Consulting skills training and medical students' interviewing efficiency

121 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

Consulting skills training and medical students’ interviewing efficiency

BJ Evans, G J Coman & B Goss’ Monash University, Caulfield, Victoria, Australia, and 1 Austin and Repatriation Medical Centre Clinical School, Heidelberg, Victoria, Australia

SUMMARY

In more traditional medical education, medical students took a patient’s medical history by asking a series of sequenced, routine questions, covering presenting medi- cal problem(s); medical history; social and personal history; systems review; and physical examination. Following this process, the student then attempted to derive the patient’s medical problems. This inductive problem-solving paradigm may not assist students to prepare for their future interviewing needs, given doctors use a hypothetico-deductive, problem-solving approach when interviewing patients and numerous researchers have developed specialized communication skills train- ing programmes designed to enhance students’ inter- viewing skills.

Students given specific consulting skills training have tended to show significantly greater interpersonal effec- tiveness and improved interview behaviours compared with students who experience traditional patient clerk- ing training. These improvements in interviewing tend to persist over the period of students’ medical training.

The aim of the present study was to determine whether specialized communication skills training helped stu- dents elicit greater quantity and quality of information from patients and if so, whether such information assisted students in improving their diagnostic skills. Videotaped history-taking interviews conducted by stu- dents trained in communication skills and untrained (control) students were rated for their interview effi- ciency.

A comparison of ratings given by experimentally naive, independent observers revealed that trained students were more efficient, but took no longer than their control group counterparts to elicit fuller, more relevant information. However, the student groups did not differ in the accuracy or scope of their medical diagnoses. It is argued that students’ lack of medical knowledge in this early phase of their clinical training militated against their being able to use their inter- viewing competence to derive more potentially accu- rate medical diagnoses.

Keywords

Australia; *clinical competence; *communication; *education, medical, undergraduate; *interviews

INTRODUCTION

The’format for medical interviews conducted by students early in their training traditionally consisted of a dia- logue between patient and medical student, in which the latter attempted to elicit in a logical and sequential manner, a description of the patient’s specific illness events and symptoms leading to their hospitalization and surgery. This approach, known as the clinical method (Roter 1983), involved the student taking the patient’s history in a highly structured interview, asking a set series of questions in a sequential fashion, to ensure they covered all possible topics from the list to be covered (Evans 1990). Having covered the patient’s presenting problem, reviewed associated and past complaints, family medical history, and social history (Evans et al. 1989; Evans et al. 1991) the student employed an induc- tive problem-solving paradigm to diagnose the patient’s presenting illness (Metcalfe 1983; Evans er al. 1991). For many years, this clinical method provided the basis for teaching a standardized history-taking procedure.

The value for students conducting their medical inter- views according to this procedure was twofold. First, it provided each student with a clear description of what they had to do in order to achieve the goals of the inter- action. Second, it provided a simple, ordered and logical format for conducting what is a very complex process (Roter 1983; Novack 1985; Evans et al. 1989; Evans er al. 1991).

Numerous studies have shown that students trained using the clinical method gradually showed progressive improvement in their ability to focus on patients’ patho- physiological complaints and elicit medically relevant information. However, in many cases, their focus on physiological aspects of medicine was often accompanied by an unwillingness or inability to explore patients’

Correspondence: Dr Barry J Evans, Psychology Department, Monash University, Caulfield Campus, Caulfield East, Victoria 3145, Australia

122 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

psychological and social concerns which impact on ill- ness states and illness reactions (Flaherty 1985; Preven 1986; Stewart & Roter 1989). Studies of medical stu- dents’ interaction skills showed that, early in their early clinical days, many students displayed an interest in talk- ing with patients and showed some skills in facilitating communication and asking appropriate questions. By the time students had reached their final years of training, however, they focused intently on diagnostic informa- tion, to the apparent exclusion of concern and empathy for the patient (Barbee & Feldman 1970; Preven et al. 1986; Stewart & Roter 1989).

