A quantitative analysis of anaesthetist–patient communication during the pre-operative visit
Summary
Previous communication research in general medical practice has shown that effective communication enhances patient compliance, satisfaction and medical outcome. It is expected that communication is equally important in anaesthesia, since patients often suffer from anxiety and lack of knowledge about anaesthetic procedures. However, little is known about the nature of communication during routine anaesthetic visits. The present study of 57 authentic anaesthetic visits provides the first results on the structure and content of communication in the pre-operative setting using the Roter Interaction Analysis System (RIAS). Patient-centred communication behaviours of anaesthetists and the extent of patient involvement were particularly investigated. From the 57 pre-operative visits, 18 267 utterances were coded. The mean (SD) [range] duration of the visit was 16.1 (7.8) [3.7–42.7] min. Anaesthetists provided 169 (68) and patients 153 (82) utterances per visit (53.5% vs. 46.5%). Physician and patient gender had no impact on the distribution of utterances and the duration of the visit. Conversation mainly focussed on biomedical issues with little psychosocial discussion (< 0.1% of all anaesthetist utterances). However, anaesthetists quite frequently used emotional comments toward patients (7%) and involved them in the conversation. The use of facilitators, open questions and emotional statements by the anaesthetist correlated with high patient involvement. The amount of patient participation in anaesthetic decisions was assessed with the Observing Patient Involvement Scale (OPTION). Compared with general practitioners, anaesthetists offered more opportunities to discuss treatment options (mean (SD) OPTION score 26.8 (16.8) vs. 16.8 (7.7)).
Effective physician–patient communication is increasingly recognised to influence a variety of patient behaviours such as patient satisfaction and understanding of medical advice. Good communication also contributes to achieving desired health care outcomes [1]. In a recent attempt to define and assess the professional competence of physicians, Epstein & Hundert [2] established basic communication skills as an important dimension of professional cognitive competence. The Accreditation Council for Graduate Medical Education in the United States specifies interpersonal skills and communication as general competencies of graduate medical education and the Council's Outcome Project documents the assessment of such competencies by American residency programs [3]. The same is true for The Royal College of Anaesthetists and The Royal College of Surgeons, both of which now require an assessment of communication skills for their certificates of specialist training [4,5]. Historically, anaesthetic training was primarily directed at acquiring technical skills. Most of today's attending anaesthetists received little formal training in communication and might lack the skills required to deal effectively with patients' psychosocial and information needs during the pre-operative visit. It is therefore not surprising that communication skills are increasingly viewed as a key component both in the medical curriculum and in the postgraduate training for specialists [6]. In the past, communication skills training as a component of formal training was only emphasised for general practitioners and internists, but the importance of such non-technical skills is now also recognised in anaesthesia and critical care medicine [7–12]. For this purpose, simulator-based modular human factor and communication training has recently been made available to anaesthetists [13].
Since pre-operative anxiety is often directed towards the anaesthetic itself [14], an anaesthetist is best able to decrease a patient's anxiety [15]. Our aim is to improve the anaesthetist's ability to elicit and respond to the patient's concerns and questions during the pre-operative anaesthetic visit. In order to define more precisely those skills that would best help an anaesthetist to become a more professional communicator, the current study investigates the anaesthetist–patient interaction in 57 videotaped real-life pre-operative visits using the standard instrument Roter Interaction Analysis System (RIAS) [16]. In this system, each statement (utterance) by either physician or patient is coded into mutually exclusive physician or patient categories. While RIAS has been extensively used and validated in general practice [17], internal medicine [18], oncology [19,20] and gynaecology [21] (for a complete list of the RIAS bibliography see [22]), it has rarely been used in the peri-operative period [23]. This study presents a detailed description and analysis of the routine pre-operative communication between anaesthetists and patients. In a patient-centred communication, physicians listen attentively to patients and show empathy and understanding for their concerns. We therefore also investigated which anaesthetist communication behaviours may correlate with increased patient involvement.
