a Using linguistic methods in clinical communication education

Analysis and reflection are important components of clinical communication learning in undergraduate medical education. Current medical consultation models do not provide an effective means to analyze interaction during consultations, compromising a conversational approach to consultations. This paper introduces a conversational analytic framework: The Clin-Com Tool (CCT), drawing on interactional linguistics. 17 medical students and six communication tutors took part in an educational intervention. A mixed-methods evaluation was conducted to compare 1) participants’ abilities to analyze consultations pre- and post-intervention, and 2) elicit their perspectives of learning and using the CCT. Results: The findings showed an improvement in participants’ analytic skills in the post-intervention test (p<0.044, 95% Confidence Interval). Participants felt that the CCT heightened awareness of interactional features and socio-cultural effects on communication, and provided a systematic approach to analysis using a set of common language. The CCT emphasizes the development of students’ critical ability to judge and act upon the constantly changing interactional communicative situations. It transforms intuitive feelings into systematic and evidence-based analysis of interaction, enabling the development of more strategic and conversational communication with patients. The Tool can become a useful addition to other communication and consultation models used in undergraduate medial education. Abstract


Introduction
Interactional linguistics (IL) and CCT development actions in social interaction (Kern and Selting 2013). It draws on several disciplines, including conversation analysis (study of systematic sequential organization in talk-in-interaction), discourse analysis (study of language as discourse in context), pragmatics (study of language and social actions it entails) and linguistics (study of language structure). Its aim is to understand how language 'is both shaped by and itself shapes the actions it is used for' (Kern and Selting 2013: 1). We convened an interdisciplinary team, consisting of: clinical communication lecturers, linguists, a clinician, a psychologist, medical students and an English language teacher, to identify relevant IL analytical concepts and adapt them for clinical communication education. The CCT consists of nine analytical terms under three themes, a) interactional analytical theme, b) discourse analytical theme and c) linguistic analytical theme. Below we explain why we selected these terms, what they are, and how they can be used in analysis.

Theme 1 Interactional analytical theme
Effective management of the interaction is the means of conducting a patient-centered consultation ( Mead and Bower 2000) and requires a conversational approach. This requires the doctor and patient to take turns to talk so they co-construct and negotiate for meanings and understandings as the conversation unfolds. Therefore, the three analytical terms the first theme introduces are turn-taking, coconstruction and recipient design (Heritage and Maynard 2006). They constitute a systematic way of analyzing an interaction. We devised a set of trigger questions to indicate what each term analyses (See Table 1

below). Theme 2 Discourse analytical theme
There are four analytic terms under this theme, namely contextualisation cues (Gumperz 1982), framing (Goffman 1974), positioning (Davies and Harré 1999) and social action (Drew and Heritage 1992) . They extend the analytic scope of Theme 1 to allow more detailed investigation. To select the relevant terms among many in Interactional Linguistics, we listed out the challenges our students report to encounter in communication learning. Not all challenges can be addressed by a detailed linguistic analysis but we identified the following 'conundrums', which we think the selected terms may address. Conundrum 1: what is a cue? 'Picking up cues' and 'being responsive to cues' are commonly mentioned by teachers and students in their feedback when discussing how to achieve attentive listening and empathy. However, cues are mainly referred to as a unilateral procedure where the patient gives cues and the doctor picks them up. Cues, however, are interactive and reciprocal. The CCT brings in Gumperz's contextualisation cues to provide a more comprehensive understanding. Gumperz (1982) defines cues as verbal and non-verbal communicative signals speakers constantly give, and listeners seek to understand in conversations. The reciprocity means students need to learn not only to pick up cues but also use cues strategically. Conundrum 2: how to provide a structure in a medical consultation? Students are taught to consult following a structure from the opening to the closing with several milestones to achieve in between ( Silverman, Kurtz and Draper 2013). Although these milestones are presented in sequential phases, they do not always occur in sequence in real consultations. This makes it hard for students to always know which phase they are in, let alone to provide a structure for the patient. The CCT therefore introduces the concept of framing. Framing is about the context, in which meaning is produced and understood. Each consultation phase constitutes a frame, making certain topics relevant and others less so in the conversation. Within each phase, there are a series of topics, which constitute sub-frames and their respective contexts. Understanding the patient's frames can help students understand what topic and Li S, Said F, O'Neill B, Ancarno C, Niksic M MedEdPublish http://dx.doi.org/10.15694/mep.2016.000119 frame the patient is on so that they can be more strategic when negotiating for a structure. Conundrum 3: How do I relate to patient's experiences? Patient-centeredness means doctors need to incorporate patient's narratives into the consultation, making both bio-medical and lifeworld topics equally relevant. Conversations about the latter rely on the doctor's own lifeworld experience, which some students are reluctant to utilize in a professional context. This, nevertheless, hinders empathy. The term positioning means the viewpoint from which people interpret as well as produce meanings in a conversation. Positioning is usually determined by the social and conversational roles interlocutors undertake. Social roles can be that of a doctor, patient, mother, daughter, sister, etc. Conversational roles can be that of a listener, speaker, eavesdropper, etc. They all contribute to meaning making. Positioning allows students to relate themselves to the wide range of viewpoints patients take. Conundrum 4: What do they want? We do not always say what we mean. Saying 'It's cold in here' may not be a statement of a fact but rather a request for someone to close the window. When people talk, they use language to accomplish social actions (e.g. greetings, questioning, comforting, diagnosing, etc.). Language is the vehicle of actions. Analyses of language use in interaction provide evidence for interpreting social actions being conducted with the language.