Researchers put forward a number of reasons to explain this shift in medical students’ diagnostic skills and focus. It was argued that the clinical method did not help stu- dents make the transition from hospital-based patient interviewing to the quite different approach needed for general practice consulting (Metcalfe 1983; Evans et al. 1991). The inherent notion in the traditional history-tak- ing procedure is to obtain all the data before attempting a diagnosis. In medical practice, however, practitioners more generally use a hypothetico-deductive, problem- solving routine, actively formulating hypotheses early in the medical interaction and rejecting or refining these hypotheses as the consultation progresses (Metcalfe 1983; Flaherty 1985). As students become more medically com- petent, they too adopt this hypothetico-deductive approach, ‘screening out’ information not considered rel- evant to the patient’s actual diagnosis. Thus, the clinical method was seen to be misdirected in its approach, with too great an emphasis on the procedure the student should use (Stewart & Roter 1989; Evans et al. 1991; Evans et al. 1993). In the experience of clinical instructors, students were not trained to respond to the cues given by patients as the history-taking proceeded, nor were they trained to modify their questions to suit the particular medical hypothesis they wished to explore (Fitzgerald 1980; Evans et al. 1991; Evans et al. 1993).

Clinicians also suggested the traditional clinical method focused too much on diagnostic skills and rela- tively little attention was given to patient management issues (Metcalfe 1983; Flaherty 1985; Meichenbaum & Turk 1987). As Roter has so cogently commented, the educational emphasis on diagnosis of patients’ present- ing illnesses meant an inappropriately small portion of time was spent learning about patient management issues and the importance of patients’ psychological states in illness and health. Diagnosis of a patient’s ill- ness is an important step in treatment and care, but the majority of doctor-patient interactions are, in fact, fol- low-up visits in which the focus is on problem manage- ment, health education and counselling, rather than on diagnosis (Roter 1983; Stewart & Roter 1989). Students’

history-taking format addresses, in some way, the task of diagnosis, but they may learn very little about psycholog- ical and social factors, which have been demonstrated to be so important in patient care (DiMatteo & Friedman 1984; Stewart & Roter 1989).

Numerous researchers and medical schools have evalu- ated the effects of specialized consulting skills training on medical students’ attitudes towards patients and their interview behaviours, compared with the effects of approaches embodied in the clinical method, beginning with the landmark work of Maguire and his associates (Maguire et al. 1978) and extending to more recent stud- ies (Alroy et al. 1984; Evans el al. 1989; Joesbury et al. 1990; Evans et al. 1993). Most studies clearly established that students given specialized consulting skills training show significantly improved interpersonal skills over their traditionally trained counterparts. They communi- cate warmth and undertaking to patients (Engler et al. 1981; Evans et al. 1992a) and are more adept at detecting and responding appropriately to patients’ verbal and non-verbal cues (Sanson-Fisher & Maguire 1980; Alroy et al. 1984), compared with their more traditionally trained counterparts. Recent studies have shown that students trained in communication retain their interest in, and caring for, patients, well into their education (Davis & Nicholaou 1992), unlike students who received more traditional training (Preven et al. 1986).

These research studies show that, as students learn interviewing techniques and skills, they relate more effectively to patients on the interpersonal level. In so doing, they may be able to elicit from patients a greater range and quality of information, which may then be used diagnostically to help the student elicit the patient’s presenting medical problem(s). Student interviews showing this greater quantity and quality of patient information have been labelled s t r u c t u d y efficient by Evans et al. (1991), that is, characterized by the student asking full, relevant and concise questions of the patient regarding presenting medical problems, incorporating an evaluation of psychological and social issues impor- tant to the patient. The greater quantity and quality of information may then be combined with the student’s medical knowledge to derive a more accurate diagnosis or possible diagnostic alternatives.

