Abstract
Background
Few studies have assessed the importance of a broad range of verbal and non-verbal consultation behaviours.
Aim
To explore the relationship of observer ratings of behaviours of videotaped consultations with patients’ perceptions.
Design and setting
Observational study in general practices close to Southampton, Southern England.
Method
Verbal and non-verbal behaviour was rated by independent observers blind to outcome. Patients competed the Medical Interview Satisfaction Scale (MISS; primary outcome) and questionnaires addressing other communication domains.
Results
In total, 275/360 consultations from 25 GPs had useable videotapes. Higher MISS scores were associated with slight forward lean (an 0.02 increase for each degree of lean, 95% confidence interval [CI] = 0.002 to 0.03), the number of gestures (0.08, 95% CI = 0.01 to 0.15), ‘back-channelling’ (for example, saying ‘mmm’) (0.11, 95% CI = 0.02 to 0.2), and social talk (0.29, 95% CI = 0.4 to 0.54). Starting the consultation with professional coolness (‘aloof’) was helpful and optimism unhelpful. Finishing with non-verbal ‘cut-offs’ (for example, looking away), being professionally cool (‘aloof’), or patronising, (‘infantilising’) resulted in poorer ratings. Physical contact was also important, but not traditional verbal communication.
Conclusion
These exploratory results require confirmation, but suggest that patients may be responding to several non-verbal behaviours and non-specific verbal behaviours, such as social talk and back-channelling, more than traditional verbal behaviours. A changing consultation dynamic may also help, from professional ‘coolness’ at the beginning of the consultation to becoming warmer and avoiding non-verbal cut-offs at the end.
Keywords: communication, consultation, general practice, non-verbal communication
INTRODUCTION
The patient-centred consultation model is widely advocated,1,2 although implementation is possibly limited,3–5 and poor communication results in complaints and lawsuits.6 Communication skills training is also probably very efficient given the long-lasting effects of training.7–10
A Cochrane Review of trials to modify patient-centredness documented mixed effects on satisfaction and small effects on health status.11 However, for satisfaction, none of the studies addressed non-verbal skills, and nearly all were intensive (‘brief’ training was up to 10 hours).11 Most models of patient-centred behaviour refer to traditional verbal skills, but a review of 22 observational studies suggested that other important factors were courtesy, empathy, positive reinforcement, reassurance and support, psychosocial talk, friendliness, humour, explanations, summarising and clarification, a direct body orientation, symmetrical legs and arms, forward lead, nodding, and gaze.12 An observational study identified important domains of patients’ perceptions — a communication and partnership approach, interest in the patient’s life, health promotion, a positive approach, and a personal relationship — each of which strongly predicts different outcomes.13,14 However, what mixture of verbal and non-verbal elements of doctor behaviour determines patients’ perceptions is less clear.
The previous literature also has significant limitations, great variability in what is rated (few include even the limited variables assessed by Beck et al),12 and outcomes. In one review of trials, meta-analysis was not possible due to the heterogeneity of interventions and outcomes,15 and in the Cochrane Review updated in 2012 heterogeneity was also moderately high.11
This study aimed to explore which aspects of GPs’ non-verbal and verbal communication are likely to be most important in determining patients’ satisfaction and perceptions of person-centred communication in the consultation.
METHOD
Unselected GP consultations for consecutive patients were videotaped and verbal and non-verbal behaviour was rated by independent observers blind to outcome. Patients completed ratings post-consultation questionnaires using the Medical Interview Satisfaction Scale (MISS) and other scales measuring the patient-centredness of the consultation.
Participants
Participants were adult patients, or children attending with their parents, for a new or ongoing problem, that is, not those attending simply for a repeat prescription. There were no exclusions apart from the inability to consent or complete questionnaires; for example, those experiencing severe distress, dementia, very severe depression, or who were very severely unwell.
How this fits in
Communication is central to every consultation and although a range of key elements have been advocated, few previous studies have assessed the impact of a wide range of verbal and non-verbal behaviours. This study suggests that clinicians may need to pay at least as much attention to non-verbal behaviours and non-specific verbal behaviours, such as social talk and back-channelling, as traditional verbal behaviours. They should avoid optimism but maintain professional ‘coolness’ at the beginning of the consultation, and end the consultation with warmth, avoiding non-verbal cut-offs.
Recruitment
Consecutive patients were recruited by 25 GPs in general practices close to Southampton in Southern England. The GPs in turn were recruited by the local postgraduate coordinator close to the Southampton postgraduate centre. GPs were asked to recruit up to 15 consecutive patients each.
