"I know what to do, but if s not possible to do it" general practitioners' perceptions of their ability to detect psychological distress

Transcription

1 Family Practice Oxford University Press 1996 VoL 13, No. 2 Printed in Great Britain "I know what to do, but if s not possible to do it" general practitioners' perceptions of their ability to detect psychological distress Amanda Howe Howe A. "I know what to do, but it's not possible to do it" general practitioners' perceptions of their ability to detect psychological distress. Family Practice 1996; 13: Background. Accurate detection of psychological distress in patients is a prerequisite of specific diagnosis and active management. Studies have shown that improved detection is related to altered management and to improved patient outcomes: there may also be a link with improved patient satisfaction. Objective. Many factors in the doctor, patient, and context of the consultation may influence whether or not a GP identifies psychological distress in a patient; whatever the triggers to detection, it has been shown that specific training in appropriate skills can alter clinician behaviour and improve detection rate. This study examined the GPs' own perceptions of the influences on their performance as detectors of psychological distress. Method. A postal questionnaire yielded nineteen GPs who were personally interviewed for the study. Results and conclusions. The study highlights GPs' sense of the difference between possessing the necessary skills and employing them in daily practice. This has implications for training and clinical practice in this area. Keywords. Consultation style, general practitioners, literature, patient factors, psychological stress. Introduction The current high profile of mental health as a major source of morbidity 1 and as a target for improved clinical performance 2 has led to close scrutiny of GP performance in this field. Numerous studies 1 " 3 have shown an apparent underperformance by GPs in their ability to detect accurately the presence or absence of psychological distress in their patients. Widely validated instruments such as the GHQ*- 7 have been used repeatedly in the primary care setting 8-9 to provide an independent estimation of the patient's mental state at the time of consultation, and have shown that GPs may fail to detect up to half the patients who show significant psychological distress on questionnaire response. 10 This is a source of considerable concern, as accurate detection has been shown to improve patient outcomes, 11 with GPs altering their management 12 and Received 18 August 1995; Accepted 9 November Department of General Practice, Community Sciences Centre, Northern General Hospital, Sheffield S5 7AU, UK. with patients being more satisfied 13 if their emotional distress is adequately identified. Factors influencing GP detection of psychological distress A number of studies have attempted to elucidate the factors that lead to the wide variation in GP performance as detectors of psychological distress. Their findings fall into three broad areas: doctor factors, patient factors, and structural or contextual factors. All these have been explored to find out why there is such a great variation in how GPs diagnose, 14 label, 13 and manage 1617 mental health problems. Doctor factors Some authors have found support for the likelihood of a link between clinical training and expertise in detection. 18 There is certainly an explicit assumption in the recent Joint Report that improved training for physicians will improve diagnosis and management of covert mental health problems." Work by Goldberg 127

2 128 Family Practice an international journal also suggests that trainings undertaken at undergraduate or postgraduate level may subsequently influence a GP's detection rate. 20 Attitude is a more complex factor: this implies whether a GP is orientated towards psychological medicine or not. It has been operationalized differently by different authors: Mechanic 21 classified doctors by their scientific and social orientation according to their use of technical procedures and views on the scope of general practice. Marks and colleagues, 5 and subsequently Boardman, 10 utilized the concept of 'bias' to indicate the likelihood of a GP making a positive psychological diagnosis, with the related indicator of the 'identification index' denoting the GP's level of concordance with the patient's score on the GHQ (corrected for prevalence and sensitivity of the GHQ). Wilmink, 22 however, found bias to be a "questionable index", and has attempted to analyse GPs instead by their attitudinal orientation to "clinical" or "family medicine''. In all studies, GPs with a greater sensitivity to psychological issues make more diagnoses of psychosocial factors being relevant to the consultation (though this does not always make them more 'accurate'). 