To test the hypothesis that students given specialized consulting skills training should be able to conduct his- tory-taking interviews with patients that were more structurally efficient than those conducted by their tradi- tionally trained counterparts, Evans and his associates (Evans et al. 1991), carried out a post hoc analysis of data obtained in an earlier phase of an on-going evaluation of consulting skills training, run by the Austin Repatriation Medical Centre Clinical School in Melbourne, Australia.

- Consulting skills and interviewing efficiency BJEvam et al.

123 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

Each of the 60 participating students was randomly allo- cated to an experimental (trained) or control (untrained) group. Students in the trained group participated in a consulting skills training programme (Evans et al. 1989; Evans 1990). Videotaped history-taking interviews com- pleted by all students pre- and post-training were rated by two trained psychologists, using the History-Taking Rating Scale (Evans et al. 1989; Evans 1990; Evans et al. 1992b). The ratings clearly showed that, following their training, experimental students received significantly higher ratings for all aspects of their communication behaviours. These students significantly improved inter- view skills over time and this contrasted markedly with the interview behaviour ratings achieved by control- group students, who maintained constant mean scores on the History-Taking Rating Scale. The duration of inter- views conducted by the trained-group students did not differ from those conducted by control-group students at the interview times (Evans et al. 1992b) and patients interviewed by trained students were significantly more satisfied with their interviews, compared with patients of control-group students (Evans et al. 1992a).

In the Evans et al. (1991) study, the students’ post-train- ing, videotaped, history-taking interviews were indepen- dently rated by a further two trained observers: a psychia- trist and a general practitioner, to evaluate the students’ interviewing efficiency, using a medical rating scale (Brockway 1978; Robbins et al. 1979), assessing the follow- ing dimensions: Commencement of the Interview; Problem Processing; Communication with the patient; Summary; and Overall Evaluation. The experimentally naive observers concluded that students in the trained group displayed significantly better history-taking skills compared with control group students. The former elicited more information from patients and their interviews were rated to be more medically relevant in that they clarified issues with patients and followed-up on medical data in a way not matched by students in the untrained group.

The aim of the present study was to determine whether students given specialized consulting skills training would be able to use the information they collected in their interviews to make more accurate diagnoses of patients’ presenting problems.

EXPERIMENTAL STUDY

Subjects

Sixty medical students completing their first year of clin- ical training at the Austin Repatriation Medical Centre Clinical School participated in the study. This group consisted of the entire class of first year students in the year. Students were randomly allocated to either experi-

mental (trained) or control (untrained) groups, and a comparison of these two groups revealed no differences on sex distribution or mean age. All students had com- pleted the first 3 years of their medical education at the University of Melbourne in the same class and were com- pleting their first clinical year together.

Experimental measure

The students’ interviews with simulated patients were observed by experimentally naive, medically trained observers, using the Medical History-Taking Scale developed in the Evans et al. (1991) study. Table 1 lists and describes the 21 items in this scale. This scale was derived from that developed by Brockway (1978), which assessed interview behaviours of medical students on the dimensions described in Table 1. These items were derived from studies of medical interviews and problem- solving literature and evaluated in a small reliability study (Brockway 1978). The scale was used in the evalua- tion of their communication skills programme by Robbins and his associates (Robbins er al. 1979) and by Evans et al. (1991) in a similar communication skills training evaluation. No other validity or reliability data have been reported, but the scale does have a good face validity and proven reliability (Evans et al. 1991).

Procedure

Early in their first term of clinical training, students were randomly allocated to either trained (experimental) or untrained (control) groups. This was done by group, as students operate in small groups of five or six students, and it was not feasible to divide student groups into con- trol and experimental subjects. Each student completed a videotaped, history-taking interview with a medical or surgical patient, to provide a pre-measure of communica- tion competency. Each tape was subsequently analysed by naive observers, using the History-Taking Rating Scale developed and validated by the researchers (Evans er al. 1989; Evans 1990; Evans et al. 1992b), to check for chance differences between the two groups. No differ- ences in interview performance were found.

