The construction of power in family medicine bedside teaching: a video observation study

Authors


Charlotte E Rees, Centre for Medical Education, University of Dundee, Tay Park House, 484 Perth Road, Dundee DD2 1LR, UK. Tel: 00 44 1382 381952; E-mail: c.rees@dundee.ac.uk

Abstract

Introduction  Bedside teaching is essential for helping students develop skills, reasoning and professionalism, and involves the learning triad of student, patient and clinical teacher. Although current rhetoric espouses the sharing of power, the medical workplace is imbued with power asymmetries. Power is context-specific and although previous research has explored some elements of the enactment and resistance of power within bedside teaching, this exploration has been conducted within hospital rather than general practice settings. Furthermore, previous research has employed audio-recorded rather than video-recorded observation and has therefore focused on language and para-language at the expense of non-verbal communication and human–material interaction.

Methods  A qualitative design was adopted employing video- and audio-recorded observations of seven bedside teaching encounters (BTEs), followed by short individual interviews with students, patients and clinical teachers. Thematic and discourse analyses of BTEs were conducted.

Results  Power is constructed by students, patients and clinical teachers throughout different BTE activities through the use of linguistic, para-linguistic and non-verbal communication. In terms of language, participants construct power through the use of questions, orders, advice, pronouns and medical/health belief talk. With reference to para-language, participants construct power through the use of interruption and laughter. In terms of non-verbal communication, participants construct power through physical positioning and the possession or control of medical materials such as the stethoscope.

Conclusions  Using this paper as a trigger for discussion, we encourage students and clinical teachers to reflect critically on how their verbal and non-verbal communication constructs power in bedside teaching. Students and clinical teachers need to develop their awareness of what power is, how it can be constructed and shared, and what it means for the student–patient–doctor relationship within bedside teaching.

Introduction

Bedside teaching involves clinical teaching in the presence of the patient and thus involves the triadic relationship among clinical teacher, student and patient.1,2 It is thought to be essential for helping students develop their understanding of the doctor–patient relationship, particularly of patient-centredness,3 and to facilitate the development of students’ clinical reasoning, clinical and communication skills and professionalism.2

These two quotations come from two different research articles:

‘…we are the master, they are the apprentice…’4

‘…doctors being up there and we being down here…’5

The first is from a skilled bedside teacher participating in a focus group study exploring perceptions of barriers to bedside teaching.4 The second is from a patient participating in a focus group study exploring multiple stakeholders’ views about patient involvement in medical education.5 Both of these quotes reflect the institutional status of doctors, students and patients within the medical workplace6,7 and imply straightforward power asymmetries whereby doctors ‘possess’ the superordinate position and thus power, and conversely students and patients possess subordinate and thus power-less positions.5

However, we know from previous audio-recorded observational research of bedside teaching within hospital settings that power is not this straightforward: individuals can change their local status (i.e. the position they negotiate within a particular interaction) through various strategies typically associated with those with higher institutional status.1,6,8,9 Drawing on multiple yet aligned theoretical perspectives, this study explores how power is constructed (enacted and resisted) throughout family medicine bedside teaching with the ultimate aim of raising the awareness of clinical teachers and students about the construction and sharing of power.

Theoretical perspectives

This study draws on several theoretical perspectives at different levels: (i) grand theory about the nature of knowing, and (ii) both macro- and micro-level theoretical perspectives on social processes such as power.10,11 Our theoretical perspectives at these different levels are conceptually aligned and therefore bring rigour to our qualitative approach.11,12

Grand theory

This study is underpinned by a social constructionist epistemological approach, in which knowledge is viewed as something that is constructed through social interaction.13

Macro- and micro-level theoretical perspectives

This study draws on symbolic interactionism (SI), which focuses on how individuals construct meaning, identity and order through social interaction.14 Although some researchers conceptualise SI as a mid-range theory focusing on local systems and including recognition of cultural and contextual variations,10 others construct SI as a ‘microsociology’ that focuses on micro (i.e. small-scale) social processes.11 Indeed, our approach to SI within the current study is commensurate with this micro-level theory focusing on the individual and interaction, and taking account of local context.10

