Realtalk Evidence-Based Communication Training: Development of a Conversation Analysis-Based Intervention Designed to be Delivered by Clinical Trainers

BACKGROUND Communication skills training for professionals caring for people with life-limiting conditions is in demand. Evidence shows that transfer of skills into practice is limited. Training can be improved by building on in-depth evidence and reecting the complexity of actual interactions. Scientic advances from direct observational ‘conversation analytic’ studies of conversations in healthcare and beyond provide this evidence, but the linguistic complexities are often inaccessible to clinical trainers. We address this dilemma via a novel training initiative, ‘RealTalk’, designed to support trainers by increasing the authenticity and evidence base of end-of-life communication training for health and social care staff and trainees. We conducted a development study to create with stakeholders, we identied the rationale for the and mode of converted research ndings into and tested and rened the to launching the nished product. School of Social Sciences and Humanities at Loughborough University (UK). LJ examines audio and video recordings of real-life interactions in a range of everyday and medical settings using conversation analysis. Laura has designed and delivered video-based communication training materials for health professionals including in end-of-life care. Her recent articles have been published in Social Science and Medicine, Applied Linguistics, Patient Education and Counselling, and BMJ Supportive and Palliative Care. sciences


Background Communication training: room for improvement
The matters about which healthcare practitioners communicate, and the manner in which they do so, can have signi cant impacts on patients' behaviours, satisfaction, attendance, clinical outcomes, and rates of complaints. [1] Practitioners' communication skills are not an innate and xed characteristic; they can be enhanced via appropriate training. [2,3][4] However, systematic reviews have found that training interventions aimed at enhancing communication in the care of people with life-limiting conditions have shown, at best, only small effect sizes on practitioner communication and patient outcomes, [5,6], and only limited transfer into practice. [7] Recent developments in scienti c knowledge suggest ways in which to improve the content and effectiveness of healthcare communication training.
Research on training from the perspective of organisational psychology has found that training's effectiveness, including its transfer into practice,[8] is more likely where training: is founded on robust, indepth evidence; [9] has content that re ects the complex, context-speci c nature of actual interactions; [10] is closely relevant to trainees' real-life practice; [11] and integrates a range of experiential and interactive components. [11] [12] In current practice, these experiential and interactive components often involve roleplay scenarios and simulations. Recent empirical research has shown that role-play and simulation differ in systematic and important ways from authentic, real-life interaction and its demands. [13] [14] Clearly, this is problematic, given that we know that for communication training to be effective and transferable, it needs to closely re ect real-life practice and the complexity of actual interactions. Furthermore, role-play and simulation in communication skills training can be time-consuming, expensive, stressful for learners, and places high demands on trainers. Role-play has been criticised for failing to help learners develop the analytical skills and exible strategies crucial to real-life, high-stakes situations. [15] As a result, the value of role-play and simulation in the training and assessment of communication practices is subject to debate, including whether and how they could be made to more closely re ect real-life healthcare interactions. [16,17] The direct study of real-life communication: Conversation analysis A substantial and growing body of ndings relevant to the content and effectiveness of healthcare communication training is to be found within communication science, and speci cally in the eld of conversation analysis, which comprises both underpinning observational research on human interaction, and applied research in which the ndings of observational research are translated into communication training interventions. Conversation analytic research has grown rapidly since the mid-1970s. [18,19] It has generated a cumulative body of ndings, many of which are highly relevant to healthcare although not yet well known amongst policymakers and practitioners. [20] Conversation analytic studies provide empirically-grounded understandings of verbal and nonverbal (bodily) communication practices, derived from systematic analysis of recorded real-life interactions. By collecting and analysing multiple episodes of speci c communication tasks (such as practices for opening the consultation, for raising end-of-life matters, or for proposing treatment recommendations), conversation analysts produce detailed evidence on the structure, functioning and consequences of both verbal and nonverbal communicative behaviours. Importantly, conversation analysis yields explicit articulation of communication practices that are normally only understood at the tacit level, and thus these are practices that are only amenable to methods that undertake more in-depth analysis as opposed to techniques that produce broad-brush descriptions. An implication of this is that conversation analytic studies can provide information that cannot be obtained from people's verbal descriptions of communication -be that within research interviews, or within feedback on role-played or simulated interactions within training.
A number of studies have reported translating conversation analytic ndings into training interventions and evaluating their effects. Findings have been promising, with such training found to signi cantly enhance practitioners' communication in actual practice. [21][22][23][24][25][26] To date, however, interventions based on conversation analysis have largely targeted highly speci c communication practices, which trainees are taught to replicate in practice (e.g. how to word speci c questions for speci c moments within consultations). [23,24] Also, many have required that a trained conversation analysis researcher deliver the intervention. [22,27] Context and rationale of RealTalk's development Intervention development studies such as ours are designed to report the "rationale, decision-making processes, methods and ndings which occur between the idea or inception of an intervention until it is ready for formal feasibility, pilot or e cacy testing prior to a full trial or evaluation"[28] (page 1; our emphases). The design and realisation of the RealTalk initiative took place within the 'VERDIS' project which entails video-based communication research and training in decision-making in supportive and palliative care [29]. We used a conversation analytic approach to communication training, to develop evidence-based resources for use in training, aiming at improving practice in conversations relating to end-of-life and palliative care.
Our aim was to develop a training resource that would: Increase the authenticity, relevance and empirical evidence base of communication training for health and social care staff and trainees.
Facilitate learning and re ection on complex, multifaceted communication tasks, whilst avoiding the shortcomings of role-play and simulation.
Be easily accessible to trainers and trainees, because it would not mandate previous expertise in conversation analysis. As a result, it would be widely and easily disseminated within public sector and charitable organisations that deliver communication training.
Be suitable for use 'off the shelf' by trainers, with written materials providing extensive guidance on using the resources.
Be usable and relevant for a wide range of face-to-face training events facilitated by diverse trainers, in diverse settings, and for diverse participants.
In the next section, we provide describe the methods by which our resources were developed, and in the results section that follows we describe the contents of our innovative training materials.