Student groups in the experimental group underwent training in consulting skills, provided by the first and second author, as part of their clinical training. This training programme is fully described in Evans et al. (1989) and Evans (1990). Students in the experimental group completed their post-training videotape within 2 weeks of having their consulting skills training and a matched control group completed their second video- taped interview at the same time, to ensure, as much as possible, for group complementarity.

Consulting skills and interviewing efficiency B y Evans et al.

124 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

Table 1 Variables rated in students’ history- Variable Description taking consultations

Commencemetu of interview 1 Initial greeting of patient 2 Explanation

Problem mocessinn 3 4 S 6 7

a

9

10

11 12

13 14 1s

Chief complaint(s) Major symptoms Onset and aetiology Location Character and severity

Course and duration

Relationship with other

Prior medical care

Patient’s interpretation Relevant psychosocial

Past health Family health Clinical perspective

Communication with parient 16 Information

17 Patient’s questions

Summary of session 18 Summary 19 Feedback

20 Thanks patient

Overall evaluation 21 Overall impression

Greeting procedure and introduction of self Statement of interview purpose

Ability to elicit main problems Ability to elicit main symptoms Ability to elicit time of onset and what occurred Ability to elicit location of symptoms Ability to elicit information about the course of illness to date Ability to elicit information about the course of illness to date Ability to explore the effect of the symptomdillness on bodily functioning Ability to elicit information about the patient’s care to date and effects of therapy Ability to establish how the patient views the illness Ability to explore effect of illness on patient’s psychosocial state Ability to detail patient’s prior medical status Ability to detail patient’s family history Overall rating of student’s grasp of presenting illness, symptomatology, aetiology and care

Ability to give accurate information and correct misapprehensions Acknowledgement of patient’s question and appropriate answering

Ability to summarise data and findings Recognition of patient’s needs by asking for questions or comments Acknowledges patient’s contribution

Global rating of student’s diagnostic efficiency

Students in the control group received their consulting skills training immediately after all students had com- pleted their second videotaped, history-taking interview.

Videotaped interviews were conducted using simu- lated patients to ensure consistency between groups. Students are dependent upon the range and type of medi- cal and surgical patients hospitalized at any one time for their interviewing experience. To ensure consistency in case presentation, it was decided to use simulated patients, who played one of three patient roles written by a senior member of the clinical school.

Researchers have used actual patients in communica- tion training evaluations (Maguire et al. 1978; Knox & Bouchier 1985), while others have used actors (Bird & Lindley 1979; Hannay 1980). Some have used both real

and simulated patients (Craig 1992; Davis & Nicholaou 1992). While concern has been expressed that students may relate differently to patients and actors, Sanson- Fisher & Poole (1979) found no differences between the students’ empathy levels in interactions with real and simulated patients, and the students could not discrimi- nate between the two. This finding was replicated by Maguire et al. (1986). The available data suggest simula- tors provide an ideal means of training programmes (Sanson-Fisher & Poole 1979; Evans 1990; Craig 1992). In the present study, each simulator played one of three roles (described in Appendix 1) and was interviewed by no more than three students, with no more than one interview being allowed on any one day. The case histo- ries were designed to confront the student with an

Consulting skills and interviewing efficiency BJEvans et al.

125 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

ambiguous set of symptoms, which would require careful questioning to elicit.

Each student was given a maximum of 20 min in which to elicit the simulated patient’s symptoms and relevant history. At the completion of the interview, the student was asked to compile a list of possible diagnoses for the case presentation. These possible diagnosis lists were summed by student group for each of the three simula- tions and checked for accuracy by a senior member of the clinical school.

Videotapes were independently analysed by the two medical practitioners who participated in the Evans et al. (1991) study. All ratings were completed independently, to minimize problems of reliability in such behavioural assessments (Kazdin 1977). To minimize problems of bias in the ratings, observer naivety regarding group member- ship and random ratings of videotapes from the two groups was ensured. When all ratings had been completed, final inter-rater correlations were computed, suing Spearman’s rho, with correlations ranging from 0.75 to 0.84. Observers’ scores were summed, given that a 5-point scoring scale gives individual variable coefficients of dubious value. The overall correlation was 0.81. On the basis of these generally high correlations, the two observers’ scores were combined, to yield a total possible score of 10.0 on each variable.

