI as research that is carried out with members of the public rather than to them. These standards include ensuring that people are involved as early as possible and that participation is made accessible. The aim of this study was to work with people with PPA and their families, from the beginning, to co-produce a CPT intervention to meet their needs.
The Medical Research Council provide a framework for developing and evaluating complex interventions 27 . The guidance outlines the importance of preliminary development and testing of an intervention's procedures prior to piloting and evaluation. This paper therefore describes how the MRC framework was used to develop Better Conversations with PPA (BCPPA), a 4-session, manualised, CPT intervention to help people with PPA and their CPs to identify and practice strategies to reduce barriers (such as interjecting when a person may not have nished) and increase facilitators in conversations (such as giving more time). A manual and an online training resource for speech and language therapists, hosted on a life-learning platform at UCL, were developed to enable speech and language therapists to deliver the intervention. In line with stages 1 and 2 of the MRC Framework the underlying theory and proposed mechanisms of change for the BCPPA program will be described as well as primary research which informed the co-production of the manual and online training resource.

Aim
To use the MRC framework for developing complex interventions to de ne and re ne a manual and an online training resource for speech and language therapists to deliver BCPPA to people with PPA and their CPs.

Methods
Intervention development activities were based on phases one and two in the MRC framework for development of complex interventions 27 . This comprised six stages including 1. examination of existing literature, 2. consultation and co-production work, 3. development of an initial draft, 4. consensus work with speech and language therapists, 5. focus groups with people with PPA and their families, 6.
Re nement of the BCPPA intervention and manual in preparation for the randomised controlled pilotfeasibility study. Figure 1 demonstrates how these activities map onto the MRC guidance. Intervention development also followed the GUIDED guidelines for reporting for intervention development studies 28 . Further patient and public involvement work undertaken to nalise outstanding training modules identi ed as supplementary to the RCT will not be discussed here. The rst author, A.V., an experienced speech and language therapist, led all stages.

Recruitment
Consultation and co-production work (Stage 2): An opportunistic sample of people with PPA and their families, specialist speech and language therapists and neuropsychologists were invited to join the project steering group. A.V. emailed people who were known to her through clinical work, and asked the facilitator of the PPA branch of the Rare Dementias Support Group based at UCL (https://www.raredementiasupport.org) to forward an invitation email to individuals in the support group, inviting them to participate.
Consensus work (Nominal Group Technique, NGT) with speech and language therapists (Stage 4): speech and language therapists were recruited to participate in the NGT consensus study through the RCSLT Dementia and Mental Health Clinical Excellence Network, of which A.V. was a committee member.
The NGT was carried out at one of the Clinical Excellence Network meetings and advertised in the RCSLT clinical practice magazine (Bulletin) and via emails circulated to members two months prior to the study day taking place.
Focus groups with people with PPA and their families (Stage 5): People with PPA and their families who attend the PPA branch of the Rare Dementias Support Group at UCL were invited to participate in one of two focus group meetings held at an accessible venue on the university campus. The aim was to recruit eight people to each focus group, totaling 16 participants. To optimize opportunities for individuals with communication di culties to contribute to discussion 29 , group numbers were capped at eight participants. Potential participants who responded to the advert were contacted by A.V. on the telephone to judge if they met the inclusion criteria of a) a diagnosis or possible diagnosis of PPA/relative with such a diagnosis, b) the ability to communicate to participate in a focus group c) see and hear well enough to participate d) English as their language of daily use. Potential participants were excluded if they had a) a history of brain lesion or major head trauma, b) major physical illness or disability which could impact on participation. criteria required.

Examination of existing literature (Stage 1)
Existing models of dementia, principles of applied Conversation Analysis (CA), behaviour change theory and frameworks for chronic disease self-management were explored. This informed the preliminary contents and focus of the intervention.
Consultation and co-production work (Stage 2) Four people with PPA and their spouses, two expert speech and language therapists, a neuropsychologist and the group facilitator (A.V.) took part in 12 formal BCPPA public involvement (PI) steering group meetings. Work to co-produce the BCPPA intervention materials and training modules was informed by feedback from people with PPA who had previously received CPT 30 , research undertaken by A.V. 10,11,12 and research into the BCA program for people with stroke aphasia 31 . In order to support communication, steering group members were informed of the topic for discussion in advance of each meeting and invited to contribute in advance, during or after meetings using verbal, written or visual means, e.g. bringing photos, drawing pictures or re-assembling draft materials.