Theme 3 Linguistic analytical theme
The last theme consists of two terms, register (Fromkin and Rodman 1998), and facework (Brown and Levinson 1987). They investigate the relationship between language structure and achievement of social actions. This section can be particularly useful for non-native English speakers to better understand how the English language functions in social interactions. Such knowledge may as well be useful for native speakers to turn intuitive feelings into more systematic understanding, especially when communication breakdowns are complicated (see examples in Dahm et al., 2015). In order to see if the use of the CCT can improve people's ability to analyze consultations, we conducted an educational intervention and evaluation. The aims were to elicit 1) if the CCT improved participants' analytical skills, and 2) participants' experience of using the CCT. • Note: If these are not realized, then the consultation is highly unlikely to be successful Theme 1 helps you analyze whether a consultation is constructed as an interactional conversation or not. To further analyze how the interaction has been conducted and whether the interactions are successful we need to go deeper to analyze the conversations through Themes 2 and 3 next.

Contextualization cues
Are the subtle signs a speaker gives the listener during conversation. They can be both verbal (e.g. actual words being said, intonation, volume of speech, speed of speech) and nonverbal (e.g. gaining or losing eye contact). They are the means by which speakers signal, and listeners interpret: What the activity is i.
How semantic content is to be understood ii.
How each sentence related to what precedes or follows? It is up to the listener to pick these up accurately and to respond accordingly. iii.

Framing
Is an inevitable process during which we put things into context. It can be a situational context or a topical context, either abstract or specific. We make sense for and of each other by putting words into such a context. It determines how you, as a doctor, communicate with others and influences how you select your recipient design in a specific phase or atopic during the consultations.

Positioning
Is the taking on of different roles during an interaction, for example, a doctor has different roles from a patient and thus also different positions. Positioning can change during the conversation and people may take up multiple positions.

Social Action
Conversation is the way we perform social actions (e.g. asking a question, clarifying a misunderstanding, restoring rapport, negotiating, diagnosing) in order to fulfil certain social functions (e.g. providing care for patient, or seeking healthcare).

Register
Is "a stylistic variant of a language appropriate to a particular social setting…" (Fromkin and Rodman, 1998:535). A clinical encounter is both a mixture of high formal and low informal registers. However, one would avoid very informal language (e.g., swearing, certain forms of slang), and very formal language (e.g. Latin words or other words in particular to medical professionals). It is very likely that one would use a lower register to talk to a patient, however, a well-informed patient may prefer to use a higher register.

Facework
Has to do with the social image a person presents and/or is thought to present to another during conversation. It can only be determined by a fellow speaker. In reality, a person is constantly working to project a positive face as well as helping others to maintain their own face, hence the term "facework" by definition. "The term face may be defined as the positive social value a person effectively claims for himself…during a particular contact [with others] (Goffman, 1955:213). It is something that is emotionally invested and that can be lost, maintained or enhanced, and must be constantly attended to in interaction. In general, people cooperate (and assume each other's cooperation) in interaction, such cooperation being based on the mutual vulnerability of face (Brown & Levinson, 1987). One can also be said to "save face" for themselves or their interactants, and it is an important part of maintaining positive social relations (i.e., to avoid conflict). Saving face is often achieved through politeness maxims (See below)

Politeness
You are probably familiar with the notion of being polite. But do you know there are two types of politeness each with a set of subsequent strategies in the English language?