Rating of videotapes
Doctors were videotaped with the video camera pointed towards the doctor. Prior findings demonstrated that short consultation ‘slices’ correlate well to total ratings; this was confirmed for most items, but where poor correlations were observed (for example, for Roter items — Roter Interaction Analysis System — such as social talk), the whole consultation was used. Thus, unless specified, the first and last minutes of the consultation were rated. The middle of the consultation was not chosen to avoid examinations performed away from the camera. Ratings were blind to the questionnaire responses by a research assistant or member of the research team, using a battery of assessments. Particular assessments were chosen based on previous significant findings. Further details of rating of videotapes and descriptive data for key variables are available from the authors.
Patient questionnaires
A similar methodology was used to the previous study:14 a post-consultation questionnaire explored the patient’s perception of communication. Questionnaires were completed by participants with help from parents as appropriate, particularly for very young children, either immediately after the consultation or, more commonly, at home and then posted back.
Primary outcome
The Medical Interview Satisfaction Scale (MISS)16 was chosen since its domains predominantly reflect communication and the doctor–patient relationship (distress–relief; communication–comfort; rapport; and compliance–intent). Patients agree or disagree on a 7-point Likert scale (very strongly agree to very strongly disagree) with items about how they rated the consultation.
Secondary outcomes
Similar 7-point Likert scales were also completed for several domains of patients’ perceptions of communication previously validated:14
communication and partnership;
a ‘personal’ relationship;
health promotion;
a ‘positive’ approach — being definite about the problem and when it would settle; and
interest in the effect on life.
The questionnaire also contained sociodemographic details, the short state anxiety questionnaire,17 the Whitley Index, the number of medical problems, current medication, enablement,18 symptom burden (Measure Yourself Medical Outcome Profile),19 whether seeing their usual doctor, the complaint (based on British National Formulary chapters), and whether or not this was ongoing.
Sample size
The sample size was calculated using an α of 0.01 and β of 0.2 using the NQuery sample size programme version 3. It was estimated that to detect a correlation of 0.25 between patient perception of communication and verbal or non-verbal communication rated in the consultation required 183 patients, or 229 allowing for 20% incomplete outcomes. Therefore 0.25 was chosen to allow for some margin of error: a correlation of 0.29 was observed between verbal behaviour and satisfaction in a similar previous UK study.20
Analysis
The data were analysed using SPSS (for data manipulation) and Stata (for modelling) statistical software for Windows™ using multiple linear regression, controlling for clustering by GP (Stata version 12; SPSS version 21). The linearity of associations was checked graphically and using ordered categorical variables. The dependent variables were the patients’ rating of the consultations and the independent variables were ratings of the physician’s behaviour (both verbal and non-verbal), as well as the rest of the data from the questionnaires.
Variables were selected manually by forward selection (to limit the development of spurious findings from automated procedures), and variables retained if there was limited inflation of standard errors, and if they were found to be significant (P<0.05) in multivariate analysis. All variables were then also checked manually in the final model, to ensure no variables that could be important had been overlooked (and which could have been spuriously thrown out by an automated procedure). Missing values were not imputed.
RESULTS
Patients were recruited from February until April each year from 2006 until 2010 by 25 GPs: 9/25 (36%) female; 6/25 (24%) not partners; and 5/25 (20%) working in deprived inner city areas.
Most patients could not be approached because of insufficient time to consent prior to the consultation. Of those who were approached, most agreed to participate (60%), with the remainder either not having the time or inclination to participate (35%), or because of the sensitive nature of the consultation (5%).
Of the 360 patients who initially agreed, 275 useable videotapes could be rated, and of these, 251/275 (91%) have useable questionnaire data, that is, a MISS questionnaire could be calculated.
The mean age of the index patients was 48 years, 138/215 (64%) were female, 145/223 (65%) were married, 13/197 (7%) were in receipt of sickness or disability benefit, 111/228 (49%) were in paid work, and had on average been to the doctor five times in the previous 12 months.
Patients rated both satisfaction and communication in the consultation highly: the mean item score for MISS on a 1–7 scale was 5.6 (standard deviation [SD] = 0.8), for the communication and partnership scale 5.5 (SD = 0.8), the personal relationship scale 5.1 (SD = 1.3), and the interest in life scale 5.2 (SD = 1.2).