'Accuracy' is another GP variable derived from Goldberg's work, and accepted by Wilmink: this can be estimated by measuring the correlation coefficient between GP opinion and GHQ score. How attitude is measured may be contributing to the lack of clarity in the literature: again, Wilmink found a surprisingly poor correlation between GP attitude and performance, but this led him to question his measures of GP attitude. However, in the field of depression, where the scales for measuring attitude have been somewhat better validated, GP attitude has been clearly shown to influence treatment as well as diagnosis, 17 which supports the view that GP attitude to mental health is important. Thirdly, consultation style has attracted a great deal of attention in the literature. Some of the first work on attitude 3 also looked at consultation content, and showed that GPs who are good detectors of psychological distress tended to make more psychosocial enquiries. Other work showed that doctors' 'identification index' (defined by Marks et a/, as "a measure of the vigilance with which doctors consider the possibility of psychiatric illness") correlated with responding in certain ways to the patient's cues, 23 and that the interviewing style of the GP seemed to influence whether the patient revealed psychological distress or not. This was confirmed by Millar and Goldberg, 24 who showed that trainee GPs who were good detectors of psychological distress would also score consistently high when the communication skills they displayed in the management phase of the consultation were analysed. This was shown to be a fairly constant factor in the GP's behaviour, whether the problem was psychological or not. Verhaak 25 also showed that GPs displayed a more open consulting style if they were dealing with psychosocial issues, though doctors who were less patient-centred made less initiatives to explore psychosocial issues. It can therefore be concluded that how a GP behaves in a consultation will alter their likelihood of detecting psychological distress in the patient, and that a GP who adopts a more open consulting style is likely to improve their performance as a detector of psychological distress. Patient factors There are two aspects to patient factors which influence GPs in detection of psychological distress: those that lie within the perception of the GP, and those over which the patient has some control. Various demographic characteristics have been consistently shown to make a GP more likely to suspect a psychological problem, gender and age being two major examples. Marital disruption 314 and unemployment 3 also seem to act as cues to GPs in some cases, and a previous mental health problem is consistently linked with a likelihood of the GP diagnosing a mental health problem. 22 Previous knowledge of, and relationship with, the patient may also aid diagnosis. 26 Other work has suggested that GPs link certain illness behaviours with a diagnosis of emotional disorder, 27 for example frequent consultations or multiple symptom patterns. It must, however, be emphasized that these are factors which alter GP detection but do not necessarily make the GPs more accurate. As regards patient factors over which patients have some control, the main issue here is how the patients present their problems to the GP. There is evidence that patients who score very high on the GHQ are more likely to present with psychological symptoms to the GP, 28 and that patients who present physical reasons for attending the GP may be less distressed clinically. Somatic presentation, however, is strongly correlated with a drop in GP detection of underlying psychological distress (by a factor of five in one study 29 ), and patients who do not offer cues of psychological disorder to the GP overtly are therefore much more likely to be misdiagnosed. Table 1 summarizes the patient factors found to influence GP detection of psychological distress. Other factors that influence GP detection Two major factors outside the doctor and patient have been examined which may influence GP detection of psychological distress: prevalence in the primary care population, and time. In theory, a GP with a high prevalence of psychological distress among her clientele would be slightly more likely to detect distress, providing their own threshold of ability remains constant, since prevalence has a major effect on predictive value. 29 There is some evidence 3 that GPs with certain personality traits have a clientele with a higher than