Educational intervention and participants
Approved by the King's College London Ethics Committee (BDM/13/14-64), we carried out a 1.5-hour face-to-face workshop and a one-hour self-directed e-learning for 17 medical students in clinical years (years 3-5) and six communication trainers. The workshop consisted of an introduction to the CCT and group exercise to analyze videoed consultations using the CCT. The e-learning contains an interactive module for learning the CCT, a collection of interactive exercises, a discussion platform and a selfevaluation tool.

Evaluation and data analysis
To understand the changes of participants' analytical skills, we asked them to analyze how well the clinician has structured the consultation, built and maintained rapport, established understanding, demonstrated empathy and patient-centered approach in a video-recorded GP consultation. The same test was done before and after the intervention. We then compared and contrasted participants' pre-and postintervention responses. Two authors (SL & CA) evaluated the quality of the participants' pre-and postintervention responses. Using qualitative data analysis software (NVivo 10), they categorized these responses into five categories, namely, excellent, good, erroneous, ambiguous and superficial (see Table  2). "There is a clear structure to the consultation, yet the clinician is clearly attentive and responsive to the patient's needs, communication style, and emotional disposition."

Erroneous
Analyses that interpreted clinical and communicative outcomes inaccurately. There was reference to language use in interaction, but it was either incorrect or insufficient.
"Very poor. The clinician focusses far too much on her own contributions to the conversation rather than trying to build more systematically on what the patient says." Our comparative analysis identified 155 and 168 relevant responses pre-and post-intervention. Because of the small number of items in some categories (see Table 3), the quantitative researcher (MN) merged the five response categories into three broader categories namely, correct (originally Excellent, Good), incorrect (originally erroneous,) and neutral (originally term related, ambiguous and superficial), in preparation for statistical analysis. MN then compared the frequency of responses of each participant for every CCT analytical term they used in their responses that have already been categorized in the three evaluation categories. We added up the percentage of correct, incorrect and neutral responses for each participant in Preintervention and Post-intervention responses (see Table 3). Wilcoxon matched-pairs signed-ranks test was used to assess if there were statistically significant differences in the frequency of correct, incorrect and neutral responses in Pre-, and Post-intervention responses.
To assess students and tutors' experiences of learning and using the CCT, we devised two questionnaires with open ended questions. Two authors (FS & BON) conducted thematic analysis of participants' responses. All authors involved in data analyses are clinical communication trainers themselves. Ambiguous Analyses referred to language use in interaction but the relationship between it and clinical and communicative outcomes was unclear.

Measureable improvement of analytic skills
"There was not much negotiation."

Superficial
Analyses referred to language use in interaction but such reference lacked sufficient explanation.
"Understands the correct cues, good turn-taking and coconstruction as well as face work and register led to very good rapport."

Results
We found that the overall reference to language use in participants' responses slightly increased in the post-intervention, regardless of their quality. The difference between correct responses in pre-, and postintervention reached borderline significance (p<0.044, 95% Confidence Interval). Although the incorrect responses decreased post-intervention, this difference was not significant (p=0.701). We found almost no change in the number of neutral responses (p=0.364). This may be attributed to the difficulties of learning and using the CCT, which were also reflected in participants' own words, as shown later in the paper. Table 3 Pre-and Post-intervention use of terms by the total participants (n=23), by evaluative categories, three main themes and the number of participants reporting each term The responses that were evaluated as good increased by 13% in the post-intervention evaluation; and ambiguous analyses decreased by 14% (Table 4). Overall, there were about 8% responses deemed as incorrect, and the results did not change after the intervention.