There is an increasing interest in forming partnerships between physicians and patients, especially with respect to the decision-making process [24]. As the interaction between anaesthetist and patient is characterised by an asymmetric distribution of expertise, the task of conducting a patient-centred interview that enables shared decision-making seems especially difficult. However, the availability of various anaesthetic techniques offers a good opportunity for patient participation in anaesthesia-specific decisions. In order to assess the practice of shared decision-making, the 57 anaesthetic visits were also coded with a novel instrument, the Observing Patient Involvement Scale (OPTION), which is designed to measure patient participation [25].
Methods
This study used data from a larger research project aimed at improving the communication skills of anaesthetists [9]. The Research Ethics Committee of the University of Basel, Switzerland, approved the analysis of videotaped pre-operative visits. Adult patients scheduled for elective operations were recruited by anaesthetists from our department and participated voluntarily. Fifty-seven real-life pre-operative anaesthetic visits were videotaped and coded by two trained raters according to RIAS [16] and OPTION [25]. RIAS consists of a detailed category system coding each verbal utterance (statement) of physician and patient with respect to both task-related (instrumental) and socio-emotional (affective exchange) behaviours [16]. The unit of analysis consists of each verbal utterance or discernible segment of speech. The length of a single unit can vary from single words (e.g. physician: ‘…mhm…I understand…’ or ‘…interesting…’, coded as ‘facilitators’) to lengthy sentences conveying a single thought (e.g. physician: ‘…I understand why you’re worried…', coded as ‘legitimise/empathy’). The original RIAS specifies 24 coding categories for the task-focused exchange (e.g. counselling, giving biomedical information) and 15 coding categories for the socio-emotional exchange (e.g. emotional responsiveness, empathy). For a complete list of the original RIAS categories see the RIAS manual at http://www.rias.org/manual (accessed 19/08/2004). While most of these categories apply for both physician and patient, some are exclusively for patients (e.g. ‘request for services or medication’) or for physicians (e.g. ‘counsels or directs therapeutic regimen’). Adaptations can be made, to some extent, to the RIAS coding scheme to accommodate the specific needs of a study. For example, the two separate categories ‘legitimise’ and ‘empathy’ are often condensed into one category (‘legitimise/empathy’) since both are very similar and often indiscernible constructs of socio-emotional talk. Alternatively, a single RIAS category might be broken down into subcategories better to reflect the communicative characteristics of specific types of medical encounters. Whereas many communication studies have used condensed versions of RIAS, in our study we used the original RIAS category system. The analysis consisted of 38 physician and 28 patient categories. In addition, we added five anaesthesia-specific items as subcategories of counselling. They were identified in a modified Delphi procedure and specify aspects of peri-operative counselling of patients by anaesthetists (Table 1).
| ‘C—Anae/Alt’ | The anaesthetist describes the planned anaesthetic technique with possible alternatives. Pros and cons of each technique are discussed. If present, the possibility of shared decision-making is clearly expressed. |
| ‘C—Anae/Pre’ | The anaesthetist gives clear pre-operative instructions, e.g. sleeping pill, nil by mouth, own medication, additional sedative medication. |
| ‘C—Anae/Pro’ | The anaesthetist describes the complete anaesthetic procedure, e.g. transport to the operating room, application of routine monitors, insertion of an intravenous cannula, description of regional anaesthesia, sedation, induction of anaesthesia. |
| ‘C—Anae/Post’ | The anaesthetist describes the postoperative procedure and the planned treatment of pain with the respective alternatives. Pros and cons of each technique are discussed. Postoperative surveillance in the recovery room or intensive care unit is also covered. |
| ‘C—Anae/Segm’ | The anaesthetist mentions the possibility of segmented anaesthetic care, i.e. another anaesthetist will perform the anaesthesia on the day of the operation. |
The two raters were trained by an experienced researcher in the field of communication analysis (W.L.) using the RIAS coding manual and training tapes. Coding was performed directly on digitised videotapes without transcripts using the official RIAS computer-entry software designed to automate scoring and entry of RIAS data. For interpretation and comparison of the results with other studies, various RIAS categories were combined into larger clusters as previously described [18,26]; for example, open and closed questions were summarised disregarding their content (either medical, therapeutic, psychosocial or lifestyle). Similarly, all utterances expressing approval and agreement between physician and patient and so-called back-channel answers (i.e. indicators of sustained interest, attentive listening or encouragement emitted by the physician when he or she does not hold the speaking floor, such as ‘yeah’ or ‘mhm…right’) were taken together as facilitators, which aim to support the process of the conversation. The frequencies for each of the coded communication behaviour were analysed by standard statistical tests and the strength of association between two variables was assessed by the Pearson product-moment correlation coefficient (r) using SPSS software (version 11.5, SPSS Inc., Chicago, IL).
