verbal analysis of doctor-patient communication

Sm. SC;. Med. Vol. 32, No. IO, pp. 1143-1150, 1991 Printed in Great Britain. All rights reserved

0277-9536/9l $3.00 + 0.00 Copyright 0 1991 Pergamon Press plc

VERBAL ANALYSIS OF DOCTOR-PATIENT COMMUNICATION

LUDW~EN MEEUWESEN, CAS SCHAAP and CEES VAN DER STAAK*

University of Nijmegen, Department of Clinical Psychology and Personality, P.O. Box 9104, 6500 HE Nijmegen, The Netherlands

Abstract-A microscopic analysis of doctor-patient communcation in the general practitioner’s surgery is presented. Verbatim transcripts of 85 medical interviews, audiotaped in a natural situation were analysed. The effects of type of complaint, patient gender and physician gender on the process of verbal communication were assessed. This study focused upon the relational aspects of communication, using Stiles’ Verbal Response Mode coding system (VRM), and. to a limited extent. uoon the content of natient’s . complaints-whether they were primarily somatic’& of a psychosocial nature.

The hypothesis of an asymmetrical relation between physician and patient was confirmed. Results partially confirmed the hypothesis that interviews of psychosocial patients take more time than those of somatic patients. There was also some evidence that psychosocial patients try to exert more control over the conversation as compared to somatic patients. Male and female patients differed in the way they elaborated their complaints. When telling their complaints, women referred more to persons (family, friends, colleagues) than did men. In case of male GP’s. the interviews of female patients took more time than those of male patients.

The largest differences were between male and female physicians. In agreement with the hypothesis males were more imposing and presumptuous (giving more advisements and interpretations). Female GP’s were more attentive and non-directive (giving more subjective and objective information and acknowledgements). The medical interviews of female GP’s took more time than the interviews of their male colleagues.

The results are discussed in the light of theoretical concepts of harmony and discrepancy with respect to doctor-patient communication and theories about gender differences.

Key wordsdoctor-patient communication, conversation analysis, gender differences, somatization

INTRODUCrION

This study focused on the doctor-patient relationship, specifically, on the verbal communication in the medical interview. It is widely agreed upon that the quality of the doctor-patient relationship is very important with respect to the outcome of medical treatment [l-6]. In the case of complaints that are difficult to diagnose medically (about 20%) presented more often by women than by men [7], these relational qualities constitute often the only treatment a physician has at his or her disposal.

Previous microscopic studies of doctor-patient communication have dealt almost exclusively with patients presenting somatic complaints. There are only a few studies of converstation between doctors and patients with psychosocial complaints [8,9]. Also, studies referring to gender differences are scarce [lo]. The research questions were:

(1) What are the communicational characteristics of patient and physician in the various phases of the medical interview?

(2) Does the type of complaint (somatic or psychosocial) influence the communicational pattern of the medical interview?

(3) Does the gender of patient and physician influence the communicational pattern of the medical interview?

*To whom all correspondence should be addressed.

THEORETICAL VIEWS OF THE DOCTOR-PATIENT RELATIONSHIP

With respect to the doctor-patient relationship there are two competing theories, the consensus- model of Parsons [l 1] and the discrepancy-model of Freidson [12, 131. The first assumes a harmonious relationship where the physician is leading and the patient is following; the physician’s role is characterized by high status and control uis-ci-vis the patient. There is a normative pattern of trust: the physician will be attentive to the needs of the patient and will act in the patient’s interests. The patient has to cooperate and to do everything the doctor advises to become healthy as quickly as possible. The discrepancy model argues in contrast that there is an inherent discrepancy between the expectations of the patient and what the doctor can really offer. Only a part of the physician’s control is used to advance the patient’s interests; the physician also uses control to maintain his institutionalized authority [14-161. It is assumed that illness is less objective than has been thought; the reactions of the environment, including the physician, are very important in the process of health and illness. Consequently, the discrepancy mode1 emphasizes the interactional and dynamic aspects of doctor-patient conversation. Both theories postulate an asymmetrical relation between doctor and patient, where the doctor is the leading party, who structures the medical conversation and gives advice, while the patient gives biographical information regarding the complaints. The latter theory

1143

1144 LUDWIEN MEEU~ESEN et al.

stresses the process of negotiation as a result, at least partly, of the interactional process between doctor and patient. This implies that the patient is also regarded as an active party in the interaction.