In this study, we focus specifically on how individuals construct power through social interaction within family medicine bedside teaching. Power is a complex concept and a range of theoretical understandings exist.11 Although power has been defined as ‘the capacity to get others to think, feel or act the way we want them to, even if they don’t want to think, feel or act that way’,14 there is no widely accepted definition or theory of power and some scholars even eschew the act of defining and therefore reifying the concept.11

From a sociological perspective, power has typically been examined at the macro level with reference to large-scale social processes such as structural inequalities, the reproduction of patriarchy, and so on.11 Commensurate with this macro level is what Foucault calls ‘sovereign power’, which applies to people in positions of authority who ‘hold’ power and wield it over others in subordinate positions.15 However, power has also been examined at the micro level through an SI lens with a focus on power relationships and their enactment in real-life settings.11 At this micro level, power is commensurate with Foucault’s conception of ‘capillary power’, in which power is viewed as something that is constantly performed (enacted and resisted) within everyday social interaction and is therefore productive, bringing about meaningful ends rather than oppression necessarily.7,15

In the current study, we take this latter conceptualisation of power, aligned with our SI theoretical perspective. Indeed, our study illustrates that although it may be difficult to change one’s position within the medical institution (i.e. institutional status),6,7 it is possible to negotiate one’s position within particular interactions (i.e. local status) by employing linguistic strategies (such as asking questions) typically associated with those with higher institutional status.6,7

Finally, and in alignment with our SI theoretical approach, we draw on Goffman’s dramaturgy theory, which employs a performance metaphor to illustrate how we construct and co-construct our roles (e.g. actor, director, audience, prop, etc.) during social interaction and how we manage others’ impressions of us through face-work.16–18 Face-work concerns the maintenance of our own and others’ positive face (i.e. positive self-image) and negative face (i.e. maintaining autonomy through freedom from imposition and freedom of choice).16–18

Bedside teaching and power

Although numerous studies have explored the perceptions of clinical teachers, students and patients about bedside teaching through qualitative interviews and questionnaire surveys,2,4,19–23 only two recent studies have employed observational methods to explore interaction within bedside teaching.1,8,9,24

The first employed audio-recorded observations of 25 hospital bedside teaching encounters (BTEs) of 12 clinical teachers identified as ‘excellent teachers of humanistic care’ to explore how humanistic behaviour is role-modelled to residents.24 The authors found that bedside teachers taught humanism and professionalism in the doctor–patient relationship mostly by role-modelling rather than overt instruction, and that the role-modelling of humanistic behaviours was demonstrated through non-verbal communication, overt demonstrations of respect, the building of personal connections, the eliciting and addressing of patients’ affective responses to illness, and clinical teacher self-awareness.

The second study employed audio-recorded observations of six hospital BTEs to explore the construction of roles, identities and power within the triadic doctor–patient–student interaction.1,8,9 The authors explored some elements of power that were constructed within hospital BTEs by clinical teachers, students and patients through linguistic and para-linguistic strategies. For example, power was enacted through the construction of roles such as the patient as a non-person (someone who is talked about as if he or she is not there).17 We see this in the following comment made by a clinical teacher who refers to the patient’s ribs using the definite article ‘the’ rather than the embodied pronoun ‘his’ within the patient’s presence: ‘…see the ribs pulling in with inspiration.’8

Another example of power construction refers to the use of wh- questions (who, what, why, etc.) and yes/no questions. We see this in the following comment made by a clinical teacher, who asks a student a question about the patient’s skin colour: ‘Well, do you think it’s yellow?’8 Another example of power construction refers to the participant’s use of pronouns such as I, we, you and they. We see an interesting pronominal shift from I to we made by a medical student to a patient in the following comment, which serves to soften the student’s directive to the patient to remove his clothes: ‘Is it alright if I-um-we unbutton your pyjama top now?’1

A final example of power construction refers to the use of laughter. The following comment from a clinical teacher demonstrates this as he teases a female medical student by withholding information from her: ‘He he he he he he he okay I’ll tell you what all might be revealed.’9 These findings are consistent with reports in the broader social science literature demonstrating that power can be enacted and resisted through questions, directives, pronominal talk and knowledge.25–28