The original OPTION scale measures patient participation in medical decisions with 12 statements concerning physician behaviour on a 5-point Likert scale (0 = the behaviour is not observed to 4 = the behaviour is exhibited to a very high degree); the final OPTION score is then presented on a 0–100 scale [25]. For the purpose of this study, items number 3 and 12 (‘the clinician assesses the patient's preferred approach to receiving information to assist decision-making' and ‘the clinician indicates the need to review the decision’) were considered as unsuitable for the pre-operative anaesthetic visit and were therefore excluded. The two raters were trained to use OPTION during a 1-day workshop by the designer of this tool using the OPTION manual and training videotapes.
To assess interrater reliability, eight random tapes were double-coded and Pearson correlation coefficients (r) were calculated for each of the RIAS communication categories. The interrater reliability for both physician communication (mean r = 0.77, range 0.61–0.92) and patient communication (mean r = 0.91, range 0.85–0.97) proved to be high and in the range of previous research using RIAS [26,27]. The average Pearson correlation coefficient for the interrater reliability of the OPTION subscales was r = 0.88 (range 0.68–0.95).
Results
Twelve female and 45 male anaesthetists and 26 male and 31 female patients participated in the study. During the 57 pre-operative visits a total of 18 267 utterances were coded. The mean (SD) number of physician and patient utterances was 169 (68) and 153 (82) per visit, equalling 53.5% and 46.5%, respectively. Physician and patient gender had no significant impact on the distribution of utterances and on the duration of the visit. The mean duration of the visit was 16.1 SD [range] (7.8) [3.7–42.7] min. The relative frequencies of the RIAS communication categories are shown in Table 2. Physicians devote the majority of their statements to asking questions (26%) and counselling (23%). On average, six of 41 physician questions per interview were classified as open questions. Patients correspondingly give a lot of information (56% of all patient statements) and engage in psychosocial conversation (15%).