First draft of the manual (Stage 3)
A draft of the BCPPA manual was developed using PowerPoint software. In order to upload these to the UCLeXtend website an online software package called Articulate was used to adapt the PowerPoint slides to an appropriate format. The work was undertaken with assistance from speech and language therapist researchers and four postgraduate researchers in speech and language sciences who were paid for their time.
Consensus work (Nominal Group Technique) with speech and language therapists (Stage 4) Draft one of the manual was made available to a group of speech and language therapists, who were attending a Clinical Excellence Network study day. In order to gain an understanding of the clinical experiences and reality of speech and language therapists a qualitative research method was identi ed as appropriate. Speech and language therapists were encouraged to review the resource and pilot it with their clients. To ensure the BCPPA intervention re ected a consensus view of the most important components to include in a CPT intervention for people with PPA and their families a Nominal Group Technique (NGT) method was chosen. Given that many of the speech and language therapists participating in the study day had pre-existing professional relationships that could result in certain voices being represented over others in discussions, the NGT method was also chosen to provide opportunities to consider ideas and experiences equally, yet allowing for clari cation and discussion prior to rating 32 .
Six weeks prior to attending the study day speech and language therapists were sent an email inviting them to anonymously complete a 12-item feedback survey comprising all open questions (supplementary document 1), hosted online on the Google Forms platform. Survey questions were developed by A.V. in consultation with the steering group and included questions about speech and language therapists' experiences and views on the content and format of the manual.
The NGT meeting itself comprised a two-stage ranking process commencing with a 90-minute group session (stage one), followed by email consultation (stage two). Meeting facilitators (AV and SB) agreed the session plan and central question for discussion in advance (see supplementary document 2), in line with guidelines for conducting NGT meetings 14 . At stage two, results of the group session were circulated via email to all participants, providing information on scores and mean rankings for each item. As per guidelines for conducting NGT meetings 32 , items describing the same ideas from the two groups were merged, following discussion and agreement between A.V. and S.B. Participants were asked to reply via email identifying and ranking their top eight items from this list (by placing a number from 1-8 to re ect which is most important -8 and least important -1). Following NGT guidelines 32 , scores were tallied and mean rankings calculated to identify the top eight ranked items overall.
Focus groups with people with PPA and their families (Stage 5) Two focus groups took place, to provide people with PPA and their families the choice of attending with or without partners. Discussion was guided by the question 'How can speech and language therapists support people with PPA to live well and maintain relationships?'. The focus groups were jointly facilitated by A.V., alongside volunteer student speech and language therapists from UCL (one per focus group). A topic guide was co-produced with the BCPPA steering group and attendees of the PPA branch of the Rare Dementia Support Group at UCL (see supplementary document 3).
Focus group discussions were video recorded and transcribed by UCL student speech and language therapists (using transcription guidance 33 ). Given the researchers objectives to understand the lived experiences of people with PPA and their families, and gather opinions from them, qualitative methods employing a realist approach to re exive thematic analysis was undertaken 34,35 . Initial codes were generated by systematically coding interesting features (phase 2), collating these into potential themes (phase 3) and reviewing them in relation to the coded extracts (phase 4). Potential themes were re ned to generate de nitions and names (phase 5), further inspected to identify and report any additional key elements (phase 6). In addition, to improve reliability of analysis, four speech and language therapist researchers with experience of thematic analysis independently extracted data from a randomly selected section of transcript, discussed and reached agreement on the coding of themes arising from the data.
Re nement of the BCPPA manual (Stage 6) The BCPPA manual was re ned based on work undertaken in stage 4 and 5 ( Figure 1).

Examination of existing literature (Stage 1)
Existing literature comprising the bio-psychosocial model of dementia, applied CA, behaviour change theory and self-management and self-e cacy theory was examined.

Bio-psychosocial model of dementia
The bio-psychosocial model 36 proposes that there are factors other than the organic causes of dementia that in uence the nature and speed of deterioration in daily functioning. These include some factors that are xed, such as PPA variant, that cannot be changed. The BCPPA manual therefore provides practice tasks, to maximise generalisation for people with svPPA, for whom this is more di cult than those with nfvPPA (insert ref to cadorio). Tractable factors, such as the way a CP interacts with a person with PPA, may be amenable to change and are directly targeted in the BCPPA intervention. Adaptive mechanisms used by the CP, such as multiple questions or test questions, may result in the person with PPA feeling incompetent 37 . On the other hand, the use of gesture and enactment (whole body gesture and pantomime) by a person with PPA when they are having di culty retrieving a spoken word 38 could be described as an effective coping strategy. The BCPPA intervention seeks to take account of xed factors whilst targeting tractable factors to support the dyad (person with PPA and their CP) achieve their potential function.

