Education and debate

Stories we hear and stories we tell: analysing talk in clinical practice

Glyn Elwyn, senior lecturer, ^a Richard Gwyn, research coordinator. ^b

^a University of Wales College of Medicine, Cardiff CF4 4XN, ^b Centre for Language and Communication Research, University of Wales, Cardiff CF1 3XB

Correspondence to: Dr Elwyn ElwynG{at}cf.ac.uk

Even at its scientific best, medicine is always a social act.¹

For all the science that underpins clinical practice, practitioners and patients make sense of the world by way of stories. ^{2 3} Even the most evidence crazed doctors have to translate their perception of "biostatistical truths" into accounts that make sense to others. Studies of the consultation process, which have largely taken place in primary care, have focused on the structure of the meeting from greeting to closure. ^{4 5} The concepts of doctor centredness or patient centredness are described⁶ and measured⁷; these concepts undoubtedly have a profound influence on professional practice.⁸ These observations have led to an ongoing exploration of the effect that communication styles have on both patient satisfaction and clinical outcome.⁹

But there is much more depth to be explored in the process of communication, and the tools normally used are insufficient to examine the layers of meaning that lie within the text of exchanges.¹⁰ The microanalysis of talk can inform the essence of medical practice, define principles for effective communication, attach meanings to a patient's story, as well as help doctors share ideas about fears and hopes for the futurein medical speak: communicate risks and benefits.¹¹ By deconstructing a piece of dialogue in this paper, we hope to illustrate the value of learning to listen carefully to the stories we hear.

	Discourse analysis

Discourse analysis is, essentially, the study of language in context. ^{12 13} There are many examples of analysis in which these techniques have identified valuable but previously hidden patterns and perspectives (for example, in outpatient clinics,¹⁴ among health visitors,¹⁵ and in transcripts of interviews conducted by HIV counsellors¹⁶). Discourse analysis has roots in linguistics, sociology, and psychology but despite these origins it is really no more than the examination of the processes of naturally occurring talk. For instance, how is one version of events selected over any other? How is a familiar reality described in such a way as to lend it an unquestionable authority? The one essential point about discourse analysis is that it follows the text which in many cases, like the following extract (box), is a piece of talk.

The transcript in the box is of a meeting between a patient and a doctor in an inner city practice. The patient is a woman aged 52. Because she has an urgent problem she has been unable to see her usual doctor and has to consult with someone she has not seen before. She begins with a torrent of symptoms: puffy eyes and legs, burning on urination, pain in the back, and a sore throat. The doctor examines the urine sample, diagnoses a "water infection" (she gets recurrent cystitis), and asks if the patient is allergic to any antibiotics. At this point the consultation might well have terminated. But the patient lets out a cough: nothing extravagant, just a little cough.

Let us consider the transcript and apply the techniques of discourse analysis at first hand. The aim is to reproduce the dialogue down to the last "um." The symbols used can seem a bit off putting at first. Interruptions, pauses, overlapping speech, and intonations are all signified to gain access to the precise dynamics of the interaction. The extract begins 2 minutes and 30 seconds into a consultation which lasted 6 minutes and 45 seconds in total. The extract itself lasts 2 minutes.

The cough (at line 052), functions as a discourse marker signalling the speaker's wish not to terminate the interaction.¹⁷ The doctor's next utterance ("Anything else?") is characteristic of doctors' preclosing moves in interactions with patients,¹⁸ but leaves closure to the patient. The patient is in a position to allow closure or to shift to a new topic. She opts to respond (055) first with a pause then a request for "water tablets." The pause here indicates a new topic and precludes any accusation of indecent haste. The patient does not wish to be perceived simply as itemising a shopping list. The ritual of correct timing is necessary to maintain the necessary gravity accorded to the ceremony of consultation and prescription. Although the pause lasts less than two seconds its significance should not be underestimated.

The patient makes her request for a repeat supply of Dyazide and links her diuretic treatment to her hormone replacement therapy, leaving the doctor seemingly bemused. She dismisses the water tablet topic while the doctor is still mulling it over(071) a prolonged "mmm"and she proceeds (072-3): "But I wanted the er Seroxat the antidepressant tablets please."

The use of the past tense ("I wanted") for a request in the present serves allows the speaker to remove herself from the here and now, a common feature of "negative politeness."¹⁹ This is consistent with a reluctance to be perceived as too pushy or demanding, and is consolidated by the "please" at the end of the utterance. The doctor might be reluctant to bluntly ask the patient about the source of her depression but at the same time the seemingly unrelated sequence of her taking diuretics, hormone replacement therapy, and her request for antidepressant drugs, needs some substantiation. The doctor asks "You take those do you?" (the slight but unexpected emphasis on "take" indicating the doctor's momentary confusion). The patient replies with a simple "yeah" (075). The doctor follows up with a question formulated out of professional concern and framed in linear time: "How long have you been taking those?"

