Doctors responses to patients concerns: testing the use of sequential analysis

Transcription

1 Doctors responses to patients concerns: testing the use of sequential analysis ATIE VAN DEN BRINK-MUINEN AND WILMA CARIS-VERHALLEN NIVEL (Netherlands Institute for Health Services Research), Utrecht, The Netherlands SUMMARY. Aims The aim of this small scale study was to explore interaction sequences during the medical consultation. Specific attention was paid to how doctors responded to patient s concerns and worries. Empathic behaviours ( e.g. concern, partnership, legitimising) and facilitating behaviours (e.g. paraphrasing, agreement) were considered as an adequate response to a patient s concern. Methods Nine consultations of nine different GPs were randomly selected from a sample of 1600 videotaped doctor-patient consultations, that were all rated with the Roter Interaction Analysis System. Each consultation contained at least 9 utterances of patient s concern. It was investigated how doctors respond within five lags of utterances after a patient s concern. Results The results showed that doctors more often responded to a patient s concern in a facilitative way than in an empathic way. When an empathic response was given, it appeared mostly during the first utterance after the patient expressed a concern. Conclusions The findings indicate that sequential analysis is appropriate to investigate a health care provider s specific style of responding. Based on the problems emerged during the sequential analysis, further exploration of the method is recommended. Declaration of Interest: none. KEY WORDS: general practice, doctor-patient communication, sequential analysis, communication style INTRODUCTION The awareness that communication is essential to both the process and the outcome of medical care has led to an increase in attention in educational programmes for doctors, in which effective communication is promoted. Several studies underlined the relationship between doctors communication skills and outcome measures, meaning that some communication behaviours are more effective in terms of meeting goals than are other behaviours (Del Piccolo et al., 2000; Gilotti et al., 2002) A number of studies showed that effective responses to patients concerns can reduce stress in patients (Butow et al., 2002). A sensitive response to emotional distress has been shown to reduce psychological morbidity in patients. However, the efficacy of such responses was dependent on the ability of health care providers to detect the patients needs for information and emotional support (Butow et al., 1996; 2002). The doctor s response to the first few patient cues, indicating concern or worry, had an important impact on subsequent disclosure (Maguire et al., 1996; Goldberg et al., 1993). It was shown that patients expressed themselves easier if General Address for correspondence: Dr. A. van den Brink-Muinen, NIVEL, PO Box 1568, 3500 BN Utrecht (The Netherlands). Fax: a.vandenbrink@nivel.nl Practitioners (GP) gave empathic statements (Del Piccolo et al., 2000). Adequate information and emotional support may result in better coping behaviour in patients, in stress reduction and even in an improved functioning of the immune system (Thorne, 1999; Spiegel & Stephton, 2001). There are indications that healthcare providers behaviours, such as showing agreements or paraphrasing, facilitate patients to express themselves (Bensing,1991; Van den Brink-Muinen et al., 2002) This, in turn, may reduce patients feelings of anxiety and uncertainty (Maguire et al., 1996), which provides space to assimilate information and can lead to the development of effective coping strategies (Van Dulmen, 1996; Kelley et al., 1997). On the other hand, doctors may inhibit patients by ignoring concerns and giving premature reassurance, or by changing the topic. These behaviours may prevent a patient from giving information and expressing their concerns (Maguire et al., 1996; Heaven, 2001). From the literature it is not clear if, and how often, doctors respond to patients in an adequate manner. So far, quantitative observation systems have been used, which hardly pay attention to the dynamic course of the interaction and give no insight into specific behaviours of doctors, e.g. to patients concerns. To uncover this behaviour it is necessary to investigate the sequences of responses given by doctors after such concerns. Moreover, the additional value of sequential analysis is also to use the results for training doctors in communication skills. The first problem when performing sequential analysis 92