A third theoretical view of the doctor-patient relationship is worth mentioning, i.e. the patient- centred approach. The relationship is an essential part of the patient centred approach, which acknow- ledges the patient as a person with unique needs and a life-history [6]. In the doctor-patient relationship, patient-centredness will be expressed when the physician is actively interested in the patient’s point of view. This is in contrast to the illness-centred approach, in which the physician is exclusively focussed on the illness or disease, at the cost of loss of interest in the patient as a person. Especially in case of psychosocial complaints a patient-centred approach may be more appropriate. In reality, however, there are strong indications that the illness- centred approach is still predominant [17].

Referring to psychosocial problems in medical practice, there are some indications that patients with psychosocial complaints should try to exert more control over the conversation and the interviews should take more time than interviews of patients with somatic complaints [8, 9, 18, 191.

Female patients present more complaints and are more prone to talk open and personal about their complaints compared with men who are presenting their complaints more concisely and reserved [7,20,21].

As far as physicians are concerned, female doctors should be more affiliative and more sensitive to the relational aspect in conversation and to psychosocial aspects of the complaints [22,23].

Based on the foregoing discussion of the literature the following hypotheses were fommlated:

(1) in general there is an asymmetrical relation between physician and patient in the medical interview;

(2) interviews of psychosocial patients take longer than those of somatic patients; psychosocial patients exercise more control over the interview;

(3) female patients present more complaints, talk longer and show more affiliative aspects as compared to male patients;

(4) female physicians take more time to talk to their patients; show more affiliative behavior and are less directive whereas male physicians are more informative, directive and presumptuous.

METHOD

Participants

The study is based on aselectly chosen audiotaped medical interviews of 85 patients with their general practitioners. In the Dutch health care system the general practitioner, comparable to a family physician, plays a very central role. Each GP takes care of a group of 2000-3000 patients. Specialists are only visited after referral by a GP. The GP’s, 10 in total, were from two middle-sized cities in the Nether- lands. To compare the physicians behavior, the GP’s had to meet criteria as practising their profession already for several years and having a medical

surgery with a heterogenous patient population. The medical consultations were arranged by a ‘IO-min- appointment’. There were 6 male GP’s and 4 female GP’s. The male GP’s were 42 years on average, and the female GP’s were 33 years old.

The patient sample included 54 female and 3 1 male volunteers.

Table 1 contains the demografic profile of the patients. Nearly twice as many women than men consulted the GP (64% vs 36%) during the period of the sample collection. The mean age of the patients in the sample was 39 years (youngest 15, oldest 75). The patients were equally divided amongst the GP’s; also concerning type of complaint and age they were equally divided.

As far as sex, marital status and familiarity of complaint is concerned, the patient sample is repre- sentative. Concerning age the group of 20-29 is overrepresented.

Female patients presented on average more complaints than male patients (2.0 vs 1.5). The seven most common presenting complaints concerned the mus- culo-skeletal system and connective tissue (back pain, muscle pain, pain in shoulders), the circulatory system (blood pressure), the respiratory system (a cold), psychosocial complaints (troubles, anxieties, de- pression, mourning), digestive system (stomach pain), skin and subcutaneous tissue, and contraception.

Coding system

As a method of coding the relational aspects of conversation, the Verbal Response Mode (VRM) system of Stiles [24-261 was used. It has been used frequently for analysis of medical interviews [5,27]. The VRM-system does not focus on the content of utterances but on the intent or relational aspect that an utterance has in communication. It consists of the following eight mutually exclusive and exhaustive categories: acknowledgement, edification, question, disclosure, reflection, confirmation, interpretation, and advisement. Table 2 contains a description of the eight modes.

Table I. Demografic profile of patients

Sex: Female Male

Age: <19 2&29 3G39 40-49 5c-59 602 N.A.

Marital status: Unmarried Married or longlasting

intimate relationship Divorced/separated Widowed N.A.

Type of complaint: Somatic Psychosocial

Familiarity with complaint: New complaint Familiar or chronic

N.A. = not available.