| RIAS communication categories | Utterances per anaesthetic visit | ||
|---|---|---|---|
| Mean | SD | % | |
| Content | |||
| Physician | 102.8 | 37.4 | 64.4 |
| Asks questions | 41.3 | 19.7 | 26.0 |
| Open questions | 5.6 | 5.1 | 3.4 |
| Gives information – medical/ therapy/lifestyle | 27.5 | 22.5 | 15.2 |
| Counsels | 33.3 | 19.3 | 23.0 |
| Anaesthetic technique | 9.6 | 16.5 | 6.6 |
| Pre-operative instructions | 5.0 | 3.3 | 3.4 |
| Anaesthetic procedure | 14.4 | 8.9 | 9.9 |
| Postoperative procedure | 3.3 | 5.1 | 2.3 |
| Segmented anaesthetic care | 1.2 | 2.1 | 0.9 |
| Psychosocial discussion (statements) | 0.5 | 1.1 | < 0.1 |
| Patient | 117.4 | 68.7 | 75.8 |
| Asks questions | 6.2 | 5.7 | 4.2 |
| Gives information – medical/ therapy/lifestyle | 83.2 | 44.9 | 56.2 |
| Psychosocial discussion (statements) | 28.0 | 32.4 | 15.4 |
| Process | |||
| Physician | 25.8 | 16.1 | 15.5 |
| Facilitators (agree, approve, back-channel, check) | 17.4 | 12.3 | 10.4 |
| Orientation | 8.4 | 6.2 | 5.1 |
| Emotional affect | |||
| Physician | 12.1 | 8.9 | 6.7 |
| Legitimise/Empathy | 1.0 | 1.7 | 0.5 |
| Reassurance/Optimism | 11.1 | 8.1 | 6.2 |
| Patient | 2.7 | 2.7 | 2 |
| Concern | 0.4 | 0.7 | 0.4 |
| Asks for reassurance | 2.3 | 2.6 | 1.6 |
- The total of patient and physician utterances does not equal 100% because rare RIAS categories such as ‘personal’, ‘laughs’, ‘compliment’, ‘disapprove’, ‘critic’, ‘concern’, ‘partnership’, ‘self-disclosure’, ‘transition word’, ‘bid for repetition’, and ‘back-channel (patient)’ are not depicted in the Table.
Given that the main focus of the pre-operative visit is related to information exchange, the extent to which physicians engage in communicative behaviours that are assumed to enhance patient involvement was investigated next. Physicians devote 26 (15%) statements to facilitate patient input (10% facilitators and 5% orientation statements). The physician can increase patient involvement not only by asking open questions or using facilitators, but also by expressing affective comments. In the present study, physicians on average made 12 (7%) affective statements per visit. This category summarises statements of ‘empathy’ in an explicit sense (e.g. ‘I can very well understand that you are afraid of the operation’; 1.0 (1.7) (0.5%) per interview) and statements of ‘reassurance/optimism’ (e.g. ‘You need not worry, the anaesthetist is usually an expert in spinal puncture; it won’t hurt'; 11.1 (8.1) (6%) per interview) (Table 2). In general, communicative anaesthetist behaviour such as using facilitators, asking open questions or expressing affective comments correlated well with patient involvement expressed as number of patient statements, questions, expressions of concern and amount of psychosocial discussion (Table 3).
| Patient involvement | Anaesthetist communication behaviour | ||
|---|---|---|---|
| No. of facilitators and orientation statements | No. of open questions | No. of affective statements (‘legitimise / empathy’, ‘reassurance / optimism’) | |
| No. of patient statements | 0.672** | 0.582** | 0.191 |
| No. of questions asked by patients | 0.481** | 0.368** | 0.347** |
| No. of affective statements by patients (‘concern’, ‘asks for reassurance’) | 0.147 | 0.012 | 0.299* |
| Psychosocial discussion (statements) | 0.434** | 0.288* | 0.037 |
- * p < 0.05;
- ** p < 0.01 (Pearson correlation coefficients r).
An alternative way to facilitate patient involvement consists of adhering to a clear communication structure during the interview. According to RIAS, a clinical visit consists of opening, history-taking, physical exam, counselling and closing segments. In 25 of the 57 (44%) anaesthetic visits, this consultation structure was followed. However, seven interviews showed a permanent alternation between the segments history-taking and counselling.
In 21 of the 57 anaesthetic visits, a clear decision-making situation was observed and the OPTION scale was used to measure patient participation in anaesthetic management decisions. Anaesthetists on average achieved a final score of 26.8 (16.8) on the modified OPTION scale. The detailed results for the OPTION subscales are listed in Table 4.