2 Doctor s responses to patients concerns: testing the use of sequential analysis is to determine what are adequate (e.g. effective or sensitive) responses to patients concerns. Concerns were defined as Any statement or non-verbal expression indicating that a condition or event is serious, worrisome, distressing or deserving special attention (such as comforting or other special consideration) and is of particular concern for the patient during the medical interview. In other words, these statements have a strong and immediate emotional or psychosocial component, and do not refer to a more general frame of mind or past issues. Voice tone, intonation or verbal content may disclose worries, concerns, stress, nervousness, personal preferences or uncertainties that are of immediate concern (Roter, 2001). There is very little literature about translating the responding behaviour in operational terms (Caris- Verhallen, 1999). Based on the above cited literature, the following responses of the doctors were included in this explorative study as adequate responses to patients concerns: concern, empathy, reassurance (empathic behaviour), and paraphrase, agreement (facilitating behaviour). The second problem is how many utterances after the patients concerns should be included. If the doctor s first response is an adequate one, there is no problem. However, it may be assumed that firstly a question is asked, the patient answers etc. Then, the doctor may respond with an adequate answer only thereafter in the third or fourth of the next utterances. It is questionable how many lags between the patient s concerns and the doctors responses should be included (or which cut-off point should be taken) with respect to a reliable and relevant result. Moreover, sequential analyses should be restricted to a manageable size in order to be able to interpret the results. Therefore, it was decided to analyse up to and including, the fifth utterance after the patient s concern. Apart from the mentioned topics, we will explore whether based on sequential analysis, specific responding styles to patients concerns can be identified. To summarise, the research questions addressed in this study are (1) Which responses do doctors give to patients concerns?, (2) What is the best cut-off point of sequential utterances for analysing adequate responses?, (3) Do doctors show a specific style of responding to patients concerns? METHODS Data were derived from consultations of general practitioners (GPs), who participated in the Second Dutch National Survey of Morbidity and Interventions in General Practice (Schellevis et al., 2003). These consultations were videotaped and rated using Roter s interaction analysis system (RIAS) (Roter, 2001), a reliable and validated instrument especially designed for measuring communication in healthcare (Ong et al., 1998; Ram et al., 1999). Each utterance was designed to one of 26 exclusive categories. The RIAS distinguishes between instrumental (cure-oriented) and affective (care-oriented) communicative behaviours on the part of the doctor and the patient. This small scale explorative study was focused on the doctors responses to patients concerns. Therefore, only the utterances of doctors were included. Nine consultations of nine different GPs containing nine or more utterances of concern by the patient were at random selected (from a sample of 1600 consultations). For the present study the following empathic categories were considered as adequate: concern/worry; empathy/partnership/legitimise; reassurance; and the following facilitating categories were considered as adequate: paraphrase/ check for understanding; sign of agreement/ understanding (see table I). The other categories, such as giving information, asking questions, counselling, directions and social talk were defined as inadequate (or neutral) responses to the patient s concern and were therefore excluded from the Table I. - Categories of adequate communication responding to a concern of a patient. Empathic categories Concern /worry A statement indicating that a condition or event is serious, worrisome, distressing, and is of special concern for the patient. Empathy/partnership/legitimize Statements that paraphrase, interpret, or recognize the emotional state of the patient/ conveying alliance with the patient in terms of support and decision-making / Statements that indicate that the other's emotional situation is understandable. Reassurance Includes statements indicating optimism, encouragement, relief of worry or reassurance. Facilitating behaviour Paraphrase/ check for understanding Repetition or re-stating information told by the other / mechanisms reflecting back information told by the other for the purpose of checking for accuracy Sign of agreement/ understanding Signs of agreement or understanding and backchannel responses. 93

3 A. Van Den Brink-Muinen and W. Caris-Verhallen analyses. The first five utterances following the patient s concern were studied, assuming that relevant responses by the doctor should be given within five lags of utterances after patients concerns. Since the study was exploratory and the sample was rather small, no statistical analyses were performed. RESULTS Doctors adequate responses to the patients concerns In total, 111 utterances of concern were given by the 9 patients to the 9 GPs (on verage about 12 concerns per consultation). The GPs gave 136 adequate responses to these concerns in the next 5 utterances (by doctors or patients). Half of the responses were agreements or signs of understanding, as yes, hmm, OK (table II). A quarter of the doctors responses consisted of paraphrasing or summarising what the patient had said, and checking for understanding (as So, you are worrying about... ). Empathic answers like This must be terrible for you and showing concern (like the patient did just before) were given less often by the doctors, while they seldom reassured the patients in response to their concerns. Paraphrases, agreements and concerns were mostly the first utterances after patients concerns, while paraphrases and agreements were also often given as next utterances. Cut-off points for adequate responses to the patients concerns Half of the adequate responses by the doctors were given as the first utterance after the patients concerns, and one fifth as the third utterance (table III). After the second utterance 61% of the adequate responses were given, and after the third one 81%. The inclusion of the fourth and fifth utterances contributed only 19% to the adequate responses. Doctors styles of responding to the patients concerns Based on the finding that 81% of adequate responses was given in the first three utterances, the division of these 81% responses was investigated per GP (table IV). They varied reasonably between each other. The percentage of adequate responses in the first utterance varied between the doctors from 48% to 72%, in the second utterance from 7% to 23% and in the third utterance between 9% and 30%. Most doctors (except for GP number 2 and 5) on average gave at least one adequate response to the patients concerns. The nine GPs also greatly differed with respect to the type of utterances given to the patients concerns (table V). GP number 1, for instance, mainly responded with agreements (9 or 82% of all adequate responses), GP number 6 with only 5 agreements of 19 responses (26%), but this doctor most often responded in an empathic way (47%) as compared to the other GPs. Concerns were most often given as response by two GPs (number 5 and 8 both 30%). Four responses (27%) of doctor number 3 were reassurances to the patient. Paraphrasing was most often used by GP number 8 (41%). Table II. - Adequate responses (N and %) by type of utterances in the first 5 utterances after patients concerns. Utterance concern empathy reassurance paraphrase agreement Total N (136) % Table III. - Adequate responses (total N and %) per utterance in the first 5 utterances after patients concerns. Utterance N % cumulative % : cut-off point N= % 100% 94