Table 1 shows the results for the MISS questionnaire. Patients’ rating of satisfaction (mean item score on the MISS questionnaire, scaled 1–7) increased with slight lean towards the patient, the number of gestures, and ‘back-channel prompts’ (such as saying ‘mmm’, ‘ah ha’, and so on), at the beginning. Social talk at some point in the consultation was associated with increased satisfaction, and infantilising (being patronising) at the end of the consultation was associated with a negative impact.
Table 1.
Variables (scored from videotapes) significantly associated with the mean item score of the MISS questionnaire (rated by patients after the consultation; each item scored 1–7)
| Predictor variablesa | Univariate beta coefficientb (95% CI) | P-value | Multivariate beta coefficient (95% CI) | P-value |
|---|---|---|---|---|
| Non-verbal | ||||
| Number of gestures (beginning consultation) | 0.11 (0.02 to 0.19) | 0.018 | 0.08 (0.001 to 0.15) | 0.046 |
|
| ||||
| Degrees of lean towards the patient (beginning consultation) | 0.014 (−0.001 to 0.029) | 0.067 | 0.018 (0.002 to 0.03) | 0.025 |
|
| ||||
| Verbal | ||||
| Back-channel prompts (beginning consultation) | 0.10 (−0.02 to 0.23) | 0.100 | 0.11 (0.02 to 0.2) | 0.020 |
|
| ||||
| Infantilising (end consultation) | −0.39 (−0.67 to −0.11) | 0.009 | −0.31 (−0.86 to 0.25) | 0.044 |
|
| ||||
| Social talk occurred during whole consultation | 0.20 (−0.04 to 0.44) | 0.092 | 0.29 (0.4 to 0.54) | 0.026 |
Unless specified, predictor variables were those rated at the beginning of the consultation.
Univariate analysis (n = 243) controlled for clustering by doctor. Multivariate analysis (n = 191) controlled for clustering by doctor, deprivation, type of problem, marital status, attitude to doctors, time on the history, and for all significant predictors listed. In this and subsequent tables the estimate of the beta coefficients from the model are quoted; the interpretation of these, taking the first line as the example are that for each gesture used the mean item score for satisfaction increases by 0.11 in univariate analysis and 0.08 in multivariate analysis. MISS = Medical Interview Satisfaction Scale.
Tables 2–4 document the associations for the key communication domains. Information on the perception of health promotion, a positive approach, and enablement domains are shown in Appendices 1–3. Variables with estimates that were either significant in both univariate analysis and multivariate analysis, or significant in multivariate analysis but a consistent direction in both univariate and multivariate analysis, are highlighted in Table 5. The pattern of findings in Table 5 suggest being aloof, dominant, or infantilising at the end of the consultation, or using non-verbal cutoffs were associated with several domains of negative perceptions of communication by patients. The negative effect of being aloof or infantilising at the end of the consultation must be put in the context that for only a minority of consultations was there a negative rating: 23/268 (9%), and 32/268 (12%), respectively. Being physically engaged (using gestures or appropriate touch) and socially engaged (social talk) had positive effects. At the beginning of the consultation being supportive was helpful, but optimism was not. Conversely, being professionally aloof at the beginning of the consultation was helpful; suggesting a cool but supportive listening approach, without injecting artificial optimism at too early a stage, could be optimal.
Table 2.
Variables associated with patient rating of a communication and partnership approach
| Predictor variablesa | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non-verbal | ||||
| Number of gestures (beginning consultation) | 0.09 (0.01 to 0.17) | 0.030 | 0.09 (0.03 to 0.16) | 0.009 |
|
| ||||
| Physical contact occurred during whole consultation | 0.90 (0.15 to 1.65) | 0.021 | 1.59 (0.25 to 2.93) | 0.022 |
|
| ||||
| Non-verbal cut-off occurred (end consultation) | −1.92 (−2.64 to −1.20) | <0.001 | −1.82 (−2.91 to −0.72) | 0.002 |
|
| ||||
| Verbal | ||||
| Psychosocial talk occurred in the whole consultation | 0.10 (−0.02 to 0.23) | 0.100 | −0.20 (−0.39 to −0.01) | 0.040 |
|
| ||||
| Social talk occurred during whole consultation | 0.24 (−0.01 to 0.49) | 0.059 | 0.43 (0.14 to 0.73) | 0.006 |
Univariate (n = 239) controlling for clustering by doctor. Multivariate (n = 214) controlling for clustering by doctor, deprivation, type of problem, attitude to doctors, and for all significant predictors listed.
Table 3.