N %

54 64 31 36

4 5 28 33 13 I5 I5 18 14 16 9 II 2 2

14 16

58 68 7 8 6 7 1 1

39 46 46 54

23 27 62 73

Verbal analysis of doctor-patient communication

Table 2. Verbal rewonse mode categories

1145

(1) Acknowledgement:

(2) Edification:

(3) Question:

(4) Disclosure:

(5) Reflection:

(6) Confirmation:

(7) Interpretation:

(8) Advisement:

Conveys reception of or receptiveness to communication from others: simple acceptance, salutations, “Yes”, “Hm-hm”, “Hello doctor”; States objective information, “The blood pressure is 120 over 70”; Requests information or guidance, “How long ago did it start bothering you?“; Reveals thoughts, feelings, perceptions, intentions, all subjective information, “I guess it’s muscular pain”; Puts other’s experience into words: repetitions, restatements, clarifications, “You are feeling miserable”; Compares speaker’s experience with other’s: agreement, disagreement, shared experience or intention, “I don’t agree”, “You are right”; Explains other to himself/herself: judgements, evaluation, or labeling of other’s experience or behavior. “You are too nervous”; Attempts to guide other’s behavior: advice, commands, suggestions, permission, prohibition, “You are not allowed to work next week”.

The classification of an utterance is determined by anwering three two-choice questions: (1) does it refer to the experience of the speaker or the listener, (2) does it refer to the speaker’s frame of reference of the listener’s, (3) does the speaker presuppose specific knowledge about the listener? These three dichotomous classification principles are labeled source of experience, frame of reference, and focus. The principles are dichotomous; each can have the value ‘speaker’ or ‘other’. The structure of the classification system is summarized in Table 3.

Beside quantification of the medical interview in terms of the eight verbal response modes (expressed in relative frequencies of utterances) it is possible to do so in terms of the three classification principles, by systematically adding up four categories each time. The resulting scores are called role-dimensions [26]. According to their psychological effect they have received the following labels: attentiveness vs informativeness, nondirectiveness (acquiescence) vs directiveness, and non-presumptuousness vs presumptuousness. Table 3 shows how each of these dimensions is composed. For example, disclosure, edification, advisement and confirmation together form the role-dimension ‘informativeness’.

Furthermore, it is possible to express a degree of imposition of an utterance of conversation fragment, based on a ranking of imposition of the eight categories. For this purpose a hierarchy of imposition has been devised [26] with values ranging from 1 (least imposing: acknowledgement) to 8 (most imposing: advisement). Table 2 contains the values of the imposition-index for each of the categories. In the case of a conversation fragment, the score can be calculated by averaging the imposition-index of all of a speaker’s utterances. The score will therefore have a value between 1 and 8.

Procedure

The 85 medical interviews were audiotaped and transcribed verbatim. The transcript were segmented according to the VRM guidelines (see below). The unit of analysis was a sentence or a part of a sentence, such as an independent clause. A speaking turn could contain more than one unit of analysis. The segmen- tation results in a dataset of 21,256 units of analysis.

Before coding all utterances of patient and physician, the medical interview was divided into three segments: the medical history, the physical examination and the conclusion segment (diagnosis and prescription). This sequential pattern is charac- teristic of medical interviews [S, 19,271. All units of analysis were coded according to the VRM manual [28], based on Stiles’ manual [24], by 4 dyads of trained raters. The intercoder reliability, assessed by Cohen’s kappa [29] was 0.78 on average.

Beside analysis of VRM, which measures the relational aspects of communication, a limited analysis on the content of the medical interview was performed. This focused on the manner of presenting and elaborating complaints by patients. Specifically, the frequency with which patients referred to persons when presenting and elaborating their complaints was assessed. These persons could be family members, friends and acquaintances, colleagues or more abstractly ‘people in general’ (e.g. “I have nice neighbors” or “My friend told me it can be dangerous”).

The risk for somatisation, a process which is characterized by a proneness to experience and com- municate physical discomfort, which can be associ- ated with a reaction to psychosocial stress [30], was assessed by a widely used Dutch general health questionaire called VOEG [31], measuring the ten- dency to somatize psychosocial stress. This measure

Table 3. Taxonomv of verbal resoonse modes

Focus

(Speaker)

Non-presumptuousness

Frame of reference

(Speaker) Directiveness

(Other) Acauiescence

Source of experience (Speaker) (Other)

Informativeness Attentiveness

Disclosure Question

Edification Acknowledaement

(Other)

Presumptuousness

(Speaker) Directiveness

(Other) Acquiescence

Advisement lnterpretatlon

Confirmation Reflection

1146 LUDWJEN MEEIJWESEN et al.

Table 4. Mean number of utterances of patient and physician

Patient Physician

Medical history 75 55. Physical examination 13 19. Conclusion segment 32 46’

Total I20 120

l P < 0.001; two-tailed r-test.

can be compared with the somatization subscale of the XL-90 [32]. The list contains 21 questions about physical health, with dichotomous answers, like “Do you frequently have headaches?” yes/no. The score lies between 0 (no risk) and 21 (very high risk). The cut-off score was based on the mean scores of a larger group: men were considered at risk for somatisation when having 5 or more complaints. Women had to have 6 or more complaints. Because the rather complicated relation between physical complaints and psychosocial reasons for it, the diagnosis of the physician was also taken into account: there had to be a referral to a psychosocial complaint or stress.

