|Home | About | Journals | Submit | Contact Us | Français|
General practitioners’ (GPs) communication with patients presenting medically unexplained symptoms (MUS) has the potential to somatize patients’ problems and intensify dependence on medical care. Several reports indicate that GPs have negative attitudes about patients with MUS. If these attitudes deter participation in training or other methods to improve communication, practitioners who most need help will not receive it.
To identify how GPs’ attitudes to patients with MUS might inhibit their participation with training to improve management.
GPs (N=33) who had declined or accepted training in reattribution techniques in the context of a research trial.
GPs were interviewed and their accounts analysed qualitatively.
Although attitudes that devalued patients with MUS were common in practitioners who had declined training, these coexisted, in the same practitioners, with evidence of intuitive and elaborate psychological work with these patients. However, these practitioners devalued their psychological skills. GPs who had accepted training also described working psychologically with MUS but devalued neither patients with MUS nor their own psychological skills.
GPs’ attitudes that suggested disengagement from patients with MUS belied their pursuit of psychological objectives. We therefore suggest that, whereas negative attitudes to patients have previously been regarded as the main barrier to involvement in measures to improve patient management, GPs devaluing of their own psychological skills with these patients may be more important.
Medically unexplained symptoms (MUS) burden the many patients who suffer them and the general practitioners (GPs) whom they consult.1,2 Patients have as poor a quality of life as those with comparable symptoms caused by disease.3,4 Many doctors find these consultations particularly challenging,5–8 and these patients consume symptomatic health care disproportionately although, because there is no disease, treatment is often ineffective.4,9–11 There has long been concern that unnecessary symptomatic treatment increases dependence on medical care.12–15 In addition, GPs often communicate with these patients in ways that have the potential to “somatize” their problems and intensify their dependence. To this extent, somatization is therefore an iatrogenic as well psychological process.16
Improved management will require change in the way that GPs communicate with patients.17 However, although GPs recognize MUS as a responsibility of primary care2 and an area of training need,18 many studies have shown negative attitudes to these patients, as the widespread designation of patients as “difficult” and “heartsink” illustrate.5–7,19,20 If such attitudes deter the take-up of interventions such as training or supervision from mental health professionals, which could change their communication, then the practitioners who most need these interventions would not receive them. Information about how practitioners’ views and experience of MUS might deter them from participation would allow future interventions to be designed in ways which increase their adoption.
Reattribution training has been successfully delivered to GPs in several countries. It leads to more positive attitudes towards MUS patients, reduces health care costs, modifies patients’ beliefs, reduces emotional distress, and improves function.21–23 However, most GPs to whom it was offered in a recent trial (MUST: the UK Medical Research Council Medically Unexplained Symptoms Trial), declined it. Extensive social psychology research confirms that attitudes, even stereotypical and negative ones, are contextually dependent.24 Nevertheless, previous studies of practitioners’ attitudes to patients with MUS have lacked a specific context for eliciting attitudes, and therefore risk generalised and idealised accounts that may not be relevant to specific opportunities for improved practice. Therefore, we interviewed samples of practitioners who had declined or accepted participation in MUST, to identify attitudes to MUS that might have led most to decline participation.
GPs were invited by letter or telephone to participate in a study in which they would receive training to help patients reattribute unexplained physical symptoms to lifestyle, psychosocial circumstances, or innocuous causes. For participating practitioners, training in three 2-hour sessions took place in the GPs’ practices and incorporated techniques that are known to be preferred by GPs, including small group teaching and video feedback.25 Practitioners were offered financial compensation to employ locum GPs for the periods of training. Of 1934 GPs invited to participate in MUST, 70 agreed, confirmed agreement when visited and were from practices in which all GPs agreed.
To obtain a sample of GPs who had accepted participation but had not received training, 11 GPs allocated to the control condition were telephoned and asked to arrange an interview before they received training. One had left the practice, but the remainder agreed and was interviewed. It is difficult to involve in research a group of practitioners who are defined by their refusal to participate. Therefore, we wrote to GPs who had declined or were unable to participate, emphasizing that the research team had much to learn about the reasons why they could not participate, and inviting them to take part in a short interview to tell the research team about these reasons. Of 76 who were asked, 16 agreed. Subsequently, telephone calls to a further 21 GPs led to interviews with a further seven. Sampling in both groups was purposive, including GPs: from each participating area (Liverpool, Greater Manchester, Lancashire and Wirral); from postgraduate training and nontraining practices; and, for those who declined to take part in MUST, including some who declined at each stage of recruitment (initial letter of invitation, subsequent telephone call, and visit from one of the research team). In accordance with accepted principles of qualitative methodology, recruitment ended when no new information was emerging.