Variables associated with patient rating of a personal relationship
| Predictor variables | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non-verbal | ||||
| Non-verbal cut-off occurred (end consultation) | −3.25 (−5.08 to −1.42) | 0.001 | −3.02 (−4.77 to −1.27) | 0.002 |
|
| ||||
| Verbal | ||||
| Social talk occurred during whole consultation | 0.40 (−0.11 to 0.91) | 0.115 | 0.49 (0.16 to 0.83) | 0.006 |
|
| ||||
| Evidence that patient history not known during whole consultation | −1.26 (−1.78 to −0.73) | <0.001 | −0.70 (−1.23 to −0.17) | 0.012 |
|
| ||||
| Overall impression | ||||
| Optimistic (beginning consultation) | −0.02 (−0.17 to 0.14) | 0.809 | −0.17 (−0.27 to −0.06) | 0.003 |
|
| ||||
| Aloof (end consultation) | −0.40 (−0.62 to −0.19) | 0.001 | −0.23 (−0.46 to 0.00) | 0.05 |
|
| ||||
| Dominant (end consultation) | −0.05 (−0.15 to 0.04) | 0.267 | −0.14 (−0.20 to −0.07) | <0.001 |
Univariate (n = 242) controlling for clustering by doctor. Multivariate (n = 213) controlling for clustering by doctor, type of problem, attitude to doctors, time on history, and for all significant predictors listed.
Table 4.
Variables associated with patient rating of perceived interest of the doctor in their life
| Predictor variables | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non-verbal | ||||
| Non-verbal cut-off occurred (end consultation) | −1.69 (−2.62 to −0.76) | 0.001 | −2.42 (−4.60 to −0.25) | 0.030 |
|
| ||||
| Verbal | ||||
| Joke or laugh during the whole consultation | 0.37 (0.01 to 0.74) | 0.047 | 0.40 (0.05 to 0.76) | 0.027 |
|
| ||||
| Overall impression | ||||
| Supportive (beginning consultation) | 0.07 (−0.11 to 0.24) | 0.451 | 0.29 (0.17 to 0.41) | <0.001 |
|
| ||||
| Optimistic (beginning consultation) | −0.04 (−0.16 to 0.09) | 0.565 | −0.17 (−0.27 to −0.06) | 0.002 |
|
| ||||
| Aloof (end consultation) | −0.14 (−0.28 to −0.01) | 0.040 | −0.30 (−0.59 to −0.01) | 0.041 |
|
| ||||
| Aloof (beginning consultation) | 0.17 (0.09 to 0.26) | <0.001 | 0.33 (0.22 to 0.44) | <0.001 |
|
| ||||
| Infantilising (end consultation) | −0.57 (−1.12 to −0.01) | 0.046 | −0.61 (−1.12 to −0.10) | 0.020 |
|
| ||||
Univariate (n = 216) controlling for clustering by doctor. Multivariate (n = 211) controlling for clustering by doctor, type of problem, and for all significant predictors listed.
Table 5.
Summary of significant positive (beneficial outcome) and negative (adverse outcome) associations of key variables and whether the variable rating refers to the beginning, end, or whole consultation
| Predictor variables | Satisfaction (MISS) | Personal relationship | Interest in life | Health promotion | Positive approach | Communication/partnership | Enable |
|---|---|---|---|---|---|---|---|
| Optimistic (beginning) | Negative | Negative | |||||
| Not knowing patient (whole) | Negativea | ||||||
| Aloof (end) | Negativea | Negativea | Negative | Negative | |||
| Dominant (end) | Negative | Negative | |||||
| Infantilising (end) | Negativea | Negativea | Negative | ||||
| Psychosocial talk (whole) | Negative | Negative | |||||
| Mismatch rate/tone of speech (end) | Negative | ||||||
| Open questions (end) | Negative | ||||||
| Non-verbal cut-off (end) | Negativea | Negativea | Negativea | ||||
| Supportive (beginning) | Positive | ||||||
| Gestures (beginning) | Positivea | Positivea | |||||
| Physical contact (whole) | Positivea | Positivea | |||||
| Social talk (whole) | Positivea | Positivea | Positivea | Positivea | |||
| Joke/laugh (whole) | Positivea | ||||||
| Back-channelling (beginning) | Positive | ||||||
| Aloof (beginning) | Positivea | ||||||
| Lean (beginning) | Positive | Negative | |||||
| Definite about problem | Positive | Negative |
Variables with estimates that were either significant in both univariate analysis and multivariate analysis or significant in multivariate analysis but a consistent direction in both univariate and multivariate analysis. MISS = Medical Interview Satisfaction Scale.