Although previous research has explored some elements of the construction of power within hospital-based encounters, we know that power is context-specific and so power relationships and enactments may differ in bedside teaching within other health care contexts, such as general practice. Indeed, primary and secondary care are thought to have different organisational cultures (i.e. assumptions, values, ideology and behaviour).29 Willcocks talks about the differences between primary and secondary care with relation to six factors: the historical background of the specialty; the nature of work tasks in the specialty and use of technology; relationships with external and internal environments; individualism and motivation; amount of communication within the specialty, and values and socialisation.29 To our knowledge, no studies have explored power fully within the organisational context of family medicine.

Finally, although many seminal anthropological and sociological studies conducted from the 1950s to the 1990s30–34 employed observational methods to help elucidate the cultures of medicine and medical education,35 none of these observational studies employed innovative visual methodologies.36 Indeed, even the two recent studies of hospital-based bedside teaching discussed above employed audio-recorded observations of BTEs.1,8,9,24 They therefore focused solely on linguistic and para-linguistic elements of power construction. The current study is not only original in the sense that it explores the construction of power within a novel context (family medicine) different to that of hospital-based medicine, but it is also original in that it employs novel methods within medical education research (i.e. video observation), thereby enabling a fuller analysis of the interplay among the construction of power and non-verbal communication and human–material interaction, in addition to sophisticated discourse analyses of language and para-language.

Methods

Study aims and research question

This study examines the triadic relationship among medical students, patients and clinical teachers during family medicine BTEs. The overarching aim is to explore the construction of roles, identities and power during BTEs. In this paper, we answer the research question: how is power constructed within family medicine bedside teaching?

Participants

Having obtained ethics approval, we invited Year 3 medical students about to commence family medicine rotations at two different medical schools (one undergraduate, one graduate-entry) to participate in this study. We then invited family doctors assigned to consenting students to participate. The researchers (RA or ADL [see Acknowledgements]) arranged to attend the two general practices on specific dates. Patients attending on the scheduled day were notified about the research study several days prior to their scheduled appointment with the participating family doctor. They were then recruited to participate in the study by the researchers prior to their appointments on the day of data collection. All students, clinical teachers and patients were required to read an information sheet and sign a consent form before participating. Participants were able to withdraw from the study at any time without penalty. A total of four doctors, two students and eight patients participated across seven BTEs. Table 1 shows the participants’ demographic characteristics.

Table 1.    Demographic characteristics of participants in bedside teaching encounters (BTEs)
ParticipantsBTEGenderAge, yearsEthnicityFirst language
  1. MD = male doctor; FD = female doctor; FS = female student; FP = female patient

Doctors
 MD11–3Male44Non-WhiteFarsi
 FD14–6Female51WhiteEnglish
 FD27Female32Non-WhiteTamil
 FD37Female53WhiteEnglish
Students
 FS11–3Female19Non-WhiteGujarati
 FS24–7Female25WhiteEnglish
Patients
 FP11Female59WhiteEnglish
 FP22Female60WhiteSpanish
 FP33Female36Non-WhiteEnglish
 FP44Female4WhiteEnglish
 FP54Female37WhiteEnglish
 FP65Female37Non-WhiteJapanese
 FP76Female35Non-WhiteFarsi
 FP87Female84WhiteEnglish

Data collection

Five BTEs were video-recorded (BTEs 1–3, 5 and 7) and two were audio-recorded (BTEs 4 and 6). The BTEs ranged in length from 16 minutes to 80 minutes. If BTE 7 is excluded as an outlier, the average length of a BTE is 19 minutes; the total for the entire sample is 193 minutes (Table 2). Individual interviews were conducted wherever possible with student, patient and clinical teacher participants after each BTE. In the interviews, the researchers asked participants for their thoughts about the encounter, such as on the roles they had played, what they had learned, and the extent to which they had felt empowered. We employed the interview data to aid our interpretation of our primary data (the BTE observations), and thus interview data are not presented here.