| Questions used in OPTION | Scores per OPTION question | |||
|---|---|---|---|---|
| Median | Range | Mean | SD | |
| Instrumental questions/talk | ||||
| Problem identification | 2 | 0–4 | 1.7 | 1.2 |
| Equipoise | 1 | 0–4 | 1.2 | 1.1 |
| Lists options | 1 | 0–3 | 1.6 | 1.0 |
| Lists pros and cons | 1 | 0–4 | 1.6 | 1.4 |
| Indicates need for a decision-making | 1 | 0–3 | 1.2 | 1.2 |
| Interpersonal questions/talk | ||||
| Explores the patient's expectations | 1 | 0–4 | 1.7 | 1.3 |
| Explores the patient's concerns | 0 | 0–4 | 0.9 | 1.2 |
| Checks the patient's understanding | 0 | 0–2 | 0.1 | 0.5 |
| Offers explicit opportunities to ask questions | 0 | 0–3 | 0.6 | 1.0 |
| Elicits the patient's preferred level of involvement | 0 | 0–1 | 0.1 | 1.2 |
- On average, anaesthetists scored overall 10.7 ± 6.7 out of a maximum of 40 points. This equals a final OPTION score of 26.8 ± 16.8 on the modified 0–100 OPTION scale. For further details of the OPTION scale see reference [25].
In general, the mean score of anaesthetists is lower for items concerning interpersonal talk than for items addressing instrumental questions. In 19 of the 21 visits coded with OPTION, the anaesthetist listed various alternatives for the anaesthetic technique; however, in only two visits did the anaesthetist also control the patient's understanding. Excellent behaviour (score 4) was observed three times for the items ‘lists pros and cons’ and ‘explores the patient's expectations'. For the items ‘problem identification’, ‘equipoise’ and ‘explores the patient's concerns' excellent behaviour was shown only once.
Discussion
In this study, we report descriptive and quantitative data of the anaesthetist–patient communication during the pre-operative visit. More than 60% of all physician statements were concerned with receiving and giving information and approximately 15% of physician statements were attributed to categories that support the process of information transfer; 7% conveyed an emotional affect and < 0.1% were psychosocial statements (Table 2). As there are no other reports from the anaesthetic literature covering this topic in a comparable way, our results cannot answer the question of whether the observed anaesthetist communication behaviour is good clinical practice. However, our results might serve as a starting point with which future studies could be compared.
For specific issues, some data from other specialities are available. Levinson et al. [23] reported that general and orthopaedic surgeons talked substantially more than their patients (60 : 40%). In this regard, anaesthetists were slightly less dominating than surgeons. Taken that the total number of utterances by the physician or patient (53.5% vs. 46.5%) reflects the dominance of one interlocutor, then pre-operative visits were not dominated by either party. However, dominance is not necessarily or exclusively linked to the share one party has in the total number of statements. If, for example, the surgeons' duty was to confer information, a 60% share of conversation might be justified. Anaesthetists face similar requirements during pre-operative visits; they need to explain different anaesthetic procedures and communicate associated risks and benefits. In light of this task, an almost equal share of utterances seems a positive result indicating some equipoise. Interestingly, both general and orthopaedic surgeons showed a greater proportion of emotional (10%) and social (4–7%) conversation than anaesthetists [23]. This finding possibly reflects the focussed and rather technical nature of pre-operative anaesthetic visits as practised today, whereas surgeons might be more inclined to explore a patient's diagnosis and social background. However, 15% of patient statements during pre-operative anaesthetic visits were concerned with psychosocial topics, which underscored the patient's wish to discuss these.
The number of questions patients ask during a visit has recently been taken as another indicator of the opportunity for patients to be involved in the process of the interview [28,29]. In the present study patients asked on average six questions; this compares favourably with the 3–5 questions patients asked during primary medical care visits [30,31]; in contrast, Kidd et al. [28] reported that patients in a specialised diabetic clinic asked 8–10 questions per consultation. However, in the case of an ‘ideal’ physician who is very sensitive in adjusting information to an individual patient's needs, the patient may not have to ask many questions to receive exactly the information he or she wants. This assumption was supported by our finding that the number of questions patients asked was negatively correlated with the amount of counselling provided by the anaesthetists (r = −0.22, p < 0.001).