Group means for the five categories of interview behaviour, described in Table 1, were then compared using analysis of variance (ANOVA), on a PC version of SPSS-X (Keppell982; SPSS 1986).

RESULTS

Table 2 provides mean ratings for the behaviour cate- gories examined by the Medical Interview Rating Scale, together with F-values and significance.

The data showed trained students were significantly more efficient on all areas assessed by the Scale: Commencement of the interview; Problem Processing; Communication; Summary; and Overall effectiveness, compared with their control-group counterparts. The mean interview time for the trained group was 15min 42 s, compared with a mean interview time of 15 min 3 1 s for the control group, a non-significant difference (t [ 13 = 0.09; P = NS).

Trained and untrained student groups did not differ in the total number of possible diagnoses they generated, nor did their diagnostic possibilities differ in probability.

DISCUSSION

Independent rating of their consultation behaviours con- firmed our previous research that trained students are

significantly better in their ability to logically and efti- ciently obtain information from patients, compared with control-group students (Evans et al. 1991).

In their introduction to interviews, trained students (M = 1656) were rated significantly better than untrained students (M = 12-03) (F = 68-07 [ 1,631; P < 0.000). Both means represent a total for the group from a maximum of 20 (two items each rated out of five, by two raters). The trained students were more effective in greeting the patient, explaining the purpose of the interview and explaining the structure of the interaction, confirming the results of our previous studies (Evans et al. 1989; Evans et al. 1992b).

In their processing of patients’ medical history and elicitation of medical data, trained students (M = 106.00) were rated significantly better than con- trol-group students (M = 74.45) (F = 96.06 [1,63]; P c 0.000). The maximum score on this variable is 130, calculated at 10.0 for each of the 13 dimensions on the sub-scale. The significant difference in mean scores represents a very important difference between groups in their ability to explore with patients the medical complaints, symptoms, aetiology, course, and treat- ment. Trained students also elicited more information related to their patients’ interpretation of their illnesses and the effects of these on psychosocial life. The data showed that the two observers consistently rated the two groups as very different in their interview behaviours, a result perhaps partially due to a ‘halo’ effect, as the raters moved down the sub-scales on the variable. Still, the results suggest a marked difference between the two student groups in terms of their ability and apparent willingness to gather medical data from patients.

Although the purpose of the interview was not to specifically give information to patients, trained-group students (M = 1653) were rated significantly better on the sub-scale, ‘Communication’ than were control- group students (M = 11.84) (F = 86.25 [ 1,631; P < 0.000). This result is consistent with our previous findings that students trained in communication give better quality information to patients and are perceived, by observers and by patients alike, to more readily answer questions (Evans et al. 1992a; Evans et ali 1992b). An important focus in our training programd is to impress upon students the need to address patients’ concerns, to readily answer questions if this is possible and appropriate, to refer the patient to the appropriate authority, or to act as an agent for the patient.

On the interview sub-scale, ‘Summary of Interview’ trained-group students were rated more effective in their interview summary and closure (M = 26-12) com-

Consulting skills and interviewing efficiency I33 Evans et a l

126 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

Trained Untrained Table 2 Mean scores on the Medical Interview Rating Scale, together with F-value and significance

Variable X SD X SD F(1,63) P

Introduction to interview’ 1656 2.38 12.03 2.03 68.07 0.OOO Problem diagnosisb 106.00 13.39 74.45 12.55 96% 0.000 Communication“ 1653 2.62 11.84 1.20 86.25 0400 Summary of interview‘ 26.12 3.95 18.60 2.39 86.44 0.OOO Overall rating“ 7.53 1.41 4.21 1.08 113.34 0.OOO

”Maximum score 20.0 (2 items); bmaximum score 1300 (13 items); ‘ maximum score 30.0 (3 items).

pared with control group students (M = 18.6) (F = 86.44 [ 1,631; P < 0.000). As part of their training, we teach students to summarize the interaction (for example, by saying: ‘Let’s see, we’ve covered the reason you’re in hospital, your previous medical history (etc.) - is there anything you’d like to tell me?’). They are then trained to offer patients an opportunity to talk and to give clear verbal and non-verbal cues that the interview is to shortly terminate.