The 10 participants who had accepted training included 7 females and 5 from practices involved in training medical students or trainee GPs. Practice size ranged from 3 to 8 GPs (median 4). They did not differ in these variables from the remainder of participating GPs. The 23 participants who had declined to take part in MUST included 5 female GPs and 10 from practices involved in medical training. Practice size ranged from 1 to 4 (median 3). Ten had not responded to the initial letter, seven had responded negatively, five had decided against participation after further information and one could not participate for procedural reasons. Gender was not available for GPs who participated neither in the trial nor the interviews, but interviewed GPs who did not participate in the trial did not differ from ones who were not interviewed on practice size, whether practices were involved in medical training or stage at which they declined to participate in MUST.
Interviews were in person in the GPs’ practice at a time of their choosing and remunerated. Although the reason for interviewing some GPs was that they had declined to participate, interviewers avoided an accusatory stance and emphasised acceptance of their decisions and belief that the research team had much to learn from practitioners who could and could not participate. The interview then included discussion of: (1) GPs’ views of the training that was offered; (2) their reasons for agreeing or not feeling able to participate; and (3) their own experience of patients with MUS. The 2 interviewers used an interview guide but aimed for a conversational style, avoiding closed questions and using open questions, prompts, and reflection. Interviews lasted from 20–55 minutes. The interview guide was subsequently modified to test or develop aspects of the analysis, which was conducted in parallel with interviews (see below). Interviews were audiorecorded, and anonymized transcripts were prepared before audiorecordings were destroyed. All text that was informative about GPs’ decisions and about their attitudes to patients with MUS and to training was retained for the present analysis.
Analysis, which proceeded in parallel with the interviews, was inductive in that no categories were specified in advance. Our procedure followed established conventions for grounding analysis in data rather than in preconceptions.26,27 Transcripts were first read and discussed by the multidisciplinary team, which encompassed expertise in psychiatry, primary care, clinical psychology, and sociology (PS, SP, LG, AR, RC, WI, and RM). Three members (PS, RC, and WI: a clinical psychologist with extensive research and clinical experience in primary care and the 2 interviewers) then separately identified recurrent patterns, testing and modifying them by constant comparison, “cycling” between sets of data and the developing account of them and by discussion. Regular discussion by the whole team was informed by their reading of the accumulating transcripts and the continuing thematic analyses by these 3 members. Periodically, the analysis was referred to the broader group of those involved in the MUS trial, which included academic and practicing GPs. Analysis focused first on GPs who had declined to participate in MUST, before examining the data from participating GPs for evidence of commonalities and divergences from these. Rather than taking a purely descriptive or phenomenological approach to analysis of practitioners’ speech, we considered its social function—the work that it did in constructing the problem of MUS and patients’ and GPs’ roles.28 Analysis was finalized when no further changes emerged from this process and when all the relevant text could be accommodated in it. In our final analysis, each category was defined by commonality of meaning across data from several GPs.
The importance of cycling between data and emerging analysis, and triangulation between perspectives of different authors in different roles, are widely recognized as procedural sources of trustworthiness of qualitative analyses. In addition, the presentation of raw data is important in exposing the analysis to scrutiny.29 Procedures are, however, insufficient to guarantee useful findings.30,31 Additional standards by which we assessed the developing analysis were coherence and “theoretical validity” whereby conclusions should connect with theoretical ideas beyond the present study.27 We were also concerned with the “catalytic validity” of the findings;32 that is, they should have the potential to render the subject more tractable to researchers or clinicians.
In reporting the final analysis, we present transcript to illustrate the range and commonality of meaning of each category of the analysis. The ellipsis indicates omitted speech.
Accounts of GPs who had agreed or declined to participate in MUST overlapped in their objectives for patients with MUS, but diverged strikingly in their attitudes to patients and to their own psychological work. Findings from these samples are therefore presented separately.
As expected, attitudes implying disengagement from patients were prominent in these GPs’ accounts, but the GPs also indicated several objectives that they had for engagement. Therefore, after first describing their negative attitudes and their objectives, we examined the tensions within their accounts.
Most GPs (N=15, 65%) described attitudes that implied disengagement from patients (Box 1 in the Appendix); patients’ complaints were not legitimate demands on medical care, reflecting the absence of “real” illness; it was impossible to help them, or it was pointless to try, because they refused what GPs thought was necessary or they were unwilling to change. Denigratory language was common and a few GPs were explicit in their dislike.