Table 2.    Length of bedside teaching encounters (BTEs) and conversational turns by type of BTE participant
BTELength, minTotal turnsDoctor turnsPatient turnsStudent turns
  1. * The two numbers refer to the numbers of turns of both doctors or both patients within this BTE

11629214012329
21830715414310
32636318515127
41927012497, 2*47
51734717015522
617219699852
7801521153, 183*571614
Total (excl. BTE 7)193 (113)3319 (1798)1178 (842)1340 (769)801 (187)

Data analysis

The video- and audio-recordings were transcribed without using participants’ names. Transcripts included what was said and how it was said; para-linguistic features of talk such as pauses, hesitations, laughter and interruption were also transcribed. We employed both thematic framework and discourse analyses.37,38 The transcripts, videos and audios of BTEs 1–3 were analysed independently by all three authors, who then met to compare, contrast and negotiate their preliminary themes. This conversation resulted in a coding framework comprising 20 higher-order themes (Table 3), which was then used by the first and second authors to code all seven BTEs using atlas.ti Version 6.2 (Scientific Software Development GmbH, Berlin, Germany). The coding was precise in order to allow us to establish the frequency of certain codes for which frequency would be meaningful (e.g. the numbers of questions asked by doctors, patients and students across the seven BTEs).

Table 3.    Higher-order themes identified by the thematic and discourse analysis
ThemeExample
  1. MD = male doctor; FD = female doctor; FS = female student; FP = female patient

Activity typeFarewell
e.g. FD1: ‘So we’re all done’ (BTE 5)
Interaction typeTriadic interaction
e.g. MD1: ‘Eh [FP1] that’s the (.) third-year medical student’
FP1: [Hi]
FS1: [Hi] (BTE 2)
Physical positioningPatient as up
e.g. MD1: ‘Come have a seat (.) please’ (0.5) (BTE 2)
Control of medical artefactsDoctor using sphygmomanometer
e.g. ‘We haven’t checked your blood pressure and (2.0) I’ll do that now (8.5)’ (BTE 7)
Construction of powerDoctor giving patient directive
e.g. MD1: ‘Take that off’ (BTE 1)
Negotiation of identityDoctor constructing medical identity as knowledgeable and experienced
e.g. FD1: ‘Look we’ve seen lots of babies (.) like this (.) lots of them’ (BTE 4)
Roles playedPatient constructed as prop by doctor
e.g. FD1: ‘Why don’t you ask her some questions about this and (.) see what your assessment is (.) okay?’ (BTE 6)
Asking questionsDoctor asking patient question
e.g. FD1: ‘And that was all okay?’ (BTE 4)
Responding to questionsPatient responding to doctor question
e.g. FP1: ‘Yeah (.) it was referred through (.) (um) preschool’ (BTE 4)
Use of laughterPatient laughing as she constructs her identity as lover of chilli
e.g. FP2: ‘I mean I’m trying to avoid all the (.) spicy food (.) I’m not um- (.) because I LOVE hot chilli huh huh huh’ (BTE 2)
Use of humourPatient jokes when student takes her pulse
e.g. FP8: ‘Am I alive?’ (BTE 7)
Use of pronounsDoctor uses exclusive ‘we’ to soften her directive
e.g. FD1: ‘Well why don’t we get him right out and have a look?’ (BTE 5)
InterruptionDoctor interrupts and over-talks patient
e.g. FP1: ‘I didn’t think [I snored] ((says laughingly))
FD1: [she (.) she snored] ((says laughingly)) (BTE 4)
DirectivesDoctor gives directive to patient
e.g. MD1: ‘What I want you to do is stand’ (BTE 3)
Giving adviceDoctor gives patient parenting advice
e.g. MD1: ‘My- my advice maybe (.) mum take care of her and go’ (BTE 3)
Medical talkPatient uses medical talk
e.g. FP1: ‘I’m on Fosamax, Warfarin (.) D- Digoxin (.) um (.) Oroxin (.) and ah (.) Xalatan eye drops’ (BTE 7)
Health belief talkPatient communicates her health beliefs
e.g. FP1: ‘I was getting dizzy (.) plus I could taste the blood in the- I can taste it when it’s [blood pressure] up’ (BTE 1)
ChallengeDoctor challenges patient
e.g. MD1: ‘This could be nothing to do with that’ (BTE 2)
ApologyDoctor apologises to patient
e.g. MD1: ‘Sorry about that’ (BTE 3)
MisunderstandingDoctor misunderstands trainee
e.g. FD2: ‘She’s on (Fosamax)’
FD1: ‘Oh she’s already on it?’ (BTE 7)