At the end of each videotape, the two observers gave the interview an overall rating for efficiency (‘Overall Evaluation’). The trained-group students achieved a mean of 7-53 (out of a possible 10)’ compared with the control-group’s mean of 4.2 1, a significant difference (F = 113.43 [ 1,631; P < 0.000). While there may well be a halo effect operating in the observers’ minds, the differ- ence between the two groups strongly reinforces the now clearly proven fact that communication skills training does promote more effective consulting behaviours (Evans 1990; Evans et al. 1991; Evans et al. 1992b; Evans et al. 1993), the effects of which persist over time (Craig 1992; Davis & Nicholaou 1992). What is all the more per- tinent here, is that medical practitioners have rated the medical interviews of students trained in communica- tion skills to be significantly better than those of un- trained students, replicating our earlier finding (Evans et al. 1991).

Despite their significantly better interview perfor- mance, as rated by the two medical observers, the lists of possible diagnoses generated by trained-group students did not differ in quantity or quality from those generated by control-group students. The mean number of possible diagnoses for each of the three case presentations for each group did not differ, and the senior clinician who checked the possible diagnoses lists rated both groups the same for accuracy of diagnosis.

The reason for this discrepancy between interview ef i - ciency and the students’ lists of diagnoses and probability of diagnoses may lie in the students’ inability, early in

their clinical training to apply their (limited) medical knowledge (Maguire 1984; Evans 1990). Many students report that, early in their training, a significant problem is being able to combine their medical knowledge with the information they elicit from patients in interviews. They suggest, over time, and with experience, that they are able to integrate their learning with patient data (Maguire 1984), in a way students in the present study may not have been able to do, given our students’ relative lack of medical competence and interview experience. We may find that students with greater medical compe- tence and patient interviewing may well be able to define patients’ problems more clearly and definitively. This is an issue for further research.

A problem in our study may lie in our failure to ask students to generate a specific number of the most proba- ble diagnoses. We asked them to generate as many diag- noses as they could think of. Doing this may have meant that students in both groups generated extensive lists of probable or possible diagnoses, clouding any difference between the groups as to their ability to correctly diag- nose the patients’ presenting problems. On-going research is aimed at determining the effect of changing the students’ instructions.

CONCLUSION

Our studies show students trained in communication skills are better interviewers and are better able to elicit reliable and quantifiably more information from patients than are students not given such specialized communica- tion training. This communication competency and accuracy occurs in interviews which do not differ in length to those conducted by less well trained students. Studies now suggest this communication competence and desire to communicate more effectively with patients persist with medical training (Davis & Nicholaou 1992).

The Medical Interview Rating Scale developed as part of our work is a useful evaluation device. It has face valid-

Consulting skills and interviewing efficiency BJEvam et al.

127 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

ity in that the behaviours and techniques rated on the Scale appear to be important components in the task of accurate interviewing. Apart from its value as a research measurement device, an additional benefit is that it allows us to assess students’ diagnostic skills and to pro- vide an evaluation of students’ specific weaknesses in their medical interviewing. As part of our clinical pro- gramme, we now routinely assess interviewing video- tapes completed by our current medical students, using the Scale. This assessment allows us to determine the specific competencies and weaknesses of individual stu- dents, with a view to providing specific remedial training in these areas of weakness.

ACKNOWLEDGEMENT

This research was funded by a grant from the Medical Defence Union.

REFERENCES

Alroy G, Ber R & Kramer D (1984) An evaluation of the short term effects of an interpersonal skills course. Medical Education 18,

Barbee R & Feldman E (1970) A three year longitudinal study of the medical interview and its relationship to student performance in clinical medicine.Journa1 of Medical Education 45,7714.