Notwithstanding their negativity, a few GPs voiced sympathy for patients ([GP19] “I think it’s a shame for them because we’re only getting an iceberg of their anxiety aren’t we, it’s probably there in the background all the time and they only present to us when they reach a peak that they can’t actually handle themselves”) and most (N=15, 65%) described clear objectives for engagement. Invariably, the priority was medical investigation. However, 14 (61%) also described explicitly psychological objectives (Box 2 in the Appendix). The overriding aim was to maintain the relationship as a vehicle for helping the patient. Several were optimistic about the value of being “straight”, in describing the absence of treatable disease. Although one observed that [GP21] “It’s very difficult to explain something when you don’t know the explanation yourself”, a few described attempts to reframe symptoms and some used aphorisms and self-disclosure. Seven GPs described explicitly seeking to identify and manage emotional problems. Whereas 1 had used antidepressants (unsuccessfully), and another had [GP23] “more of a counselling model”, most described informal approaches (Box 2 in the Appendix).
Negative and dismissive attitudes and psychological objectives and strategies were generally described by the same GPs (Box 3 in the Appendix). Of 15 (65%) GPs who voiced attitudes of “disengagement”, i.e., not taking responsibility for patients’ problems, or regarding them negatively, 11 (73%) also indicated engagement with patients by describing sympathy for them, or objectives or strategies for working with them psychologically.
GPs used contrasting language to describe their psychological work and excluding disease. Whereas the latter was described as highly technical and important, psychological approaches were typically preceded by “just”, as in [GP20] “it’s just supporting those patients” or [GP24] “they just want to be acknowledged that they have an issue”. One GP explicitly denigrated the ways that he had learned to support a patient: [GP32] “Basically my advice to anyone is that you’ve got to appear to listen because whether you do or not it doesn’t matter, but you’ve got to appear to listen ... I’ve got one woman and she goes all round the houses and she just wants me to give her a phrase of the week, something like ‘all downhill’ or ‘what goes around comes around’. She likes sayings like that and she uses them then and relies on them to get sympathy I think... She just needs someone to come and talk to. She just wants a chat.”
These GPs regarded patients as making legitimate demands that merited GPs’ attention. GPs saw them as [GP14] “part of the job... if it’s there, it’s got to be dealt with”, several described them as “interesting” and one felt [GP1]“quite fond of them really”. GPs’ empathy with these patients was demonstrated by the complexity of their accounts of MUS (Box 4 in the Appendix), in which they linked psychological problems to specific difficulties in patients’ current or earlier life and to cultural factors and medicalization.
GPs’ objectives for MUS patients included those described by the nonparticipating GPs, and extended to include asking patients to complete diaries or a life story, problem-solving, giving coping skills, helping patients to gain insight, scheduling regular reviews or inviting patients to book longer appointments. Two had explicitly addressed problems in the GP–patient relationship, including dependence.
Like the nonparticipating GPs, they distinguished the biomedical and psychological aspects of their role, emphasizing the importance of excluding disease. However, unlike the nonparticipating GPs, none denigrated their psychological work. Instead, they described this aspect of their work in ways that indicated skill, such as [GP1] “active listening” or in observing that [GP10] “the skill is getting the balance right” between meeting patients’ support needs and encouraging dependency. Many comments indicated enthusiasm or interest. One described [GP14] “the pleasure of repeat visits... time on your side” and several referred to “giving” time to the patients. For one [GP1] “the most fascinating part of your work is the consultation”. Another described the importance of the relationship to the patient [GP10] “It would be easy as soon as you suspect something to say go and see the psychologist. That I don‘t think would help quite a lot of patients because they would feel they were rejected by the doctor.”
Previous reports, based on interviews about generalized attitudes in unselected practitioners, suggested that they find patients with MUS difficult or have negative views of them.5,7,8,20,33 Our study indicates a more complex picture, characterized by variability between practitioners and by tension between different parts of their accounts.
GPs described psychological objectives for working with patients with MUS, some of which have been reported previously.34 Many aimed for honesty, and even therapeutic effect, in being “straight” that disease was absent. Some described acknowledging the reality of patients’ symptoms and helping them reattribute symptoms. Several described probing about emotional and social problems. The overriding psychological aim was to maintain a relationship, and to use this as the vehicle for addressing psychological problems and providing support. Support was generally informal and involved listening to the patient, as well as idiosyncratic techniques including self-disclosure and aphorisms to help patients make sense of their problems.
Negative and dismissing attitudes to these patients did not indicate lack of engagement with them. Although GPs who had declined training described not taking responsibility for patients’ problems, considering patients unrewarding or to be making illegitimate demands, they also reported sympathy, acceptance of responsibility, and careful attempts at engagement. Dismissing attitudes cannot, therefore, be taken at face value as explaining why GPs did not accept training.