Methods
The process is described here by which we transformed our underpinning analysis into the RealTalk communication training resources. Our description is organised into phases: The initial phase involved identifying relevant communication topics, after which we undertook systematic conversation analysis of our dataset of audio and video recordings. Next, we used our ndings to build video resources with accompanying documentation, and piloted our materials with experienced clinical educators.

Identifying communication topics for the training intervention
We decided which broad communication topics to focus upon, on the basis of: reviews of qualitative studies investigating preferences and perspectives of patients, companions and practitioners; [30,31] priorities articulated by the UK's National Palliative and End of Life Care Partnership; [32] discussions with educator and clinical colleagues at four UK hospices who had experience of running extensive communication training programmes; discussions with teaching colleagues and our project advisory group, which comprised lay consultees, academics, educators and clinicians.

Conducting underpinning analyses
The VERDIS research programme involved collecting audio and video recordings of actual consultations within a large UK hospice providing inpatient, outpatient and day therapy services to people with a variety of life-limiting conditions and their companions. The dataset comprises recordings of 37 consultations, which involved ve experienced palliative medicine doctors, 37 terminally ill patients, and 17 accompanying companions.
The VERDIS research project comprised a series of interlinked studies of palliative care communication using the systematic method of conversation analysis to examine these real-life recorded healthcare scenarios. [29] Our focus was on the ways in which doctors facilitate talk about end of life, [33] and how patients ask, and doctors respond to, questions about remaining life expectancy. [34] Compiling documentation Learning points were identi ed based on our previous analyses, [29,33] ndings from conversation analytic studies, [35] and our previous work informed by the conversation analytic literature.
[36] We selected a series of video clips to exemplify these learning points. We did not disguise participants' faces or physical features in the clips, since bodily gestures and facial expression are vital features of communication. All participants who appear in the videos had given written informed consent for their image to be shown in this fashion. Any references to people or places were edited out in order to ensure con dentiality. The participants' real names and organisation place names were replaced with pseudonyms.
A detailed synopsis was added to accompany each case study, providing background and description of what occurred over the course of each patient consultation. Working in conjunction with technical teams specialising in digital educational materials, [37] branding, and web-design,[38] we compiled the resulting written documentation and video clips to develop a comprehensive written manual [39] and an accompanying digital media resource.
The manual contained all of the training materials developed, with the exception of video clips. All materials, including video clips, were housed on a password-protected website www.RealTalkTraining.co.uk, designed to be accessed only by eligible registered parties.
All video clips could be played with or without subtitles, and included the capacity to stop, start and replay the recordings.