There is a pause, and then the patient chooses to respond not in linear time, but in event time (077): "Well my son was killed." This is the event which led to her being prescribed antidepressants. These opening phrases are interspersed by lengthy pauses: "( . ) Uh well my son was killed (2.0) five years ago (2.0)."

Linear time (five years) is only relevant in relation to event time (her son's death). Mishler made the distinction between the "voice of medicine" and the "voice of the lifeworld."¹⁰ The following excerpt is drawn from a consultation between a general practitioner and a young woman who is abusing alcohol.¹⁰

Doctor: How long have you been drinking that heavily?

Patient: Since I've been married.

Doctor: How long is that?

Patient: (giggle) Four years. (giggle)

Dismissing the importance of a biomedical time frame for clinical judgements, Mishler argues that the practitioner above, by insisting on a real time scale (four years) over a more meaningful, personal one subordinates the voice of the "life world" to the voice of medicine.

In our excerpt, the doctor does not interrupt the patient; he allows the voice of the life world to take precedence ("life meaning" comes before "time meaning"). By doing so he gives the patient the opportunity to fill in the kinds of linear detail which she thinks might be relevant and which she immediately does anyway ("five years ago").

In the transcript the introduction of biographical detail helps establish the narrative basis of the patient's depression and legitimises her continued use of antidepressant drugs. The account, with its litany of deaths, provides an idea of this patient's "sustaining fiction,"²⁰ of the explanatory causes that underlie her story. We are all continuously involved in the process of adding new stories to our own sustaining fictions. Stories are renewed, reconstructed, or abandoned but are always central to the individual's presentation of self and sense of personal identity.

We find that the patient's son did not simply "die." He was "killed" (077), that is died as the victim of a particular agent or set of circumstances. Implicit in the pauses is an opportunity for the doctor to ask how her son was killed, an opportunity that he chooses not to take. The pauses act as a rhetorical device allowing the gravity of her loss to sink in and gives an accounting for the prescribed drugs. But that is not all. Seeing that the doctor does not request further information about the circumstances of her son's death (a request which would be highly threatening to both doctor and patient), the patient then enumerates two other losses in her family: the death of a baby granddaughter from meningitis and the death of a son in law from a heart complaint. The fact that the causes of death and the ages of the dead are enumerated in both these other cases only draws attention to the lack of explanation regarding the killing of her son.

By emphasising within a short space of time the extent of her losses, the patient avoids the possibility of being categorised as somebody requesting antidepressant drugs without good cause. Hanging over every patient is the potential accusation of malingering²¹ resulting in an obligation to prove that a malady is not contrived and to express a wish to get well. Moreover, in this transcript, the patient insists that it was her doctor who "wanted her" to take the tablets (reinforcing her own passivity in this decision despite their effectiveness). Then (as if further evidence of her good intentions were needed) she states her wish to reduce the dose, thus maintaining her contractual responsibilities to recovery. This wish to lower the dose is shown as her choice, a choice unaided (indeed hindered) by her practitioner ("but doctor Y said she still wanted me to take those antidepressants"), which strengthens the representation of herself as a responsible member of society; she states later in the consultation: "I wouldn't like the thought of being on them forever."

	Reflective thoughts

For the doctor, the narrative appeared out of the blue. He records: "I hadn't expected this: three deaths and a request to withdraw from antidepressants during a routine repeat prescription. Would that be all right? To participate in a shared decision about the end of grief, about a symbolic farewell to a son, killed five years ago. I attempted to give her autonomy over her decision, hoping not to abandon her.²² But it wasn't enough.²³ How could I tell her that I didn't know. That if I had lost a son I can't imagine surviving at all, never mind coming off tablets."

	Discussion

This transcript reveals intricate communication strategies, informs us how patients construct their roles within consultations, and opens up a new way of listening to the signals which so often pass unnoticed; this analysis gets us that step nearer to reconstructing "the imaginative universe in which human acts are signs."²⁴ Mishler objects to mere code-category assessments of consultations (that is, ascribing coding formats to subunits of the interaction) and argues for a more eclectic approach using detailed textual assessment. There is also a need to capture the thoughts of both patients and clinicians. As more studies show that patients' perceptions of what happens within consultations are probably more valid than measures based on coding structures,^25-27 and that "finding common ground" is more of a perceived event than a quantifiable finding, those who are interested in this sort of analysis need methods that will illuminate the subtextthe white space that signifies thoughts, disagreements, distress, and indecision. Evidence suggests that patient participation in decisions reduces costs for the health service and emphasises the critical, but almost neglected, part that the patient-doctor interaction plays in the use of health resources.²⁸

Clinicians may have to go beyond the superficial assessment of the consultation to examine the perceived messages that patients take away into the longitudinal discourse of their own lives.²⁹ By becoming interested in talk, clinicians might be able to listen more constructively to their patients' stories³⁰ and might be able to allow a more "democratic arrangement of voices."³¹ Lest we forget, for countless patients it is the telling of their stories that helps to make them well.