Results

Several different types of activity occurred within the BTEs, including welcomes, listening to and exploring patient histories, examining patients, explaining information to patients, mini-teaching sessions, case presentations and farewells. Across these activity types, students, patients and clinical tutors constructed power through multiple linguistic, para-linguistic and non-verbal communication strategies. In the paragraphs that follow, we quote short excerpts from the seven BTEs to illustrate the various strategies involved in the co-construction of power. In the final section, we include an extended power-referent excerpt to illustrate the complex interplay among multiple strategies and speakers. Note that we present another two extended excerpts and their interpretations in Table S1 (online).

Language

Students, patients and clinical teachers employed five key linguistic strategies to construct power within family medicine bedside teaching. Firstly, all participant groups employed questions to enact or resist power, although clinical teachers (MD [male doctor] 1: ‘Now what can I do for you today?’) employed disproportionately more questions according to their number of turns (233 questions, 20% of turns [Table 2]). Students (FS [female student] 2: ‘Do you need a new referral?) and patients (FP [female patient] 3: ‘Should I not try the Infacol drops?’), by contrast, asked disproportionately fewer questions according to their number of turns (students, 64 questions [6% of turns]; patients, 46 questions [5% of turns]).

Secondly, all participant groups employed directives to enact or resist power. As with questions, clinical teachers (MD1: ‘Look at that now’) gave disproportionately larger amounts of directives according to their number of turns (99 directives, 8% of turns). Patients (FP1: ‘You’d better check my blood pressure’) and students (FS2: ‘Poke your tongue out’) gave disproportionately fewer directives according to their number of turns (students, 14 directives [2% of turns]; patients, 15 directives [1% of turns]).

Thirdly, although all participant groups gave information, only clinical teachers gave advice, which could be seen as a softer, more polite way of giving a directive (FD [female doctor] 1: ‘I’d suggest to you that you might try only bathing him every second day’).

Fourthly, all participants employed personal pronouns in their talk and some of these were strategically related to the construction of power. For example, participants employed the first-person singular pronouns I and me when they wanted to construct their authority, identity, independence and knowledge (FP1: ‘Yeah but to me it feels low’). They employed the first-person plural forms we and us inclusively (meaning me and you together: ‘FD1: So, how are we going to figure this out?’) or exclusively (meaning I or you: ‘MD1: ‘Let’s get your blood pressure’) in such a way that the former constructs collegiality and the latter masks a power asymmetry (we is employed to soften directives). Participants also employed third-person pronouns (FS2: ‘She’s got large tonsils’) to refer to another of their BTE interlocutors; this talking about someone (typically the patient or student) as if she were not present served to exclude, disempower and ultimately construct that other as a non-person.

Finally, all participant groups employed medical talk to demonstrate their knowledge, thereby constructing power within bedside teaching. Although the amount of medical talk employed was higher for doctors (FD2: ‘I thought he said it was BCC’) than for students (FS1: ‘…pyloric stenosis?’) and patients (FP1: ‘They put me on MicardisPlus plus Adalat’) at 32, 19 and 19 instances, respectively, the proportions of medical talk according to the number of turns was not terribly dissimilar across the three groups (clinical teachers, 3%; students, 2%; patients, 1%). Mirroring medical talk was health belief talk, which refers to talk in which patients drew on their experiential knowledge of their own bodies and health conditions to construct power (FP3: ‘I thought she was sick or something’). However, such health beliefs were often contested by clinical teachers, which served to allow the latter to regain power within the encounter (MD1: ‘I think changing the milk will not make a difference’).