Piloting of the training resource
Trainers were identi ed and recruited from the NHS, higher education sector and hospices, via meetings, conferences, and organisational newsletters, and through direct contact with communication trainers known to the project team and advisory group. not make a pro t from the training they provide.
Trainers were supplied with a registered password to access the online resources, and were required to agree in writing not to make copies of the video recordings.
Evaluation of the acceptability and utility of the resources was carried out as part of each training event.

Results
In this section, we describe the content of the RealTalk materials, and where these have been used. The speci c learning points identi ed for RealTalk were organised into topic modules. Each module focuses on a speci c area of communication and comprises a number of cases from the recorded video consultations. Two core modules formed the basis of the training resource, entitled 'Broaching dying' and 'How long have I got?' (or, more formally, 'Life expectancy enquiries').
The resources were organised such that individual components, cases or communication themes could be selected and used as needed, depending on the participants, the topics needing to be covered, and the level and/or duration of the training being provided.
An overview of the resulting structure and contents of the RealTalk training package is provided in Figure   One.
The initial chapters of the manual introduce the materials by including: general instructions for use; a summary of the safeguards for use (as described in more detail below); a checklist for preparing and delivering training using RealTalk; an evaluation form for trainees; and suggested reading. The training modules follow, each containing several patient cases. For each of the patient cases there is a synopsis which includes: keywords to enable users to cross-reference patient conditions and communication topics across the resources; a case summary, including the patient's condition and reason for attending; description of the clips which are included in the patient case; and an overview of the patient's entire consultation. In addition, for each patient case there are transcripts for each clip, speci c evidence-based learning points, and further reading.
Trainers were required to agree to uphold and share the list of ethical safeguards with trainees before showing video clips. These safeguards were based on a literature review [20] and empirical interview study, [40] and are available as a supplementary le.
The RealTalk website was accessed by a range of trainers from diverse organisations, including Higher Education Institutions, NHS Trusts, and hospices, across the breadth of the UK and Channel Islands. The resources were used in events for trainees including specialist palliative care professionals and students studying medicine and nursing. To date, more than 350 trainers have registered to use RealTalk.

Discussion
The RealTalk training resources were designed to be evidence-based, authentic, relevant to real-life practice, and exible enough to be used in delivering face-to-face communication training for health and social care professionals and trainees. Rather than training people to adopt a speci c model or script in communicating with patients, RealTalk facilitates re ective learning of principles and patterns, which may be adapted to individual circumstances and conversations. Crucially, despite being grounded in ndings derived from the close and rigorous analysis of interaction by experienced conversation analysts, the RealTalk resources do not require that trainers have advanced expertise in the complex methods and theories of conversation analysis. We have achieved this by uniquely translating technical linguistic ndings from communication science into accessible and usable learning points within each training module.
Feedback to date has indicated that prior face-to-face training in use of the RealTalk resources is useful, but not essential. For example, a series of 'train the trainer' events that we have run was positively received. [41] Our aim is that the resources should ultimately be suitable for use 'off the shelf', and not necessarily require prior specialist training. The training instils fundamental conversation analytic principles to approaching communication. RealTalk trainers ask trainees to avoid making evaluative judgements about the conversations within the video clips. Instead, trainers encourage trainees to describe the talk by asking "What did you see?" and "What did you hear?" The value of the RealTalk resources lies in the foundation of robust, in-depth evidence. Further, the training materials re ect the complex, context-speci c nature of actual medical interactions; they are closely relevant to real-life practice; and integrate a range of experiential and interactive components. These features have been found to improve training's effectiveness, including its transfer into practice. The key limitation of our ndings at present is that the effectiveness of RealTalk in enhancing trainers' or trainees' practice cannot yet be assumed, and this will require full evaluation on a larger scale. An evaluation of the acceptability and utility of the RealTalk training resources is currently in progress.

Conclusions
RealTalk is a training intervention that encompasses learning points from cutting-edge conversation analytic communication research. Its development provides a blueprint that may be used by others in developing training resources that draw on recorded real-life interactions. Not applicable (manuscript does not contain data from any individual person).

Availability of data and materials
The full dataset of video and audio recordings is deposited in Loughborough University's secure archive.