DISCUSSION
Summary
This exploratory study is one of the largest to assess a range of verbal and non-verbal behaviours. It suggests that several non-verbal behaviours and non-specific verbal behaviours, such as social talk and back-channelling, may impact more than traditional verbal behaviours. It also proposes that a changing consultation dynamic may be important — from being professionally supportive but ‘cooler’ at the beginning to being warmer and avoiding non-verbal cut-offs at the end.
Strengths and limitations
This was one of the largest studies to assess a very broad range of verbal and non-verbal behaviours in primary care.
The main potential limitations of this study are confounding and type I errors. Type I error is less likely for variables that had multiple associations, and more likely for variables only associated with one outcome and only in multivariate analysis (for example, the rate of speech) or with mixed effects (for example, lean towards the patient and being definite about the nature of the problem). Confounding was controlled as necessary for age, sex of the doctor, type of problem, being in receipt of sickness and unemployment benefit, whether the patient was seeing their usual doctor, and attitudes to doctors.
Even with the range of variables controlled for it is difficult to deal with residual confounding, and there is also the danger of over-fitting. However variables, such as non-verbal cut-offs or being aloof or infantilising, which were significant and had similar estimates in both univariate and multivariate analysis (that is, stable estimates and less likely to be confounded), provide stronger evidence of causality.
Conversely, concerns about confounding are stronger when the estimates are unstable, hence variables with more consistent estimates were highlighted. Type II error (β) is also possible, even though this is one of the largest studies to assess such a complete range of variables (very large studies are challenging due to the very intensive nature of the video assessment). Reverse causality is likely to be relevant for some variables, for example, psychosocial talk (which is likely to occur in consultations with more distress and emotion expressed) was associated with negative perception of communication. Although Rosenthal et al have shown that naïve raters can use their scales reliably,21–23 judgements about each item will inevitably be subjective, although the impact of a single rater’s judgement was minimised by using several raters. Similarly, despite using the Stewart method which has an extensive manual,2 the estimation of reliability in this study suggests that for many variables reliability is only likely to be moderate, and so the associations may have been underestimated. Booking interval was not controlled for but total time in the consultations was not an important predictor.
These results should be confirmed in a wider group of GPs: although patients from GPs working in deprived areas had slightly higher MISS ratings and a ‘positive’ doctor approach, and partners had higher ratings than non-partners, these variables were controlled for in analysis. Selection bias potentially applies to both doctor and patient: Mead and Bower suggested that doctors who consider themselves to be good communicators, are thus likely to have better verbal communication skills and are more likely to take part, which will have potentially underestimated the importance of verbal skills.24 Patients with a sensitive nature to their problem/s are more reluctant to be filmed, and it is in such areas where patient-centredness is likely to be most important.24
Recruitment and rating of videotapes was slow due to the day-to-day running being coordinated each year by medical students as part of their research projects, and students could only provide concentrated input for 1–2 months per year. The requirement to use students in part reflects the difficulty of obtaining funding for this research. Furthermore, due to the logistic requirement to approach patients and obtain consent, a large proportion of consultations were not recorded.
Comparison with existing literature
Several variables in the current study were apparently important but only at the end of the consultation. The importance of avoiding non-verbal cut-offs supports Mehrabian,25,26 and suggests that non-verbal cut-offs give the patient the impression that the doctors communication skills are poor and that the doctor is not interested in their life. As might be expected, avoiding any sense of distance at the end of the consultation, such as aloof or infantilising/patronising,22,23 was powerful, although few doctor consultations in this study were judged to be very aloof or infantilising. The apparently surprising findings of some benefit from being ‘aloof’ and not overly optimistic at the beginning of the consultation suggests a changing dynamic throughout the consultation. This may reflect the importance early in the consultation of a cooler but supportive professional manner in helping patients feel listened to. The use of gestures and touch, that is, appropriate physical contact, are both supported by previous work,26,27 and were consistent between univariate and multivariate analysis.
The use of social conversations was important and is one of the central measures in the Roter interaction analysis system28,29 Where there was social conversation, patients were more likely to feel there was a communication and partnership approach, a personal relationship, and that the doctor was positive. Some of this may be reverse causality given that patients who know their doctor better are more likely to be engaged in social conversation and vice versa. However, controlling for whether this was the patient’s normal doctor did not modify the estimates, so reverse causality seems a less likely explanation. As expected demonstrating knowledge of the patient and their history was relevant for patients having a sense of a personal relationship, and this finding was also not affected by whether the doctor was the patient’s usual doctor. This highlights the importance when the GP is not the usual doctor of quickly checking the key elements of the patient’s past history.