Additional negative attitudes also distinguished the GPs who declined training. Although they valued the biomedical aspect of their role—excluding disease—and described it technically and as an important service to patients, they typically described their psychological work as “just” support or “just” reassurance and sometimes explicitly denigrated it. These GPs thereby devalued their own skills and expertise in psychological work and in maintaining effective clinical relationships.35 By contrast, GPs who accepted training were all positive about their psychological work, describing it as skilful, interesting, and valuable to patients. We cannot say whether these GPs devalued psychological work in general or only with MUS patients.
Negative attitudes to patients in GPs who declined training may be a way to cope psychologically with the emotional challenges that patients present.36 As well as by describing protean symptoms that defy biomedical explanation,6,20,37 patients with MUS challenge practitioners by their psychological distress6,20,38 and demand for emotional support.39 The cognitive dissonance40 that arises when practitioners say that patients cannot or should not be helped, but then try to help the patient, may further compound GPs’ discomfort. Devaluing their own psychological skills might also be a coping response. An alternative explanation lies in these GPs’ dualistic thinking, which contrasted with the holistic and biographical approach that characterizes academic accounts of primary care.41,42 They identified their professional role with the biomedical opportunities of physical medicine rather than the psychological aims and methods of a psychosocial approach.
The importance to GPs of preserving clinical relationships, which has been blamed for promoting collusive communication with patients with MUS,43,44 might be expected to deter adoption of approaches that are perceived to formalize or threaten the relationship. Emphasis on clinical relationship did not, however, distinguish those who declined training from those who accepted it. Negative attitudes to patients suggest barriers to engagement, and previous writers have suggested that these should be explicitly targeted by programs to improve the care of MUS.5 However, these attitudes were belied by evidence that the GPs who displayed them did engage psychologically with their MUS patients.
More fundamental barriers should be considered. If GPs are negative about psychological skills because they are emotionally challenged by patients with emotional demands, such as those with MUS, then approaches that are perceived to require increased involvement with such patients would be unattractive. Ongoing supervision and support may be required before GPs such as these engage with such approaches. If GPs are negative about their psychological skills because they devalue psychological work by comparison with biomedicine, interventions to improve these skills would be unattractive compared to ones focused on enhancing technical or biomedical skills. Efforts to enhance the professional status of psychological skills would need to precede the offer of training or other interventions.
Whereas previous reports described practitioners’ attitudes to MUS,5,7,8,20,33,34 our analysis extended to an interpretative level that suggested psychological functions of attitudes that have hitherto been regarded as primary. Interview studies can be compromised by respondents’ tendency to speak in socially desirable ways. However, GPs in our study readily voiced negative and critical attitudes. Our findings arose in the context of the offer of a specific form of training as part of a research trial, and the offer of a different intervention, or the absence of a research context, might evoke different responses. Nevertheless, our design avoids generalized or idealized accounts. Moreover, the intervention implemented treatment and training principles that are widely advocated for these patients and for primary care, and the key barrier that we identified—devaluing of one’s own psychological skills—would impede any formal attempt to engage practitioners in psychological care for MUS. Practitioners who have declined to participate in training or research are, while potentially critical informants, inevitably hard to recruit for a study of the reasons for their nonparticipation. Only 24% of GPs in this group who were asked to take part agreed. Because of this, and the small sample size in both groups, the attitudes that we have described cannot be regarded as representative. Ultimately, the general applicability of our analysis can be tested in future work that seeks to engage GPs in improving management of MUS.
Our findings have implications for attempts to extend training or other interventions to improve management of MUS in primary care to GPs such as those who declined to participate in MUST. To focus on changing GPs’ negative attitudes to these patients, as previously advocated,5 is probably mistaken. First, we need to understand whether some GPs are particularly challenged by psychological demands that patients with MUS present, and whether this accounts for their devaluing psychological skills. This will require the future study of practitioner, as much as patient, characteristics,39,45,46 including their reactions to the emotional needs of patients with MUS and their tolerance or intolerance of uncertainty.45 Secondly, we need to understand the role of GPs’ dualistic separation of physical and psychological care in their deprecation of psychological skills, and we need to explore how these putative effects of dualism can be overcome.45,47 Only if GPs value their psychological role will they accept help to enhance it.
The study was supported by the UK MRC. We are grateful for the collaboration of Professor Francis Creed, Professor Graham Dunn, and Dr Barry Lewis in the design of the MUST trial, and to Ms Judith Hogg for the management of the trial and for the cooperation of the participating general practitioners.
Contributors RM led the MUST study, from which this report arises, LG, SP, CFD, PS, and AR contributing to the overall design and management. PS led analysis of these data, to which SP, LG, AR, RC, WI, and RM contributed, and drafting of this report, to which all authors contributed.
Potential Financial Conflicts of Interest None disclosed.