For comparisons between physician and patient behavior, paired t-tests were performed. To assess the effect of type of complaint, patient gender and physician gender on the communicational variables, uni- variate analysis of variance was performed. The design controlled for individual differences between the physicians.

General profile

RESULTS

The mean duration of the interview was 9’13”. The shortest one lasted 2’25” and the longest one 23’15”. The mean number of utterances of the whole interview was 250. Of this number 95.9% took place during the medical history (segment 1), the physical examination (segment 2) or the conclusion segment (segment 3). The other 4.1% referred to greeting ceremonies at the beginning and the end of the interview, and were not used in the analysis. As Table 4 is showing, within segments there is an asymmetry between physicians* and patients’ numbers of utterances, which is not found for the whole interview.

As far as the VRM categories are concerned, patients presented subjective and objective information (disclosures and edifications) at the start of the interview and also during the physical examin-

ation, and gave more acknowledgements in the conclusion segment. At the beginning of the interview, physicians engaged in active listening (giving many acknowledgements) and probing for specific data (asking questions). At the end of the interview, they gave medical information (edifications) and prescribed a therapeutic regime (advisements). At the beginning of the interview the patients appeared to have the opportunity to present his or her complaints and elaborate on them. As the interview progressed the physician became the dominant partner, by delivering more utterances like advisements, interpretations and edifications and by giving less acknowledgements.

The behavior of patient and physician corresponds therefore with the functions of the three segments of medical conversation. The results are in line with Stiles’ data [27]; the same can be said about the results of the role-dimensions (see Table 5): in almost every phase patients and physicians differed greatly with regard to the role-dimensions. The scores of the role-dimensions correspond to the different tasks of the three phases of the medical interview. The results of the imposition-index do underline these findings, (see Table 6) and again confirm the hypothesis of the asymmetrical relationship between patient and physician: In the medical history the patient is more directive, and more talkative; however as the conversation progresses the physician becomes the dominant, both verbally and pragmatically.

Type of complaint

Overall, interviews of patients with psychosocial problems lasted not significantly longer than those of patients with somatic problems (9’54” vs 8’23”, n.s.). There was however a significant interaction between type of complaint and physician gender on this variable (F(l,3) = 38.87, P < 0.01). This interaction reflects the fact that female physicians talked longer with psychosocial patients (12’18”) compared to somatic patients (8’57”) while for male physicians no such difference was found (7’48” and 7’31”).

Furthermore, psychosocial patients gave more interpretations in the conclusion segment than did somatic patients (0.12 vs 0.07; F( 1,35) = 13.58, P < 0.001). The latter group gave more reflections in the conclusion segment (1.93 vs 0.40; F( 1,35) = 4.17, P < 0.05). These results suggest that at the end of the interview somatic patients shared the doctor’s frame of reference more than the psychosocial group did.

Table 5. Mean values on role dimensions (mean proportions of scored utterances) of 85 medical interviews

Medical Physical

Phases history examination’ Conclusion

Role-dimensions P D P D P D

Attentiveness-informativeness (other’s experience)

Acquiescence-directiveness (other’s frame of reference)

0.10 0.83.’ 0.27 0.50** 0.46 0.44

0.42 0.65.. 0.40 0.42 0.62 0.43.’

Presumptuousness-Non-presumptuousness* (focus on other)

0.03 0.19.’ 0.06 0.36” 0.10 0.45’.

*A high score indicates a high score on the pole of the dimension mentmned first. l *P < 0.001; two-tailed I-test. P = patient; D = doctor. ‘Not all medical interviews contained the physical examination segment. In that case the number

of interviews was 61.

Verbal analysis of doctor-patient communication 1147

Table 6. Imposition-index of patient and physician, for the three phases of the medical interview

Phases Patient Phvsician

Medical history Physical examination Conclusion segment

3.02 2.W 2.66 3.94’ 2.43 4.08’

Total 2.70 3.53.

l P < 0.001; two-tailed I-test

No other meaningful significant differences were found regarding the effect of type of complaint on VRM codes or dimensions.