Para-language

Participants in BTEs employed two key para-linguistic strategies to construct power within family medicine bedside teaching: laughter and interruption. All participant groups employed laughter to construct power across the seven BTEs. For example, we see patients employing laughter to soften directives (FP1: ‘And you’d better check my blood pressure uh ha ha ha huh’); we see clinical teachers employing laughter to contest a patient’s health beliefs (FD1: ‘With that being said, some of the medicine will be coming through your breast milk and in some ways that can be an okay thing ((says laughingly))’), and we see students using laughter to soften face-threatening acts (FS2: ‘Have I made you nervous? ... uh-huh-huh ((closed mouth laughter))’). In terms of interruption, we see all participant groups interrupting one another and over-talking in order to involve themselves in the encounter and wrestle back some conversational power(Fig. 1).

Figure 1.

‘Can I just have a feel of your pulse while she does that?’ Extract from bedside teaching encounter 7. FD = female doctor; FS = female student; FP = female patient. Transcription notations: >words in sharp brackets< = spoken faster than typical speech for the individual; (.) = micropause; (2.0) = pause for two seconds; ((words in double brackets)) = further information; [words in square brackets] = overlapping speech

Non-verbal communication

The final type of strategy identified by our analysis was non-verbal communication; BTE participants employed two key non-verbal strategies to construct power. Firstly, we see doctors and one of the students (FS2) controlling medical equipment (e.g. sphygmomanometer, stethoscope, otoscope, ophthalmoscope, etc.) and other medical artefacts (e.g. diagrams, computers, computer print-outs, patient notes, etc.) across the BTEs, in a manner that serves to construct their medical knowledge, identities and therefore power within the encounters. We also see how power can be constructed through the physical positioning of participants within the room. For example, we often see FS1 disempowered by her physical positioning within BTEs 1–3: she typically sits against the wall at the periphery of the patient–doctor interaction, and is told where to sit by the male doctor in BTE 2 (MD1: ‘[FS1], you can sit there’).

‘Can I just have a feel of your pulse while she does that?’

We can see the interplay among linguistic, para-linguistic and non-verbal strategies to construct power in an excerpt from BTE 7 (Fig. 1). (See Table S1 for another two extended extracts showing the construction of power and associated interpretations.) In BTE 7, a female patient is attending the practice for her 75-year-plus health check, which the student mostly completes, unsupervised. However, the patient then sees a female general practice (GP) trainee (FD2) and the student together in order to complete some further tests and examinations, before the GP trainee and student report back to the clinical teacher (FD3). This excerpt begins with the GP trainee discussing the patient’s medication with the patient, while the student is out of the room collecting the results of the health check. The student enters the room as the GP trainee and patient are discussing her Fosamax prescription.

By contrast with the excerpts in Table S1 and the conversational turns outlined in Table 2, we see a fairly unusual combination of BTE roles in this encounter: the GP trainee, student and patient are all actors within this BTE, and the student is momentarily co-constructed as director through the GP trainee’s question: ‘What else did you want?’ (turn 11) and her own response: ‘We need to do height weight’ (turn 12). We think this atypical construction probably relates to the fact that the GP trainee and student are both ‘learners’, albeit at different stages of the medical education continuum.

In this excerpt, we see the GP trainee and student simultaneously examining the patient. Although it is not technically correct to take blood pressure and pulse measurements simultaneously because it may result in higher than normal readings for both, this is an example of the sharing of power. We first see the GP trainee put the sphygmomanometer cuff on the patient’s arm (turn 1) and later we see her pump up the cuff and use her stethoscope (turn 15), which serves to construct her medical authority and power within the encounter. While she does this, the patient and medical student chat about the benefits of the ‘old-style’ mercury sphygmomanometer compared with its digital counterpart (turns 15–18) and then the student quickly asks the patient a question: ‘Can I just have a feel of your pulse while she does that?’ (turn 18). By asking this question and then checking the patient’s pulse, the student constructs her medical identity and authority, and shares power with the GP trainee. From a positional perspective, we also see the GP trainee, patient and student in close physical proximity, thereby serving to construct all participants as active participants and ‘insiders’ within the encounter.