Applied Conversation Analysis (CA)
CA is an approach to the study of human social interaction through the analysis of spontaneous, naturally occurring talk 39 . A number of CA informed stroke aphasia intervention studies and clinical resources have been developed 40 such as SCA 18, SPPARC 41 and BCA 42 . These have in common the analysis of video recordings of natural conversations between the person with aphasia and their CP, and providing these as video feedback, as a foundation for targeting therapy 40 . The speech and language therapist (who typically delivers such an intervention) analyses 10-15 minute video-recorded interaction to identify behaviours resulting in conversational breakdown, known as barriers, and ways in which members of a dyad successfully resolve or circumvent troubles to maintain interaction, known as facilitators. The aim of video feedback is to increase awareness in one or both members of the dyad of the impact of their behaviours, and jointly agree on goals for therapy. Once the goals of therapy are agreed upon, a process of practice, through supported conversations, role play and re ection, is commonly employed 40 . The BCPPA intervention is informed by this well-described 43 , CA-underpinned approach to CPT.

Behaviour change theory
Recognising conversational barrier behaviours in video recordings of oneself and setting a goal to cease these, or adopt facilitative strategies instead, does not guarantee that a change in behaviour will occur 44 .
Behaviour change theory, speci cally the COM-B model 45 accounts for an individual's behaviour change as the product of three equally weighted components namely Capability, Opportunity and Motivation. Researchers examined video recordings of CA-underpinned CPT being delivered to people with stroke aphasia and their CPs 46 and used the COM-B model 45 to identify the essential change processes and the core procedures that serve them 47 . The BCPPA intervention incorporates the seven core mechanisms that have been identi ed as essential to behaviour change in a CPT 46 , speci cally the processes to motivate change and those that embed changes (See supplementary material 4).
Self-management and self-e cacy Central to self-management is the concept of the client as an active participant whose current status is in uenced not only by diagnosis but by psychological responses and experiences. This implies interventions should address the ability to self-manage daily activities and the emotional journey, not just medical symptoms 48, 49 . Taking action to accomplish a plan to self-manage their condition is more likely to succeed if a person has the con dence or self-e cacy to achieve it 50 . Self-e cacy is a mechanism that directs behaviour change, for if one feels in control of a behaviour it becomes easier to make a change to it 50 . Five core self-management skills and four key self-e cacy mechanisms have been highlighted for inclusion in speech and language therapist interventions with people with progressive communication di culties 49 and these have been considered in the development of the BCPPA intervention (see supplementary le 4).
Consultation and co-production work with the steering group (Stage 2) Decisions made included: 1. Identi cation of seven subjects to form distinct training modules within the BCPPA program. Table 1 provides an overview of the learning objectives and how these were co-produced. The three modules required for the phase II NHS based randomised controlled pilot-feasibility study (Module 3: How to make a video, Module 4: What to target in therapy and Module 5: the BCPPA therapy) were prioritised for development over the four only needed for the future general release of the online BCPPA program. Table 2 provides an overview of the content of these three modules.   intervention components as either core or non-essential components that can be tailored to an individual's needs.
The draft manual was evaluated by the steering group to ensure information was presented in an accessible way. This included decisions on images and formatting.
The rst draft of the manual was uploaded to a secure area on the UCLeXtend website and made available to speech and language therapists participating in the stage 4 consensus work via a bespoke URL. It was not publicly accessible.

Consensus work (Nominal Group Technique) with speech and language therapists (Stage 4)
Demographics and characteristics of speech and language therapist participants Thirty-six speech and language therapists took part. Of these, 17 had completed the pre-NGT meeting survey, 22 had viewed the rst draft of the BCPPA manual and training program prior to attending, and two had been able to use the BCPPA manual with a client with PPA. Table 3 presents speech and language therapist participant demographics and their familiarity with the BCPPA manual and training program. Following the meeting, 20 of the 36 participants completed the nal NGT ranking task by email.   Figure 2. Notably, access issues were generally related to glitches in the program, though some local NHS browser systems posed restrictions.