3 TABLE 1 Patient characteristics reported in the literature as influencing GP detection of psychological distress Age Gender Marital disruption Unemployment Educational status and social class Mode of symptom presentation Best in middle age Females > males Separated and bereaved > married or single Unemployed > employed Findings conflict Psychological > somatic Detection of psychological distress 129 less likely to detect or manage psychological distress, though patient and doctor factors probably play a more important part in the overall diagnostic process. Do GPs agree with the literature? A background to the study of GP perceptions The extensive literature summarized above contains very little subjective material from GPs themselves. As part of a study into GP detection of psychological distress which took place throughout 1993 in Sheffield, semi-structured interviews were undertaken with 19 GPs. Their perceptions of the reasons for GPs failing to detect psychological distress were specifically enquired about, and the results are presented here. Past history mental health problem Previous relationship with GP GP more likely to diagnose a mental health problem now More accurate diagnosis average psychological morbidity, but how or why these factors are related, and how consistent a finding it is, remains unclear. Time pressures have also been shown to alter GP detection, though it appears to be a minor source of variation compared to factors such as doctor consultation behaviours (3% variance 3 as opposed to 45% in one study). There are methodological difficulties in measuring consultation length: 30 approaches such as simple booking interval are inaccurate, but direct timing is complex technically and can be a major source of study effect. However, studies where consultation lengths have been assessed yield interesting findings. There seems to be clear evidence that consultations with a psychosocial content take longer, 31 and that longer consultations are also associated with better detection of psychological distress by GPs. 3 In another study, Howie and colleagues 31 showed that longer consultations were associated with a more comprehensive assessment of psychosocial issues. GPs may have a consulting style which leads to longer consultations, but this has been shown to vary with the type of problem presented, sociodemographic features of the patient, 33 workload within the surgery 32 and the booking intervals of the appointments. 33 This last factor has also been shown to influence the GP's stress level, 34 and in this way the speed of working which a GP is required to undertake may interfere with their performance. Interestingly, patients who are significantly emotionally distressed at consultation have expressed dissatisfaction with the amount of time they perceived the GP to have, and indicated that this led them to conceal issues they would have liked to discuss. 33 To summarize these findings, it seems possible that shorter time available in the consultation may make GPs Method As part of a postal questionnaire sent to a random sample of GP principals in Sheffield, GPs were invited to participate in a study of GP detection of psychological distress. Nineteen GPs were recruited from those expressing interest, and all these GPs were personally interviewed by the researcher prior to the commencement of the quantitative data collection. Interviews followed a standard semi-structured protocol, and were transcribed within five days of recording. Data were analysed using a grounded methodology approach 34 to link themes in the GPs' responses. No educational activity which might have influenced the GPs' opinions was undertaken before these interviews. The two questions asked in every interview which are pertinent here were: "Some of the literature suggests that GPs miss quite a few patients who are psychologically distressed when they consult. What do you think about that?" ' 'Why do you think GPs might miss or fail to diagnose these patients?" Results The characteristics of the GP cohort are presented in Table 2, and the comparative figures for Sheffield presented in Table 3. This shows that the GPs were younger than average for Sheffield, with a higher proportion of female GPs. These data are presented to put the qualitative data in context. All 19 GPs agreed with the literature suggesting that GPs fail to detect psychological distress in some patients. The findings at interview are presented in parallel to the literature, as doctor, patient and other factors. Doctor factors The commonest factors mentioned were GP attitude, previous experience and consultation behaviour.