Conventional approaches in conceptualising verbal aspects of patient-centredness — exploring the disease, understanding the person as a whole, and finding common ground — were not strong findings. This is closer to Mead and Bower who found very limited associations,24 but the correlations in Kinnersley et al’s study were also not large.20 This may reflect the limited range of GPs — studies of communication tend to enlist those already most interested in communication — but the range of scores for verbal communication perhaps makes this explanation less likely.
Implications for research and practice
Given the exploratory nature of this study, and the likely self-selection of GPs, very firm recommendations cannot be made. However, if these results can be confirmed, they suggest that health professionals should pay attention to non-verbal skills in the consultation, particularly the use of gestures and physical contact, and to non-specific verbal elements such as back-channel prompts and social conversation. A professionally cool approach may be helpful at the beginning of the consultation, but not at the end where a warmer approach and particular care to avoid non-verbal cut-offs are needed.
Acknowledgments
On behalf of the patient communication project team: Marinar Paramour, Anna Jackson, Olayiwola Olatawura, Becky Hawes, Emma Marston, Jennifer Walsh, Hannah Ebdon, Corrina Quinney and Fiona Hignett (all medical students at the time).
Appendix 1. Variables associated with patient rating of health promotion
| Predictor variables | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non verbal | ||||
| Non verbal cut–off occurred (end consultation) | −2.28 (−3.20 to −1.37) | <0.001 | −2.89 (−5.35 to −0.43) | 0.023 |
|
| ||||
| Object manipulation (end consultation) | −0.13 (−0.27 to 0.01) | 0.074 | −0.21 (−0.38 to −0.04) | 0.020 |
|
| ||||
| Computer use (beginning and end combined) | −0.34 (−0.70 to 0.02) | 0.066 | −0.59 (−1.02 to −0.16) | 0.010 |
|
| ||||
| Verbal | ||||
| Feedback examination (findings unsolicited) | 0.001 (−0.08 to 0.08) | 0.976 | 0.41 (0.13 to 0.68) | 0.006 |
|
| ||||
| Overall impression | ||||
| Aloof (beginning consultation) | 0.07 (−0.17 to 0.31) | 0.546 | 0.34 (0.21 to 0.48) | <0.001 |
|
| ||||
| Aloof (end consultation) | −0.04 (−0.36 to 0.29) | 0.816 | −0.28 (−0.49 to −0.07) | 0.011 |
|
| ||||
| Infantilising (end consultation) | −0.34 (−0.88 to 0.20) | 0.210 | −0.69 (−1.15 to −0.23) | 0.005 |
Univariate (n = 224) controlling for clustering by doctor. Multivariate (n = 195) controlling for clustering by doctor, type of problem, history time, total number of medical problems, and for all significant predictors listed.
Appendix 2. Variables associated with patient rating of the doctor being positive about the problem and its natural history
| Predictor variables | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non verbal | ||||
| Negative facial expression (end consultation) | −2.13 (−3.53 to −0.72) | 0.005 | −2.06 (−3.61 to −0.52) | 0.011 |
|
| ||||
| Lean towards the patient (end consultation) | −0.01 (−0.02 to 0.01) | 0.002 | −0.02 (−0.03 to −0.01) | 0.006 |
|
| ||||
| Physical contact occurred during whole consultation | 2.39 (0.53 to 4.23) | 0.014 | 2.52 (0.03 to 5.00) | 0.047 |
|
| ||||
| Negative head movement (beginning consultation) | 1.05 (0.58 to 1.52) | <0.001 | 1.22 (0.62 to 1.82) | <0.001 |
|
| ||||
| Verbal | ||||
| How definite about natural history | 0.12 (−0.01 to 0.25) | 0.077 | 0.13 (0.02 to 0.23) | 0.021 |
|
| ||||
| Social talk during whole consultation | 0.28 (−0.14 to 0.70) | 0.177 | 0.62 (0.10 to 1.13) | 0.021 |
|
| ||||
| Psychosocial talk during whole consultation | −0.20 (−0.48 to 0.07) | 0.135 | −0.55 (−0.88 to −0.22) | 0.002 |
|
| ||||
| Open questions (end consultation) | −1.23 (−2.69 to 0.23) | 0.095 | −1.35 (−2.65 to −0.04) | 0.044 |
|
| ||||
| Overall impression | ||||
| Aloof (beginning consultation) | 0.01 (−0.28 to 0.29) | 0.968 | 0.17 (0.05 to 0.29) | 0.007 |
|
| ||||
| Aloof (end consultation) | −0.18 (−0.64 to 0.29) | 0.437 | −0.34 (−0.64 to −0.04) | 0.026 |
Univariate controlling for clustering by doctor. Multivariate controlling for clustering by doctor, type of problem, deprived practice, state anxiety and for all significant predictors listed.