4 Doctor s responses to patients concerns: testing the use of sequential analysis Table IV. - Adequate responses (N and %) in the first 3 utterances per GP. GP Utt 1 (N) Utt 1 (%) Utt 2 (N) Utt 2 (%) Utt 3 (N) Utt 3 (%) Utt 1-3 (N) Patients Concerns (N) Table V. - Adequate responses (N) by type of utterances in the first 3 utterances per GP. GP concern empathy reassurance paraphrase agreement _ CONCLUSION AND DISCUSSION This small scale explorative study in general practice has shown that sequential analysis can give insight into the way in which doctors respond to patients concerns. However, several problems have arisen during the analysing process, which are addressed later on. Firstly, the research questions are answered and discussed. The first research question was: Which responses do doctors give to patients concerns? Doctors answered less often to the patients in a empathic way (showing concern and empathy and giving reassurance), than in a more facilitating way. They more often gave signs of agreement and understanding and paraphrased or checked for understanding of the patients concerns. These responses may give more opportunities to the patients to explore and tell their feelings and concerns and the doctors also show their interest in the patients story. Therefore, this facilitating behaviour of doctors is also important, although less directly aimed at showing that the patients feelings and concerns are understandable or that the patients need to be reassured. If empathic responses were given, they mostly appeared to be the first utterances after the patients concerns, while agreements and paraphrases were also given in latter utterances. The next question was: What is the best cut-off point of sequential utterances for analysing adequate responses? As half of the adequate responses appeared to be given in the first utterance, while 81% was given in the first three utterances after the patients concerns, it appears a good choice to include the first three utterances in the sequential analyses in order to investigate the doctors responding adequate behaviour. This is the more appropriate because in the fourth and second utterances, accounting for only 20% of the responses, mainly facilitating behaviour was shown by the doctors. The third research question was formulated as Do doctors show a specific style of responding? The doctors did differ between each other in the percentage of adequate responses per utterance as well as in the type of responses. Some doctors gave most of the adequate responses in the first utterance after the patients concerns, while other doctors also gave much adequate responses in the third utterance. The second utterance seems to be seldom used, mainly because the patients responded to the doctors first response. The different type of responding was most clearly shown by the type of responses. One doctor specifically gave concerns, another mainly responded empathically and again another GP tried to reassure the patient. Therefore, sequential analysis seems appropriate to 95