Alongside these non-verbal strategies, we also see the use of linguistic and para-linguistic strategies to construct power. We see the GP trainee employing the pronoun I rather than we to construct her agency, authority and power in terms of taking the patient’s blood pressure (turn 1); we see her repeated questions of the patient (turns 1, 3, 5) and student (turn 11), and we see her contest the patient’s health beliefs about Fosamax in turn 7 (‘I’m sure it’s doing you some good’). We also see the GP trainee refer to the patient as ‘her’ to the student (turn 11), which serves to disempower and construct the patient as a non-person.

We also see the student constructing power through her questions of the GP trainee (turn 10) and patient (turns 18, 23) and through her medical talk about digital sphygmomanometers to the patient (turn 18). Finally, we see the patient construct power throughout this excerpt. Although she directs the GP trainee to give her a Fosamax prescription in turn 2, she simultaneously complains that she doesn’t really want it, and questions the GP trainee on whether there is an alternative. Later, in turn 15, she interrupts the student–GP trainee conversation with her medical talk about sphygmomanometers. Both this interruption and her medical talk serve to construct her knowledge, authority and power within the encounter, and this powerful construction continues throughout the excerpt as she makes yet more interruptions of the student (turns 17, 19), and GP trainee (turn 22), and yet more medical talk (turn 22: ‘Oh that’s [blood pressure] a bit high today’). The student and patient share a mutual tease at the end of this excerpt, which simultaneously constructs their own power. The student first teases the patient (‘Have I made you nervous?’), to which the patient retorts with a fallibility tease: ‘Probably’ (turn 24). This could be considered a face-threatening act and so, after a brief pause, the patient softens this with a quick retraction (‘Well, not really’), while the student simultaneously laughs, which serves to construct her identity as good humoured (turn 25).

Discussion

Patients were found to have more conversational input into the family medicine BTEs in this current study compared with that found in previous research in a hospital setting,8 which possibly reflects the different cultural contexts of primary and secondary care.29 However, we still see a fairly typical construction of roles across the family medicine BTEs, in which the clinical teacher is constructed as director, the student as audience or actor, and the patient as prop or non-person.8 Clinical teachers, students and patients employed both linguistic (e.g. pronominal talk) and para-linguistic (e.g. laughter) strategies to enact and resist power across multiple activity types in the BTEs, as has been found in hospital BTEs.1,8,9 However, by using video-recorded observation in the current study, we also found that clinical teachers, patients and students employ non-verbal communication strategies (e.g. use of medical artefacts and physical positioning) to construct power.

Allied with their higher institutional status, clinical teachers enact power by asking more questions, giving more directives and advice, employing more medical talk, and using and controlling more medical artefacts across the BTEs. Consistent with their lower institutional status within the BTEs, we see numerous examples of the construction of patients and students as non-persons (i.e. they are talked about in the third person [e.g. ‘she’] as if they were not present), as has been found in previous research.8 Although students and patients have lower institutional status within the medical workplace, and typically employ these strategies less frequently (e.g. they ask fewer questions and make fewer directives, etc.), they still do employ these strategies, thereby negotiating their local status and thus power within the BTE. Using linguistic strategies such as questions or directives or medical talk to someone with higher institutional status can threaten the positive and negative face of those with superordinate positions; it was not uncommon for these face-threatening acts to be accompanied by laughter, as has been found previously in hospital BTEs.9 Ultimately, our findings are commensurate with a Foucauldian understanding of power relations and enactments and extend the previous literature by illustrating the importance of non-verbal communication and human–material interaction in the construction of power and in the novel context of primary care.

Methodological issues

This study has a number of methodological issues which must be taken into consideration when interpreting the results. Firstly, although it is not unusual for in-depth analyses of interactional data to be based on small sample sizes,8 this study includes just seven BTEs within two different general practices associated with two different medical schools in Australia. Although our sample size is adequate for an in-depth qualitative study underpinned by an interpretive approach, we realise that our sample is fairly homogeneous in some ways. Although our sample of participants was fairly diverse in terms of age, ethnicity and first language, it was almost exclusively female, and this gender homogeneity may explain why we found no obvious construction of gendered identities within our BTEs, unlike in our previous work within hospital BTEs.9 Therefore, further observational work with greater diversity in the gender composition of BTEs is required. However, despite this fairly homogeneous sample, our findings largely support those of a previous study conducted within the context of hospital BTEs, and support the theoretical perspectives of Foucault and Goffman, suggesting that our study has strong conceptual ‘generalisability’.10