Appendix 3. Variables associated with patient rating of enablement
| Predictor variables | Univariate | P-value | Multivariate | P-value |
|---|---|---|---|---|
| Non verbal | ||||
| Lean (beginning consultation) | −0.01 (−0.02 to 0.00) | 0.041 | −0.01 (−0.02 to 0.00) | 0.045 |
|
| ||||
| Match of rate of speech (end consultation) | −1.08 (−1.94 to − 0.21) | 0.017 | −1.92 (−3.22 to −0.62) | 0.006 |
|
| ||||
| Match of patient and doctor tone (beginning consultation) | −0.19 (−0.53 to 0.15) | 0.262 | −0.35 (−0.52 to 0.17) | 0.001 |
|
| ||||
| Verbal | ||||
| Feedback examination | −0.01 (−0.05 to 0.02) | 0.463 | −0.04 (−0.07 to −0.01) | 0.013 |
|
| ||||
| Definite about the problem | −0.03 (−0.06 to −0.01) | 0.167 | −0.05 (−0.09 to −0.02) | 0.005 |
|
| ||||
| Overall impression | ||||
| Hostile (beginning consultation) | −0.62 (−1.00 to −0.24) | 0.003 | −0.39 (−0.59 to −0.18) | 0.001 |
Univariate (n = 234) controlling for clustering by doctor. Multivariate (n = 206) controlling for clustering by doctor, type of problem, deprived practice, attitude to doctors and for all significant predictors listed.
Funding
We are grateful to the Scientific Foundation of the RCGP and the NIHR South West Regional R+D panel for partly funding this work. (Reference number SFB 2003/44).
Ethical approval
The study was approved by the Salisbury and South East Hampshire local research ethics committees (Southampton Local Research Ethics Committee number: 230/97).
Provenance
Freely submitted; externally peer reviewed.
Competing interests
The authors have declared no competing interests.
Discuss this article
Contribute and read comments about this article: bjgp.org/letters
REFERENCES
- 1.Stewart MA. Effective physician-patient communication and health outcomes: a review. CMAJ. 1995;152(9):1423–1433. [PMC free article] [PubMed] [Google Scholar]
- 2.Brown JB, Stewart M, Tessier S. Assessing communication between patients and doctors: a manual for scoring patient-centred communication. Working Paper Series 95-2. London: Centre for Studies in Family Medicine and Thames Valley Family Practice Research Unit; 1995. [Google Scholar]
- 3.Barry CA, Bradley CP, Britten N, et al. Patients’ unvoiced agendas in general practice consultations: qualitative study. BMJ. 2000;320(7244):1246–1250. doi: 10.1136/bmj.320.7244.1246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Britten N, Stevenson FA, Barry CA, et al. Misunderstandings in prescribing decisions in general practice: qualitative study. BMJ. 2000;320(7233):484–488. doi: 10.1136/bmj.320.7233.484. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Law SA, Britten N. Factors that influence the patient centredness of a consultation. Br J Gen Pract. 1995;45(399):520–524. [PMC free article] [PubMed] [Google Scholar]
- 6.Meryn S. Improving doctor–patient communication. Not an option, but a necessity. BMJ. 1998;316(7149):1922–1930. doi: 10.1136/bmj.316.7149.1922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Maguire P, Fairbairn S, Fletcher C. Consultation skills of young doctors: Benefits of feedback training in interviewing as students persist. Br Med J (Clin Res Ed) 1986;292(6535):1573–1576. doi: 10.1136/bmj.292.6535.1573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Little P, Stuart B, Moore M, et al. Amoxicillin for acute lower-respiratory-tract infection in primary care when pneumonia is not suspected: a 12-country, randomised, placebo-controlled trial. Lancet Infect Dis. 2013;13(2):123–129. doi: 10.1016/S1473-3099(12)70300-6. [DOI] [PubMed] [Google Scholar]
- 9.Butler CC, Simpson SA, Dunstan F, et al. Effectiveness of multifaceted educational programme to reduce antibiotic dispensing in primary care: practice based randomised controlled trial. BMJ. 2012 doi: 10.1136/bmj.d8173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cals JW, de Bock L, Beckers PJ, et al. Enhanced communication skills and C-reactive protein point-of-care testing for respiratory tract infection: 3.5-year follow-up of a cluster randomized trial. Ann Fam Med. 2013;11(2):157–164. doi: 10.1370/afm.1477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dwamena F, Holmes-Rovner M, Gaulden CM, et al. Interventions for providers to promote a patient-centred approach in clinical consultations. Cochrane Database Syst Rev. 2012;12(12):CD003267. doi: 10.1002/14651858.CD003267.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Beck RS, Daughtridge R, Sloane PD. Physician-patient communication in the primary care office: a systematic review. J Am Board Fam Pract. 2002;15(1):25–38. [PubMed] [Google Scholar]