Patient gender

Interviews of female patients took longer than male patients’ interviews (9’56” vs 7’58”; F( 1,35) = 3.11, P < 0.1). However, the interaction with physician gender was significant (F(1,3) = 11.96, P < 0.05). Male physicians talked longer with female than with male patients (8’42” vs 5’44”, F( 1,24) = 8.11, P < 0.01). In case of female physicians the difference was not significant (11’9” vs 10’3”).

With respect to the VRM-categories the following observations were made. During the medical history, female patients used more agreements or disagree- ments with their doctor than male patients did (confirmations: 1.1 vs 0.8; F( 1,35) = 6.08, P < 0.05). During the physical examination female patients gave more acknowledgements than male patients (13.4 vs 5.6; F(1,35) = 6.13, P < 0.05). In the conclusion segment as well as for the whole interview, male patients gave more interpretations than females. The scores were respectively 0.17 vs 0.05 (F(1,35) = 9.93, P < 0.01) and 0.14 vs 0.05 (F(1,35) = 5.57, P < 0.05).

Male and female patients did not differ significantly ‘on the imposition-index and role dimensions. How- ever, in the medical history, physicians were more presumptuous with female patients as compared to maiepatients (0.21~~0.16) (F(1,35) = 7.37, P < 0.01). In particular, male physicians presumed more knowledge about female patients (0.25 vs 0.16; F(1,24) = 11.66, P c 0.01; female physicians: 0.17 vs 0.16, n.s.).

Furthermore, physicians were on the whole more attentive to male patients compared to female patients (0.63 vs 0.59, F(1,35) = 5.21, P x0.05). More edifications (objective information) were given by physicians to female patients than to male patients during the physical examination (13.1 vs 5.3; F( 1,35) = 6.80, P < 0.05).

During the medical history, physicans asked more questions to male patients than to female patients (23.6 vs 18.6; F(1,35)=4.17, P x0.05).

Regarding the presentation and elaboration on their complaints it turned out that female patients referred nearly twice as much to persons compared to male patients (1.25 vs 0.67; F( 1,35) = 4.65, P c 0.05). Women were more prone to elaborate on their complaints like talking about contacts with their neighbors, diseases or dead members of the family, opinions of friends or specialists. Type of complaint was not relevant, neither was gender of physician.

Physician gender

The interviews of female physicians lasted longer than those of the male physicians (10’45” vs 7’38”;

Table 7. Imposition-index of male and female physicians

Male Female Phases physician physician

Medical history 2.59 2.40 Physical examination 4.46 3.50” Conclusion segment 4.33 3.82’

*P < 0.05; **p < 0.01.

F(1,3) = 15.66, P < 0.05). As reported above, male physicians discriminated more with respect to gender of the patient, whereas female physicians discriminated more with respect to type of complaint as far as the duration of the conversation is concerned.

Male physicians showed a higher degree of imposition than female physicians in all segments of the interview (see Table 7).

During the physical examination and in the conclusion segment male physicians were more presumptuous than their female colleagues (0.42 vs 0.30 (F(1,3) = 11.93, P ~0.05) and 0.51 vs 0.38 (F(1,3) = 13.27, P < 0.05) respectively). In the conclusion segment male physicians were more directive than female physicians (0.61 vs 0.53; F(1,3) = 12.77, P < 0.05).

With respect to VRM categories of the whole interview: Male GP’s gave more interpretations (10.0 vs 6.0; F(1,3) = 33.00, P < 0.05) and advisements (18.4 vs 13.8; F(1,3) = 9.68, P < 0.05) compared with their female colleagues. On the other hand, female GP’s gave more disclosures (7.6 vs 6.6; F(1,3) = 10.89, P < 0.05) and edifications (15.0 vs 13.0; F(1,3) = 10.66, P < 0.05). Disclosures of physicians were divided is 7 subcategories: sharing feelings/experiences (“I’m also afraid of going to the dentist”), structuring comments (“First, I shall measure the blood pressure”), showing sympathy (“I can imagine how you are feeling”), critical comments (for example about medical operations), here-and- now comments “Oh, I see”), phrases of politeness, and ‘I-don’t-know’ comments [33]. A closer look revealed that the differences between male and female physicians referred to ‘structuring comments’ and ‘showing sympathy’. These utterances were used by female physicians more frequently.