Nominal Group Technique
After two iterations of consensus work with speech and language therapists, focused on the question "What components of the BCPPA therapy sessions are important for people with PPA and their conversation partners?", eight components were identi ed, and ranked in order of importance, see Table  4.  6 Working with both the person with PPA and the CP together.
7 Providing opportunities to practice strategies and get feedback from the SLT.

Providing an opportunity to discuss their communication difficulties
Focus groups with people with PPA and their families (Stage 5)

Demographics of participants
Thirteen participants, six people with PPA and seven family members, responded to the advertisement. All were eligible and agreed to participate but one couple withdrew the day before the focus group due to a con icting commitment. The remaining 11 participants attended two focus groups (NB: these were mixed groups, whereby people with PPA and their CPs attended together, alongside some CPs and people with PPA who attended independently, group 1: seven participants; group 2: four participants). Participants with PPA represented all three variants, and atypical mixed variants. Demographic information is outlined in Table 5.  Re nements that arose from stages 4 and 5 of intervention development were decided on in conjunction with the steering group and are presented in Table 6. The re ned BCPPA program was consequently made available to participating local speech and language therapist collaborators on UCLeXtend as part of their training in preparation for delivering the intervention during the randomised controlled pilotfeasibility study. The nal intervention is described in detail, using the template for Intervention Description and Replication (TiDIER), in the authors PhD thesis which this paper is based on 51 , and a published protocol for study which remains currently underway 52 . Further to this, the project steering group made plans to continue working to co-produce the remaining four modules, in anticipation of a future launch of the BCPPA program. This paper is based on work from the authors PhD thesis. Addition of Home based task 2: Strategies to help turntaking and expansion of session plan 3 to include a list of 11 optional additional strategy practice ideas based on ideas collated from SLTs, people with PPA and their families and a review of manuals for stroke aphasia CPT manuals.
Provide more information on resources and other services.
Expansion of session plan 4 to include a list of resources and other services for SLTs making recommendations for the future.
Develop video examples of the intervention being delivered.
Addition of video recordings of conversation breakdown and intervention being delivered inserted to Module 5: The BCPPA therapy. These included:

Discussion
The BCPPA manual and training program were developed using the framework described in the MRC guidelines for development of complex interventions 27  Speech and language therapists report seeing people with PPA in their clinics who feel incompetent in conversations, whilst their CPs feel helpless to support them in these situations 53 . Addressing this by exploring meaningful strategies to maintain conversation via CPT that involves both a person with PPA and their CP has been recommended by expert speech and language therapists 26 . Currently, speech and language therapists delivering CPT to people with PPA and their CPs report using tools designed for people with stroke aphasia because there are no PPA-speci c materials 11,12 . The BCPPA manual and training program address this gap in the speech and language therapists' "toolkit" (described as such by participants in the focus groups) of interventions for PPA, and provides an evidence based, manualised training resource designed by and for people with PPA and their CPs.