- 13.Little P, Everitt H, Williamson I, et al. Preferences of patients for patient centred approach to consultation in primary care: observational study. BMJ. 2001;322(7284):468–472. doi: 10.1136/bmj.322.7284.468. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Little P, Everitt H, Williamson I, et al. Obervational study of effect of patient centredness and positive approach on outcomes of general practice consultations. BMJ. 2001;323(7318):908–911. doi: 10.1136/bmj.323.7318.908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Griffin SJ, Kinmonth AL, Veltman MW, et al. Effect on health-related outcomes of interventions to alter the interaction between patients and practitioners: a systematic review of trials. Ann Fam Med. 2004;2(6):595–608. doi: 10.1370/afm.142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hallam D, Doggett L, Wilkin MA. Measures of need and outcome for primary health care. Oxford: Oxford University Press; 1992. [Google Scholar]
- 17.Marteau TM, Bekker H. The development of a six-item short-form of the state scale of the Spielberger State-Trait Anxiety Inventory (STAI) Br J Clin Psychol. 1992;31(3):301–306. doi: 10.1111/j.2044-8260.1992.tb00997.x. [DOI] [PubMed] [Google Scholar]
- 18.Howie JG, Heaney DJ, Maxwell M, et al. Quality at general practice consultations: cross sectional survey. BMJ. 1999;319(7212):738–743. doi: 10.1136/bmj.319.7212.738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Paterson C. Measuring outcomes in primary care: a patient generated measure, MYMOP, compared with the SF-36 health survey. BMJ. 1996;312(7037):1016–1020. doi: 10.1136/bmj.312.7037.1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kinnersley P, Stott N, Peters TJ, Harvey I. The patient-centredness of consultations and outcome in primary care. Br J Gen Pract. 1999;49(446):711–716. [PMC free article] [PubMed] [Google Scholar]
- 21.Ambady N, Laplante D, Nguyen T, et al. Surgeons’ tone of voice: a clue to malpractice history. Surgery. 2002;132(1):5–9. doi: 10.1067/msy.2002.124733. [DOI] [PubMed] [Google Scholar]
- 22.Ambady N, Koo J, Rosenthal R, Winograd CH. Physical therapists’ nonverbal communication predicts geriatric patients’ health outcomes. Psychol Aging. 2002;17(3):443–452. doi: 10.1037/0882-7974.17.3.443. [DOI] [PubMed] [Google Scholar]
- 23.Ambady N, Rosenthal R. Thin slices of expressive behavior as predictors of interpersonal consequences: a meta-analysis. Psychological Bulletin. 1992;111(2):256–274. [Google Scholar]
- 24.Mead N, Bower P. Measuring patient-centredness: a comparison of three observation-based instruments. Patient Educ Couns. 2000;39(1):71–80. doi: 10.1016/s0738-3991(99)00092-0. [DOI] [PubMed] [Google Scholar]
- 25.Mehrabian A. Nonverbal communication. Chicago, IL: Aldine Atherton Inc.; 1972. [Google Scholar]
- 26.Mehrabian A. Some referents and measures of nonverbal behavior. Behavior Res Meth Instrument. 1968;1:203–207. [Google Scholar]
- 27.Gleeson M, Timmins F. A review of the use and clinical effectiveness of touch as a nursing intervention. Clin Eff Nurs. 2005;9:69–77. [Google Scholar]
- 28.Roter D, Larson S. The Roter interaction analysis system (RIAS): utility and flexibility for analysis of medical interactions. Patient Educ Couns. 2002;46(4):243–251. doi: 10.1016/s0738-3991(02)00012-5. [DOI] [PubMed] [Google Scholar]
- 29.Sandvik M, Eide H, Lind M, et al. Analyzing medical dialogues: strength and weakness of Roter’s interaction analysis system (RIAS) Patient Educ Couns. 2002;46(4):235–241. doi: 10.1016/s0738-3991(02)00014-9. [DOI] [PubMed] [Google Scholar]