Table 8 gives a summary of the differences between male and female GP’s with respect to the VRM categories. Almost all differences are according to our expectations except those regarding the category edification. Contrary to the expectations, female

Table 8. VRM categories of the whole intervIew of male and female physicians (percentages of utterances)

Male Female VRM categories physician physiclan

(I) Acknowledgement 25.1 32. I (2) Edification 13.0 15.0’ (3) Question 15.3 14.8 (4) Disclosure 6.2 7.6’ (5) Reflection 4.9 5.3 (6) Confirmation 1.0 0.5 (7) Interpretation 10.0 6.0* (8) Advisement 18.4 13.8’ Unscorable 5.1 4.9

Total 100.0 100.0

l P < 0.05. ‘The numbers refer to the percentages of each category in the whole

interview.

1148 LUDWEN MFEUWESEN ef al.

physicians gave more edifications, especially in the It can be concluded that gender-specific socializa- conclusion segment. Male physicians, in this segment, tion processes influence the behavior of patients as gave more advisements and interpretations. Table 8 well as physicians. On the one hand female GP’s also offers a closer look at the source of the differ- have, by choosing a traditional male profession, ences in the imposition index between male and broken with traditional sex roles. They have received female GP’s. Female GP’s scored higher on the less the same professional training as their male col- imposing categories such as acknowledgements and leagues. However, in executing their profession they disclosures, and lower on the more imposing ones like integrate patient-centred relational qualities with the interpretations and advisements. directive character of a medical interview.

CONCLUSIONS AND DISCUSSIONS

A micro-analysis of the conversation of 85 medical interviews confirmed the hypothesis of the asymmetrical relation between physician and patient, in terms of duration, VRM codes, role dimensions and imposition index. The results of the general profile of the medical interview correspond with the results reported by Stiles and co-workers [5,27] that the role-dimensions of patient and physician in the three major segments of the medical interview varied systematically according to their function. The findings of the study can be considered as a validation of VRM taxonomy for studying medical conversation.

There are some indications [6] that a more facilitat- ing rather than a domineering conversation style results in greater compliance and satisfaction of the patients. Stiles et al. [5] noted that patients were more satisfied when they were allowed to express them- selves in their own words during the medical history and when physicians were more informative in the conclusion segment. The results of this study suggest that female physicians met the criteria of a patient- centred approach more often than their male colleagues.

The observed differences in presumptuousness and imposition index between patient and physician support the views of Parsons [1 1] and Freidson [13] who posit a large gap in status between patient and physician. It serves the controlling function in medical encounters but the two views differ in their prediction of the patients’ reactions to it. The general pattern of the medical interview showed consider- able variation, depending on type of complaint, and gender of patient and physician.

Some methodological issues of this study are worth mentioning. The greatest variation in behavior by using VRM turned out with respect to physicians’ role and not the patients’. Apparently, VRM is more suited for analysing the physician’s role (see also Ref. [39]). On the basis of our research material we do not gain the impression that patients behave less varied, but that the VRM categories capture more of the physician’s behavior. It is recommended to make a more detailed subdivision of the disclosures and edifications, the most frequent patient categories, in order to get more insight in the content of the narratives of the patients. This would be especially relevant in case of psychosocial complaints.

There was some evidence that psychosocial patients tried to exert more control over the communicative interaction. This more active role of the psychosocial patient could be interpreted as part of a process of negotiation, which is more in accordance with Freidson’s discrepancy model [ 131.

There are also some indications that female patients are more interactionally oriented compared to male patients. The consequences for the communicative interaction depend on type of complaint and physician gender. Maybe female patients can be regarded as a more active party in the communicative interaction than the prescribed patient role behavior in Parsons’ model [1 l] would permit.

As far as the reliability of VRM was concerned, we already mentioned that it was high. Referring to the validity it can be said that VRM only measures just one level of communication [40]. Especially in the case of complex patterns of communication the no- tion of the multifunctionality of an utterance [41] comes into focus. It is recommended to present case studies in order to further unravel the specific interactional patterns of the complex communication in medical interviews. VRM could then be applied in combination with other research methods such as micro-analysis [41] and/or sequential analysis [42,43].

Depending on the gender of the physician these differences can be enlarged. In the context of a very directive conversation, as the medical interview is, female physicians showed more affiliative behavior. Their male colleagues in contrast, expressed more controlling behavior. At the same time there was some evidence that female GP’s were more sensitive to the specific conversational needs of patients with psychosocial complaints.