Participants' perspectives of the CCT:
Insights into participants' experiences of using the CCT were elicited via the free text comments section of 15 student questionnaires. We asked the following four themed questions: How do you think this tool can inform your clinical practice?
Responses to Q. 1 revealed that students' 'usual' methods of analyzing communication centered on observing for specific communication skills and processes (n = 5) or use of the 'history-taking' structure (n = 5) taught as part of the undergraduate communication curriculum. The skills and processes referred to included rapport, listening, expressions of empathy, and observations of 'body language'. Five students appeared to have limited or no systematic approach, instead focusing on whether specific information was elicited (without reference to the process of how this was achieved) or in one example, on 'gut-instinct'. In response to Q. 2, 11 students confirmed that they thought using the analytical tool would change the way they analyze communication in future, four were unsure and one student did not think they would use it. Of the 11 positive responses, students referred to having a heightened awareness of specific interactional features (e.g. turn-taking, framing, co-construction) and of adopting the conversation analytic terms to articulate their analysis. The students, who were unsure, cited the complexity of the tool and the time needed to become familiar with it as the main challenges to its use. Overall, the majority of students commented on the usefulness of the tool in providing a more structured and systematic approach to analysis. All 15 students identified how they could use the tool for future learning and development (Q. 3). Key themes included using the tool to interpret and improve upon their own skills (n = 5) and to develop the habit of using the tool in clinical practice (n= 5). Other comments referred to using the tool to become more aware of the intricacies of communication such as positioning, pauses, cues (n = 3) and using the language of the tool to provide feedback to others. One student expressed they would use the tool for OSCE preparation by recording and analyzing practice OSCE stations. 13 students provided clear examples of how they thought the tool could inform their clinical practice (Q. 4). These spanned a range of areas such as responding flexibly during interactions, awareness of sociocultural factors and being able to offer structured feedback to peers. Two students were unsure of how the tool could inform future practice other than referring to 'turn-taking' and staying 'focused'. Five students commented on having difficulty learning and applying the tool due to the level of detail and content it contained. Sample responses can be found in Figure 1. Additional written feedback was gained from five clinical communication tutors, who participated in the training. Two tutors had a disciplinary background in linguistics, one in psychology and two in nursing. When asked if the tool would change the way they analyze communication, responses included that it would provide additional points for consideration and language to aid analysis, as a way to systematize feedback and deepen learning. Tutors' responses to the question of how the tool might be used in future teaching included asking students to revise the tool prior to experiential teaching, as an aid to postteaching reflection and for reference. Four of the five tutors stated that the tool could be used as an additional means, rather than as an alternative to their 'usual' method. Figure 1: Samples of student and tutor feedback Li S, Said F, O'Neill B, Ancarno C, Niksic M  2.

3.
We developed the Clin-Com Tool (CCT) to assist more systematic analyses of interactions in clinical contexts from a conversational perspective. After teaching 17 students and six communication tutors to use the CCT, we found that their ability to analyze communication in medical consultations improved. Participants felt that even with little or no linguistic background, they were able to use the CCT to analyze recorded consultations more effectively. Most suggested that they felt confident to use the tool to also analyze live interactions and improve their clinical practice or teaching. This, however, was not tested in this study although we believe that with increased familiarity with the tool, such use may be possible. Video reflection has been used by many UK medical schools in addition to the conventional method of role-play in clinical communication training. We think that this method should be adopted by more medical schools in that scrutinizing the nitty-gritty of the turn-by-turn conversation provides a 'slow motion' view of communication, which shows us what's normally overlooked and yet plays a significant role in social interaction (Rampton 2001: 97).
We recognize that our evaluation study has several limitations. First, it draws on a small number of participants and an even smaller number of tutors. To mitigate this limitation, we adopted two evaluation methods to analyze data, which provided a more reliable understanding of whether and how participants' analytical abilities have changed as a result of using the Tool. The data were analyzed both qualitatively and quantitatively to allow for a comprehensive understanding of the changes incurred by the introduction of CCT. We did not distinguish the differences among student year groups or between students and facilitators. Future studies with more participants could explore the differences, taking into account participants' different experiences in learning and teaching. The only international student suggested that the Tool could be particularly useful for them. This deserves further investigation.
Reflection and analysis are at the center of the epistemology of professional practice (General Medical Council 2009, Schon 1983, Schon 1987. Analysing language use in interaction is quintessential in clinical communication learning. It can help learners develop a critical ability to judge and act upon the constantly changing interactional communicative situations they find themselves. The CCT was developed to support the development of a more systematic analysis and reflection so that surface learning becomes deep learning, allowing effective knowing-in-action (Kaufman andMann 2014, Moon 1999). Focusing on the features of language use in interaction, the CCT puts forward the view of clinical communication as a creative and dynamic interactive process interaction, allowing students to sustainably and continuously develop their creativity and critical ability to manage medical consultations.
Analysis and reflection are important skills for undergraduate medical students to learn clinical communication.
Interactional linguistics can provide a systematic framework for developing such skills.
Improving trainees' understanding of interactional mechanism of communication improves patientcentred approach to consultations.

Discussion and conclusion
Take Home Messages