Strengths And Limitations
Drawing on the best available evidence and appropriate theory to develop the BCPPA manual, in accordance with MRC guidance 27 , should increase the likelihood that components of the intervention result in behaviour change. Extensive use of theory has been associated with larger effect sizes in a review of online behaviour change interventions 54 . This work has involved new research with those targeted by the intervention as well as those delivering it.
There are, however, some methodological limitations. NGT does not allow for anonymisation in the way that other consensus methods such as Delphi do, and can thus bias the responses of participants.
Though, the NGT did provide a method of involving large participant numbers and incorporating mathematical voting techniques to aggregate group judgements equally 32 . Making the intervention manual available for scrutiny of its practicality for clinical practice in this way has allowed for prioritisation of key components in anticipation of the phase II NHS based randomised controlled pilotfeasibility study. Notably, only two males were recruited to the NGT, though this is generally representative of the current SLT community 55 . Despite being a useful method for eliciting participant's genuine and honest opinions, a focus group can be a challenging communication environment 56 . The role of the speech and language therapist facilitator and the student speech and language therapist co-facilitators was to mitigate this by enabling participants to contribute to discussion. The option to attend with CPs to support communication was also provided, but instead participants prioritised the convenience of meeting dates and times. Given the steering group was established a number of years prior to the recently published practice standards for PI 57 it is likely that the methods employed may have limited the effectiveness of the co-produced work. Some have criticised the steering group model for consulting with only a small number of individuals. There were only three couples with PPA in this group and that may have limited its value. PPA is, however, a relatively rare condition and people were approached to re ect the known diversity within the condition. Additionally, new members were sought when others withdrew due to disease progression, and the author sought to gather perspectives of other people and their families through individual telephone contact. Despite approaching professionals from other disciplines, including medicine and social work, interested individuals were not able to attend steering group meetings. The author was able to consult with the research team, including neurology colleagues, to gather feedback and ideas.
A manualised approach enables standardised delivery of the intervention for a future trial. Given that speech and language therapists in clinical practice may have limited experience of working with people with PPA 10,11 , this helps to maximise ease and delity of delivery for future implementation. However, a manualised intervention may limit the potential to tailor an intervention to individual clients, for example by deciding not to use video recording or by delivering the intervention to a person accompanied by two CPs. Person-centred components have been identi ed as important for functional communication interventions for people with PPA, and have been highlighted as important for behaviour change 46,49 . The development of this intervention took behaviour change theory into account and embedded the core processes and mechanisms that had been identi ed in previous CPT research as essential components. These were clearly signposted in the manual and distinguished from non-essential components that were amendable to tailoring. Furthermore, expecting four 1-hour therapy sessions to result in a change may seem ambitious. However, the decision on dosage was made based on the average number of sessions that speech and language therapists reported having available to deliver functional communication interventions for PPA 11 . Developing an intervention that meets this requirement increases the chance of implementation.

Conclusions
The six-stage process of development included a review of existing literature, and consultation and coproduction with the project steering group to develop an initial draft. Consensus work undertaken with speech and language therapists and focus groups with people with PPA and their families identi ed further re nements. The BCPPA manual was re ned in preparation for a phase II NHS based randomised controlled pilot-feasibility study which is currently underway 52 .

Declarations
Ethics approval and consent to participant All work undertaken in this study was was conducted in accordance with the Declaration of Helsinki.
Consultation and co-production work (Stage 2): Ethical approval is not required when involving individuals in the planning or design of research, for example when they are members of a research steering or advisory group (Health Research Authority, 2019). In order to equalize participation and power issues speci c strategies were used in and outside of the steering group meetings. All communication, written and spoken, was made accessible to ensure no individuals were disadvantaged. All steering group members were required to use the same methods to contribute (raising a card to indicate they had a question or comment). People with communication di culties were invited to contribute before others, and the author made contact with individuals with communication di culties prior to the meeting to gather initial thoughts, to support facilitation during group discussions.
Consensus work with speech and language therapists (Stage 4): The UCL Research Ethics Committee con rmed the NGT consensus work (Stage 4) with speech and language therapist participants to be service evaluation. Participants were informed that all responses would be anonymous and at the start of the event they provided written informed consent to participate. from people with communication di culties. A caregiver (a friend or relative) was asked to witness the informed consent process whenever possible.
Participant information sheets, consultee information sheets, consent forms and consultee declaration forms were designed to be accessible to support the process of gaining informed consent. They were designed using a resource for researchers in communication disability "Engaging people who have aphasia" (Pearl, 2014) and modi ed with advice from the project steering group. Transcriptions of focus group data were anonymised via the allocation to each participant of a unique research number, used at all times. All names, places and personal information mentioned in the discussions were pseudonymised.
The addition of video recordings demonstrating delivery of the intervention had ethical implications, requiring a minor amendment to HRA ethical approval. Having received this approval, a separate dyad were recruited through an email advert to members of the PPA branch of the UCL Rare Dementia Support Group. During the consent process it was made clear to the dyad there would be a risk that their faces and voices may be recognized from their video recordings. Information was provided regarding the course registration process and expected registrants, such as health professionals and people with PPA and their families. After consenting to participate, the dyad made four pre-intervention video recordings of their conversations, and received BCPPA therapy from the author, an experienced speech and language therapist. All four therapy sessions were video recorded. The author then identi ed a selection of short video clips that illustrated key components of the intervention such as the process of supporting dyads to identify barriers and facilitators in their conversation sample, goal setting, and discussion about planning for the future. The dyad viewed these clips prior to giving nal consent for their inclusion in Module 5.

Consent for Publication
Not applicable Availability of data and materials Not applicable Figure 1 The six stages in the development BCPPA intervention and manual drafting, mapped on to the MRC framework for development of complex interventions.