In what way do the differences in verbal behavior have an impact on the outcome of the medical treatment? This question has to be investigated by relating the effect of medical treatment to *verbal behavior during the interview. Stiles [44] rightly warns about the pitfalls of process-outcome corre- lations. On the basis of this study one can only conclude that the VRM-taxonomy offers a useful framework for studying gender typed differences in the context of a medical interview.

The gender differences in verbal behavior regarding the physician’s role, are in accordance with state- ments on language differences in general [34,35] and related to physician differences specifically [22-23, 36-381; female physicians are more egalitarian and interactionally oriented and they pay more attention to the relationship as well as a psychosocial factors of the complaints without neglecting the medical- technical aspects.

The small number of GP’s enabled us to investigate the patterns of communication in a very detailed way. Because of this small number it is not unequivocally possible to generalize the results. It is recommended to replicate the investigation with a greater number of GP’s. Possible effects of confounding factors like age or health care attitude have to be taken into account.

The findings which suggest that Parsons’ model [1 1] is more applicable to the communicative pattern of physicians with somatic patients and less in the

Verbal analysis of doctor-patient communication 1149

case of psychosocial patients need to be investigated further, as well as the findings that female patients and female physicians are more interactionally oriented, which is more in line with Freidson’s model [13] and the patient centred approach [6].

Acknowledgemenrs-We wish to thank the Nijmegen Uni- versity Department of General Practice (NUHI) for giving us autiotapes to conduct this research project with, and William B. Stiles for his helpful comments on earlier drafts of this paper.

1.

2.

3.

4.

5.

6.

7.

8.

9.

IO.

11.

12.

13.

14.

15.

16.

17.

18.

19

REFERENCES

Korsch B. M., Gozzi E. K. and Francis V. Gaps in doctor-patient communication: I. Doctor-patient interaction and patient satisfaction. Pediatrics 42, 855, 1968. Korsch B. M. and Negrete V. F. Doctor-patient communication. Sci. A. 227, 66, 1972. Francis V., Korch B. M. and Morris M. J. Gaps in doctor-patient communications: Patients’ response to medical advice. New Engl. J. Med. 280, 535, 1969. Freemon B., Negrete V. F., Davis M. and Korsch B. M. Gaps in doctor-patient communication: Doctor-patient interaction analysis. Pediuf. Res. 5, 298, 1971. Stiles W. B., Putnam S. M., Wolf M. H. and Sherman A. J. Interaction exchange structure and patient satisfaction with medical interviews. Med. Cure 17, 667, 1979. Stewart M. A. What is a successful doctor-patient interview? A study of interactions and outcomes. Sot. Sci. Med. 19, 167, 1984. Verbrugge L. M. Sex differences in complaints and diagnoses. J. Behac. Med. 3, 327, 1980. Bain D. J. G. The content of physician/patient communication in family practice. J. Pam. Practice 8, 745, 1979. Bain D. J. G. Doctor-patient communication in general practice consultations. Med. Educr. 10, 125, 1976. Weisman C. S. and Teitelbaum M. A. Women and health care communication. Potienr Educi. Couns. 13, 183, 1989. Parsons T. The Sociul Sysrem. Free Press, New York, 1951. Freidson E. Patients’ View of Medical Practice. Russell Sage Foundation, New York, 1961. Freidson E. Profession of Medicine: A study of rhe Sociology of Applied Knonsledge. Dodd, Mead & Com- pany, New York, 1970. Waitzkin H. and Stoeckle J. D. The communication of information about illness: Clinical, sociological and methodological considerations. Adv. psychosomuric Med. 8, 180, 1972. Hughes D. Control in the medical consultation: Orga- nizing talk in a situation where co-participants have different competence. Sociology 16, 359, 1982. Waitzkin H. Information giving in medical care. J. Hlth sot. Behav. 26, 81. 1985. Thomassen J. F. M. Consultanalyse: Communicatieve en kwalitatieve aspecten van hulpverlenen in de huisart- spraktijk (Analysis of medical consults: Communicative and qualitative aspects of care in general practice). Dissertation. VU Uitgeverij. Amsterdam, 1986. Nordmeyer J.. Steinmann G., Deneke F. W., Nordmeyer J. P. and Kerekjarto von M. Verbale und nonverbale Kommunikation zwischen Problempatien- ten und Arzten wahrend der Visite (Verbal and non- verbal communication between problem patients and physicians during the visit). Medizinische Psvchologie 8, 20, 1982. Byrne P. S. and Long B. E. L. Doctors Talking IO Putienrs: A Study of the Verbal Behavior of General