Secondly, in order to address the methodological challenges of previous audio-recorded observations of BTEs, we employed video-recording wherever possible and although the video enabled us to capture non-verbal communication involving gross body movements, it was sometimes hard to capture smaller and more subtle non-verbal communication (e.g. smiles) of all BTE participants all of the time. Furthermore, two of the seven BTEs were audio-recorded only (at the request of patients) and therefore it was not possible to code the transcripts of those two BTEs precisely for several of the more visually orientated themes (e.g. physical positioning). Further observational work employing video observation with multiple cameras is now needed to explore in greater detail non-verbal communication and human–material interaction.

Finally, we quantitised some of our qualitative data in this study in order to reveal patterns in the data.39,40 Although discourse analysts sometimes explore patterns using numbers,31 we know that some methodological purists find this problematic because different world views underpin qualitative and quantitative approaches. However, despite our use of numbers, we retain a process-orientated qualitative approach underpinned by social constructionism.13,41

Implications for educational practice and research

Although our study illustrates the typical power asymmetries between those with high and low institutional status within bedside teaching, it simultaneously demonstrates how participants with lower institutional status (i.e. students and patients) enact and resist power by employing the same strategies used by those with higher status (i.e. doctors). Given the current rhetoric on patient-centredness and shared leadership within, respectively, medicine and academic medicine,42,43 both of which espouse the sharing of power, it is essential that clinical teachers and students reflect critically on how their communication enacts and resists power in bedside teaching. Therefore, this paper (discussing a type of research typically published in social science rather than medical education journals) could act as trigger material for discussions in faculty development and student learning workshops about power within bedside teaching. Video-reflexivity methods may be useful here44 and may help clinical teachers and students to make visible their communicative practices and discuss how their use of language, para-language and non-verbal communication might serve to construct power. Specific questions to address in educational sessions could include: what roles are played by clinical teachers, students and patients and how are those roles constructed? What linguistic, para-linguistic and non-verbal communication strategies are employed to construct power asymmetry or shared power between doctor and patient, doctor and student, and student and patient? Fundamentally, both educators and students need to develop an understanding of what power is and how it can be constructed and shared within BTEs, and need to be given opportunities to discuss what asymmetrical and shared power means for the student–patient–doctor relationship within bedside teaching. This awareness raising may be influential not only in terms of developing clinical teachers’ and students’ understandings of power, but also in helping teachers and students develop their interpersonal communication, thereby simultaneously benefiting patient care and student learning within bedside teaching.

In terms of further research, we are currently conducting video-recorded observations of a larger number of BTEs across multiple health care settings (primary and secondary care) and a diverse range of specialties within the UK. As part of our steering group activities, we are also asking BTE participants (clinical teachers, students and patients) which issues within their BTEs they would like us to explore. Indeed, if we are serious about espousing a model of power sharing among BTE participants, we also need to consider the sharing of power between the researcher and the researched.36,45,46

Contributors:  all three authors contributed to the study design and data analysis. CER and RA applied for ethics approval at both study sites. RA contributed to data collection. CER wrote the first draft of the paper. All authors contributed to the editing of the paper and approved the final manuscript for publication.

Acknowledgments

Acknowledgements:  we thank Professor Ian Wilson, Graduate School of Medicine, University of Wollongong, Australia, and Professor Simon Willcock, Central Clinical School, University of Sydney, Australia, for helping in the recruitment of participants, and Dr Amy De Laroche, Beaumont Health System, Michigan, USA, for helping in the collection of data. We also thank the tutors, students and patients who participated in this study. Finally, we thank our family medicine colleague, Dr Madawa Chandratilake, Centre for Medical Education, University of Dundee, Scotland, for his helpful peer review comments on an earlier draft of this manuscript.

Funding:  this study was supported by a bridging support grant issued by the University of Sydney, Sydney, NSW, Australia.

Conflicts of interest:  none.

Ethical approval:  this study was approved by the ethics committees of the two participating schools, who are not named here to maintain anonymity.

Ancillary