5 A. Van Den Brink-Muinen and W. Caris-Verhallen investigate the doctors style of responding. However, this study was explorative and based on only few consultations of general practitioners. Future research is necessary to determine whether the present findings are valid. Problems with sequential analysis The first problem is the definition of what is an adequate response and what is inadequate (or neutral). In the present study the choice was made on a theoretical basis. The starting point was the concern of the patient, which was defined as a statement indicating that a condition or event is serious, worrisome, distressing or deserves special attention. This definition was applied in the medical context of general practices, but may be not appropriate in every context. The adequateness of a response may also differ because of the definition of the patients cue, i.e. focus of this study, and because of the context. The choice for both the empathic and facilitating behaviour as adequate responses is arbitrary. It could be argued that facilitating behaviour is not adequate, because the doctor does not really go into the concern. Besides, one might argue that psychosocial or medical questions, information or advice, may be suitable as responses to the patients concerns, depending on the context of the encounter, e.g. general practices, hospitals or nursing homes. Thus, qualitative analysis may be necessary to solve this problem, additionally to RIAS (used in the present study) or another comparable observation method. Perhaps it would be preferable to develop, or to build on another observation system taking into account the context of the cues. In this study the cut-off point of the inclusion of sequential utterances was determined after the third utterance. In this way, four fifths of the adequate responses were covered. The empathic responses were included for the greater part, while the somewhat less important - facilitating utterances more often occurred also after the third utterance. Another question is whether these facilitating behaviours (agreement and paraphrases) are important to include if these responses are given after the first utterance of the doctor. For, they then may be given in response to a patient s utterance and can, therefore, not be considered as an adequate response to the patient s cue. So, maybe there is a need to determine different cutoff points for different types of response. There are some problems to be solved that are related to each other. Sometimes, more than one cue was given in succession by the patients, or the patients had other utterances directly after their cues. Regarding the doctors, often more than one adequate response was given to the patients cues. Should these double utterances be neglected in the sequential analyses? Maybe it would be a better solution to decide to rate per turn of the speakers instead of per utterance, and then to investigate whether one or more adequate answers are given during this turn. Back channel responses (yeah, hmmm) are often given during a series of cues of the patient. One could argue that these agreements are less (or not) important compared to agreements given after concerns of the patients, dependent on the research question and the context of the communication. By rating only the responses of interest this problem could be solved. Performing sequential analysis per utterance for a whole consultation in general practice or in another context is not always relevant for the research question of a study. Thus, sometimes if may be sufficient to analyse only phases of provider-patient communication, especially if the communication lasts a long time, e.g. more than 30 minutes. REFERENCES Bensing J.M. (1991). Doctor-Patient Communication and the Quality of Care. NIVEL: Utrecht. Butow P.N., Kazemi J.N., Beeney L.J., Griffin A.M., Dunn S.M. & Tattersall M.H. (1996). When the diagnosis is cancer: patient communication experiences and preferences. Cancer 77, Butow P.N., Brown R.F., Cogar S., Tattersall M.H. & Dunn S.M. (2002). Oncologists reactions to cancer patients verbal cues. Psychooncology 11, Caris-Verhallen W.M.C.M. (1999). Nurse-Patient Communication in Elderly Care. NIVEL: Utrecht. Del Piccolo L., Saltini A., Zimmermann C. & Dunn G. (2000). Differences in verbal behaviours of patients with and without emotional distress during primary care consultations. Psychological Medicine 30, Gilotti C., Thompson T. & McNeilis K.S. (2002). Communicative competence in the delivery of bad news. Social Science and Medicine 54, Goldberg D.P., Jenkins L., Millar T. & Faragher E.B. (1993). The ability of trainee general practitioners to identify psychological distress among their patients. Psychological Medicine 23, Heaven C.M. (2001). Measurement of Skills. (Unpub PhD thesis) Psychological Medicine Group, Stanley House, Christie Hospital: Manchester. Kelley J.E., Lumley M.A. & Leisen J.C.C. (1997). Health effects of emotional disclosure in rheumatoid arthritis patients. Health Psychology 16, Maguire P., Faulkner A., Booth K., Elliott C. & Hillier V.F. (1996). Helping cancer patients disclose their concerns. European Journal of Cancer 32, Ong L.M.L., Visser M.R.M., Kruijver I.P.M., Bensing J.M., van den Brink-Muinen A., Stouthard J.M.L., Lammes F.B. & de Haes J.C.J.M. (1998).The Roter Interaction Analysis System (RIAS) in 96

6 Doctor s responses to patients concerns: testing the use of sequential analysis oncological consultations: psychometric properties. Psycho- Oncology 7, Ram P., Grol R., Rethans J.J. & van der Vleuten C. (1999). Videotoetsing van consulten van huisartsen in de eigen praktijk [Assessment of communicative and medical performance of general practitioners in daily practice. Validity, reliability and feasibility of video observation]. Huisarts & Wetenschap 42, Roter D.L. (2001). The Roter Method of Interaction Process Analysis. Johns Hopkins University: Baltimore. Schellevis F.G., Westert G.P., de Bakker D.H., Groenewegen P.P., van der Zee J. & Bensing J.M. (2003). De tweede Nationale Studie naar ziekten en verrichtingen in de huisartsenpraktijk: aanleiding en methoden. [Second Dutch National Survey of Morbidity and Interventions in General Practice] Huisarts & Wetenschap 46(1),7-12. Spiegel D. & Stephton S.E. (2001). Psychoneuroimmune and endocrine pathways in cancer: effects of stress and support. Seminars in Clinical Neuropsychiatry 6, Thorne S.E. (1999). Communication in cancer care: what science can and cannot teach us. Cancer Nursing 22, Van den Brink-Muinen A., Van Dulmen A.M., Messerli-Rohrbach V. & Bensing J.M. (2002). Do gender-dyads have different communication patterns? A comparative study in Western-European general practices. Patient Education and Counseling 48, Van Dulmen A.M.(1996). Exploring Cognitions in Irritable Bowel Syndrome: Implications for the Role of the Doctor. KUN: Nijmegen. 97