Bird J & Lindley P (1979) Interviewing skills: the effect of ultra- brief training for general practitioners: a preliminary report. Medical Education 17,235-9.

difference does it make? Evaluation and Program Planning 1, 211-20.

Craig J L (1992) Retention of interviewing skills learned by first- year medical students: a longitudinal study. Medical Education 26,276-81.

Davis H & Nicholaou T (1992) A comparison of the interviewing skills of first- and final-year medical students. Medical Education

85-9.

Brockway B S (1978) Evaluating physician competency: what

26,441-7. DiMatteo M R & Friedman H S (1984) Social Psychology and

Medicine. Oelgeschlager Gunn & Hain, Cambridge, MA. Engler C M, Saltzman G A, Walker M L & Wolf F M (1981)

Medical student acquisition and retention of communication and interviewing skills.Journa1 of Medical Education 56,572-9.

PhD Thesis, University of Melbourne, 1990.

skills training and patients’ satisfaction. Health Communication

Evans B J, Stanley R 0, Coman G J & Sinnott V (1992b) Measuring

Evans B J (1990) Medical interaction skills training: an evaluation.

Evans B J, Stanley R 0 & Burrows G D (1992a) Communication

4,155-70.

medical students’ communication skills: development and evaluation of an interview rating scale. Psychoba and Health 6, 213-25.

Evans B J. Stanley R 0, Mestrovic R & Rose L (1991) Effects of

communication skills training on students’ diagnostic eficiency. Medical Education 25,517-26.

Evans B J, Sweet B & Coman G J (1993) Behavioural assessment of the effectiveness of a communication programme for medical students. Medical Education 21,344-50.

Forum on Medicine 348-9.

skills. In: The Znterpersml Lhmension in Medical Education (ed. by A Rezler & J Flaherty), pp. 10146). Springer, New York.

Hannay D R (1980) Teaching interviewing with simulated patients. Medical Education 14,246-8.

Joesbury H E, Bax N D S & Hannay D R (1990) Communication skills and clinical methods: a new introductory course. Medical Education 24,433-7.

Kazdin A E (1977) Artefact, bias and complexity of assessment: the ABCs of reliability.Journa1 ofApplied Behamoral Analysis 10,

Fitzgerald F T (1980) The clinical examination -a dying art?

Flaherty J A (1985) Education and evaluation of interpersonal

14 1-50. Keppel G (1982)Design and Analysis, 2nd edn. Prentice Hall,

Englewmd Cliffs, NJ. Knox J D & Bouchier J A (1985) Communication skills teaching,

learning and assessment. Medical Education 19,285-9. Maguire G P, Fairbairn S & Fletcher C M (1986) Benefits of

feedback training in interviewing as students persist. British MedicalJoumal 1,268-70.

Maguire P (1984) Communication skills and patient care. In: Health Care and Human Behaviour (ed. by A Steptoe & A Mathews), pp. 153-74. Academic Press, London.

Maguire P, Roe P, Goldberg D, Jones S, Hyde C & O’Dowd T (1978) The value of feedback in teaching interviewing skills to medical students. Postgraduate Medicine 8,695-704.

Meichenbaum D & Turk D C (1987) Faciliraring Patinu Adherence: a Practitioner’s Gurdebook. Plenum Press, New York.

Metcalfe D (1983) The mismatch between undergraduate education and the medical task. In: Doctor-Patient Communication (ed. by D Pendleton & J Hasler), pp. 227-32. Academic Press, London

Novack D H (1985) Beyond data gathering: twelve functions of the medical history. Hospital Practice 20,ll-12.

Preven D W, Kachur E K, Kupfer R B & Waters J A (1986) Interviewing skills of first year medical students.Journa1 of Medical Education 61,842-4.