	Acknowledgments

Series editor: Trisha Greenhalgh

	References

1. Davidoff F. Who has seen a blood sugar? Philadelphia, PA: American College of Physicians , 1996.
2. Hunter KM. Doctors' stories: the narrative structure of medical knowledge. Princeton, NJ: Princeton University Press , 1991.
3. Brody H. Stories of sickness. New Haven, CT: Yale University Press , 1987.
4. Byrne PS, Long BEL. Doctors talking to patients. London: HMSO , 1976.
5. Pendleton D, Schofield T, Tate P, Havelock P. The consultation: an approach to learning and teaching. Oxford: Oxford University Press , 1984.
6. Levenstein JH. The patient-centred general practice consultation. S Afr Fam Pract 1984; 5: 276-282.
7. Stewart M, Brown JB, Weston WW, McWhinney IR, McWilliam CL, Freeman T. Patient centred medicine: transforming the clinical method. Thousand Oaks, CA: Sage , 1995.
8. Laine C, Davidoff F. Patient-centered medicine: a professional evolution. JAMA 1996; 275: 152-156[Abstract].
9. Kinnersley P. The patient-centredness of consultations and the relationship to outcomes in primary care. In: Bristol: University of Bristol , 1997[Unpublished MD thesis.]
10. Mishler E. The discourse of medicine: dialectics of medical interviews. Norwood, NJ: Ablex , 1984.
11. Calman KC, Royston GHD. Risk language and dialectics. BMJ 1997; 315: 939-942[Free Full Text].
12. Edwards D, Potter J. Discursive psychology. London: Sage , 1992.
13. Potter J, Wetherell M. Discourse and social psychology. London: Sage , 1987.
14. Wodak R. Disorders of discourse. London: Longman , 1996.
15. Drew P, Heritage J, eds. Analyzing talk at work. Cambridge: Cambridge University Press, 1992.
16. Silverman D. Discourses of counselling. London: Sage , 1997.
17. Coupland J, Robinson J, Coupland N. Frame negotiation in doctor-elderly patient consultations. Discourse and Society 1994; 5: 89-124.
18. Coulthard RM, Ashby MC. A linguistic description of doctor-patient interviews. In: Wadsworth M, Robinson D, eds. Studies in everyday medical life. London: Robinson , 1976.
19. Brown P, Levinson SC. Politeness: some universals in language usage. Cambridge: Cambridge University Press , 1978.
20. Hillman J. Healing fiction. Woodstock, CT: Spring Publications , 1983.
21. Parson T. The social system. Glencoe, IL: Free Press , 1951.
22. Quill TE, Cassel CK. Nonabandonment: a central obligation for physicians. Ann Intern Med 1995; 122: 368-374[Abstract/Free Full Text].
23. Quill TE, Brody H. Physician recommendations and patient autonomy: finding a balance between physician power and patient choice. Ann Intern Med 1996; 125: 763-769[Abstract/Free Full Text].
24. Geertz C. The interpretation of cultures. New York: Basic Books , 1973.
25. Tuckett D, Boulton M, Olson I, Williams A. Meetings between experts: an approach to sharing ideas in medical consultations. London: Tavistock Publications , 1985.
26. Margalith I, Shapiro A. Anxiety and patient participation in clinical decision-making: the case of patients with ureteral calculi. Soc Sci Med 1997; 45: 419-427.
27. Stewart M, Brown JB, Donner A, McWhinney IR, Oates J, Weston W. The impact of patient-centred care on patient outcomes in family therapy. London, ON: Centre for Studies in Family Medicine, University of Western Ontario , 1997.
28. Redelmeier DA, Molin JP, Tibshirani RJ. A randomised trial of compassionate care for the homeless in an emergency department. Lancet 1995; 345: 1131-1134[Medline].
29. Charles C, Gafni A, Whelan T. Shared decision-making in the medical encounter: what does it mean? (Or it takes at least two to tango). Soc Sci Med 1997; 44: 681-692.
30. Kleinman A. The illness narratives. New York: Basic Books , 1988.
31. Silverman D. Communication and medical practice: social relations and the clinic. Bristol: Sage , 1987.