Since the majority of anaesthetist–patient communication is about receiving or imparting information, the observed amount of empathic statements (12 anaesthetist statements per interview were attributed to ‘legitimise/empathy’ or ‘reassurance/optimism’) represents a high proportion of non-instrumental communication; in gynaecological interviews (personal communication, S. van Dulmen, June 2004) and visits in internal medicine only 2–3 physician utterances were classified as empathic statements [18]. One can argue whether empathy is truly reflected by the frequency of ‘reassurance/optimism’ statements. The current analysis does not differentiate between statements that responded specifically to a patient's concern and those that were made spontaneously as possibly premature consolation and which may or may not relieve the patient's concerns. Further studies will have to take this methodological problem into account by analysing the sequence of patient and physician statements in more detail following the methodology suggested by Langewitz et al. [18]. In the present study the degree of patient involvement in the anaesthetist–patient conversation was assessed with different indicators, namely the number of patient statements and questions, the extent to which patients addressed concerns or asked for reassurance, and the amount of psychosocial discussion. The use of facilitators, open questions and emotional statements by the anaesthetist coincided with high patient involvement (Table 3).
In addition to the above mentioned limitations of the analysis of empathic conversation, pure information giving presents its own problems. Patients do not always recall what they were told and what they do remember is not always what the physician believes to be the most important information [32]. As the total amount of information given during a pre-operative visit is far beyond the number of new items human beings can remember [33], we investigated whether information was at least given in a structured manner. In 25 of 57 videos, a typical physician–patient communication structure as outlined in RIAS was followed, although in many, no explicit opening or closing segments were observed, which is perhaps attributable to the anaesthetist operating the video camera.
Involving patients in decision-making has recently become an important task [24]. There is a shift away from decisions made solely by physicians, reflecting an increased respect for the autonomy of the patient. Three models of physician–patient interaction have been described: paternalism, shared decision-making and informed choice [34]. In a paternalistic approach, the physician takes the responsibility for decision-making; such behaviour is widely observed for anaesthetists (‘I will stick a small needle in your back, it will not hurt…’). Informed choice is at the other end of the spectrum; the patient takes responsibility for the decision after being provided with adequate information. The OPTION (Observing Patient Involvement) scale provides reliable and validated scores for detecting differences in consultations concerning the extent to which patients participated in the decision-making process [25]. Mean (SD) anaesthetist scores on a slightly modified scale of possible patient participation (26.8 (16.8)) were higher than previously reported scores of general practitioners (16.8 (7.7)) [25]. However, in the present study of anaesthetic visits, two of the original OPTION items were excluded, which may have contributed to the difference in the results despite presenting the scores on a percent scale from 0 to 100. It must also be noted that there are limits to patient choice in the field of anaesthesia. Offering choices to patients that the anaesthetist is unfamiliar with might interfere with patient safety or increase patient anxiety [35–37].
In summary, conversation during the pre-operative anaesthetic visit mainly focussed on biomedical issues, with little psychosocial discussion by the anaesthetist. Although the conversation under study served a specific purpose, namely the preparation for an operation and anaesthesia, anaesthetists and patients were at a surprisingly high level of equipoise. Anaesthetists also quite frequently expressed empathy toward patients and involved them in the decision-making process. In addition, anaesthetist communicative behaviours that were positively related with patient involvement could be identified. This research can serve as a basis for future studies, which will need to explore the relationship between anaesthetist–patient communication and medical outcomes and evaluate communication interventions in the field of anaesthesia. Since information giving and counselling by the anaesthetist represent substantial parts of the pre-operative visit, communication skills training in anaesthesia should primarily focus on how to improve the methods used to present important pre-operative information to patients [38].
Acknowledgements
The authors thank Matches Nübling, Joan Etlinger and Glyn Elwyn for valuable help.