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Practitioners Consulting in their Surgeries. Her Majesty’s Stationary Office, London, 1976. Nathanson C. Illness and the feminine role: a theoretical review. Sot. Sci. Med. 9, 57, 1975. Wodak R. Women relate, men report: Sex differences in language behavior in a therapeutic group. J. Prugmurics 5, 261, 1981. Weisman C. S. and Teitelbaum M. A. Physician gender and the physician-patient relationship: Recent evidence and relevant questions. Sot. Sci. Med. 20, 1119, 1985. Gray J. The effect of the doctor’s sex on the doctor-patient relationship. J. R. C. gen. Pratt. 32, 167, 1982. Stiles W. B. Manualfor u Taxonomy of Verbal Response Modes. Chapel Hill, Institute for Research in Social Science, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, 1978. Stiles W. B. Verbal response modes and psychothera- peutic technique. Psychiatry 10, 49, 1979. Stiles W. B. Classification of intersubjective illocution- ary acts. Language Sot. 10, 227. 1981. Stiles W. B., Putnam S. M. and Jacob M. C. Verbal exchange structure of initial medical interviews. Hlrh Psychol. 1, 315, 1982. Meeuwesen L. Handleiding voor een taxonomie van verbale antwoordwijzen (Manual for a taxonomy of verbal response modes, second edition). Internal Report 89 KP 01, University of Nijmegen. Department of Clinical Psychology and Personality, Nijmegen. 1989. Cohen J. Weighted kappa: Nominal scale agreement with provision for scaled disagreement or partial credit. Psychol. Bull. 70, 213, 1968. Lipowski Z. J. Somatization: The concept and its clinical application. A. J. Psychiut. 145, 1358. 1988. Visser A. P. De betekenis van de VOEG: Enkele gengevens over de begripsvaliditeit (The meaning of the VOEG: Some data about the construct validation). Gezondheid en Sumenleving 3, 177. 1983. Derogatis L. R. and Cleary P. A. Confirmation of the dimensional structure of the SCL-90: A study in construct validation. J. clin. Psycho/. 33. 981, 1977. Seuntjens M. Sekseverschillen in communicatie bij artsen tijdens het medisch consult, toegespitst op disclosure (Gender differences in communication of physicians during the medical interview, focused on disclosure). Doctoral theseis, University of Nijmegen. Department of Clinical Psychology and Personality, Nijmegen, 1988. Maltz D. N. and Borker R. A. A cultural approach to male-female miscommunication. In Language and Social Identity: Studies in Interacrional Sociolinguistics 2 (Edited by Gumperz J.). Cambridge University Press, Cambridge, 1982. Smith P. M. Language, rhe Sexes and Sociery. Blackwell. Oxford, 1985. Leserman J. Sex differences in the professional orien- tation of first-year medical students, Se.x Roles 6, 645. 1980. Comstock L. M., Hooper E. H., Goodwin J. M. and Goodwin J. S. Physician behaviors that correlate with patient satisfaction. J. med. Educar 57, 105, 1982. West C. Routine Complications: Troubles with Talk Between Doctors and Puriems. Indiana University Press, Bloomington, 1984. Stiles W. B. Development of a toxonomy of verbal response modes. In The Psvchorherupeutic Process: A Research Handbook (Edited. by Greenberg L. S. and Pinsof W. M.). The Guildford Press. New York. 1986. Stiles W. B. Levels of intended meaning of utterances. Br. J. clin. Psychol. 25, 213, 1986. Labov W. and Fanshel D. Therapeutic Discourse: Psychotherapy us Conversation. Academic Press, New York, 1977.

1150 LUDWIEN MEEIJWESEN et al.

42. Meeuwesen W. and Schaap C. Het uiteenlopend 43. Meeuwesen L. Spreekuur of xwijguur: Somatische perspectief van arts en patient(e) tijdens het fixatie en sekse-asymmetrie tijdens het medisch consult medisch interview (Different perspectives of physician (Speaking or being silent: Somatic fixation and gender and patient during the medical interview). In asymmetry during the medical interview). Dissertation, Gespreksanalyse (Conversation analysis) (Edited by Nijmegen University, Nijmegen, 1988. Scholtens A. and Springorum D.), p. 195. Nederasselt, 44. Stiles W. B. Evaluating medical interview process com- 1986. ponents. Med. Care 27, 212, 1989.

verbal analysis of doctor-patient communication

Documents