Robbins A S, Kauss D R, Heinrich R, Abrass I, Dreyer J & Clyman B (1979) Interpersonal skills training: evaluation in an internal medicine residency. Journal of Medical Education 54,885-94.

Roter D L (1983) Physiciadpatient communication: transmission of information and patient effects. Maryland State Medical Journal 32,260-71.

communicating with patients be taught in medical schools? The Lancet 6,523-6.

Sanson-Fisher R W & Poole A (1979) Teaching medical students communication skills: an appraisal of the short and long-term benefits. In: Research in Psychologv and Medicine (ed. by D Oborne, M Gruneberg & J Eiser). Academic Press, London.

SPSS-X User’s Guide, 2nd edn (1986) SPSS Inc and McGraw-Hill, New York.

Stewart M & Roter D (1989) Communicating with Medicul Patients Sage, Newbury Park, CA.

Sanson-Fisher R W & Maguire P (1980) Should skills in

Consulting skills and interviewing efficiency B J E w n s et al.

128 MEDICAL EDUCATION 1996, 30, 121-128 0 1996 Blackwell Science Ltd

APPENDIX

Case histories

Patient with abdominal pain. Young female in late 20s-30s. Has had abdominal pain for 1 year, but now getting worse. Seems to occur any- where in the stomach - high up or low down - not severe usually but at times the pain becomes colicky and more severe - no vomiting - occasionally off food - good appetite usually.

Weight, if anything, has gone up a little. Often her stom- ach feels sore to touch.

Saw a doctor a few years ago who diagnosed gallstones, but tests were negative. Bowels tend to be all over the place - can be constipated for a few days -sometimes bowels loose and slime on them - rarely regular - but has lived with irregularity for years. Food intake a bit irregular but ade- quate - eats mean and vegetables at least once a day. Other meals often on the run. Never any blood in stools.

Sleeps fairly well. No medication. Pain not related to meals - NOT relieved by food - does not wake at night. When goes on holidays, pain may be easier but still pre- sent. Not relevant past or family history - non-smoker.

Patient with headache Young person in early 30s. Troublesome headaches for last few months -across front of head - also back of head and temples. Occurring now 3 4 times a week - under a lot of stress (young family, working, trying to make it, children upsetting, new baby last year).

Knows headaches are going to occur as feels ill. If person takes panadeine and lies down feels better. Headache can last most of the day - sick with the pain, never vomited - cannot concentrate - eyes go blurry and cannot see well - once over, pain feels OK again. Headache thumping feel- ing with tight head.

Has always had headaches from teen years onwards, but never really bad and not frequent - every few months - takes a painkiller and lies down. Cannot think what brings pain on - probably stress - feels that headaches can occur if drinks milk but not certain.

Parents divorced - does not see much of father - but remembers that father used to have severe headaches and was tense. Has been told that headaches due to tension, but not happy with that opinion.

Patient with angina Male in his 40s - smokes a packet of cigarettes a day. Work - not relevant. Family history - father died sud- denly in his early 60s about 15 years ago - had been smoker. Mother alive in her 70s but in nursing home with stroke.

General health and condition - not relevant. Does not see doctor regularly - last seen 7 years ago. If asked can state that doctor told him BP was up a little, but no need for treatment.

Patient has noticed ache in left upper arm when walking up hill to the station on way to work. Needs to slow up and ache settles - getting a little worse and not relieved now till he gets to the station and sits down - began to get this worse 1-2 months ago.

Gets a bit short of breath on exertion but puts this down to the cigarettes - always been a bit puffed for the last few years when he hurries, but not a problem. Also a bit short of breath and tight in chest when (L) arm ache comes on. Not wheeze - not S.O.B. at other times although he has some difficulty keeping up with mates when playing golf. First noticed ache in L arm about 6 months ago when mowing the lawn, but felt it was a sprained muscle from pushing the motor mower - not a great problem. Never had cholesterol checked. No medication.

Received 20June 1995; accepted fmpublicatwn 31 August 1995

Consulting skills and interviewing efficiency BJEvans et al.