4 130 Family Practice an international journal TABLE 2 Individual GP background characteristics in the study cohort Characteristic Mean age 36.8 years (SD ± 6.6) Gender ratio 8 male: 11 female Mean years in practice 7.6 (SD ± 8.0) Membership of the RCGP 10/19 Other postgraduate qualifications Consultation behaviour (10) included dimensions of active enquiry about psychosocial aspects, acting on one's own feelings, and choosing to make an effort to detect psychological problems (9). In addition, GPs mentioned the issue of inaccuracies in rating patients as distressed a particular requirement of the research study which may also relate to routine clinical care. Five GPs mentioned the likelihood of rating patients as not distressed if they themselves cannot face dealing with that aspect of care; a comment which relates to the broader issue of choice above. Five GPs also thought they might rate the patient incorrectly according to their previous relationship with the patient a patient factor explored below. Diploma in Obstetrics/ Gynaecology 10/19 Diploma in Child Health Miscellaneous 3/19 2/19 Further training in counselling or communication skills 10/19 Previous experience of mental health questionnaires such as GHQ 8/19 Previous experience of consultation analysis 15/19 TABLE 3 Comparison with average figures for Sheffield (1993 FHSA data) Characteristic Number of GPs Mean age (years) Gender ratio (male/female) Mean years in pracice Sheffield :9 9 Study cohort :11 The category of GP attitude comprised several different properties: whether the GP was orientated towards psychological problems (10 comments); what they regarded as the legitimate role of the GP (2); their willingness to label patients (2); and also the GP's frame of mind at the time of consultation (10). Previous experience (11) involved training (7), clinical experience (6) and knowledge of management options (5). 8 Patient factors Factors relating to the patient which might influence whether the GP would detect their distress fell into three main groups: background factors, pattern of presentation, and behaviour during consultation. Background factors included the dimensions of: previous relation with the GP (10 comments); personal background (9), including previous history of a mental health problem (5); and the patient's personal and cultural expectations of the GP (6). The issue of previous relationship was felt to help some GPs but to hinder others. Pattern of presentation had three dimensions: the pattern of contacts with the practice (4); deviation from the patient's normal behaviour (4); and the pattern of symptom presentation (9). In particular, a predominance of physical complaints (6) or a very routine reason for attending (2) was felt to reduce likelihood of detection. Behaviour during consultation (7 general comments) also included evidence of emotional distress (3) and response to the GP raising psychological issues (6). It was felt to be possible that patient and GP collude to avoid discussion of emotional factors (3). Structural factors The main structural factors felt to interfere with accurate detection of psychological distress were time, time of week or year, the type of appointment, the overall levels of psychological distress in the practice area, and the background culture of medical practice. Shortage of time (15 comments) was by far the commonest, though some GPs felt they performed worse in winter (4), and the first, last and any 'drop-in' surgeries of the week were quoted as times when distress might be missed (4). Practices where the clientele were thought to carry a lot of psychosocial distress, or where the workload was high (10), were perceived as likely to alter detection rates, and the norms of the practice (3) were also considered relevant. In addition two GPs felt medical culture did not encourage GPs to be good at detecting psychological problems.

5 Discussion The findings from interview show clearly that GPs are aware of many of the factors which influence detection of psychological distress. They also acknowledge the actions of the GP to be crucial in the process of accurate detection. In this sense, the opinion of GPs in the cohort broadly parallels the literature findings on why GPs may or may not detect psychological distress in their patients. However, there appears to be a difference in emphasis concerning which factors are most influential: whereas the literature, and subsequent training interventions 37 have focused on GP skills as the essential element which predicts GP performance, the GPs in this cohort were more inclined to perceive the problem as being one of time and energy to utilize skills they already possessed. One of the most interesting findings was the extent to which GPs perceived a 'choice' as to whether they did or did not gather evidence for a diagnosis of psychological distress during a consultation. In addition, GPs clearly felt that their performance would be strongly influenced by their workload and its vagaries, and also the practice environment. Conclusion It may be that the low level of GP performance as detectors of psychological distress is as much related to the constraints of time and state of mind as it is to whether GPs possess the necessary skills. More research could be done into whether detection fluctuates with length of consultation or overall duration of surgery and numbers of patients seen. Undertaking training in consultation skills is likely to improve a GP's ideal performance, but translating this into daily use may be difficult if the conditions of consultation do not allow adequate time for GPs to utilize their skills appropriately. References 1 Sartorius N et al An international study of psychological problems in primary care. Arch Gen Psychiatry 1993; 50: The Secretary of State for Health. The health of the nation: a strategy for health in England. London: HMSO, Whitehouse C. Psychosocial illness management in general practice. J RCoUGen Proa 1987; 37: Goldberg DP, Blackwell B. Psychiatric illness in general practice. Br MedJ 1970; 2: Marks JN, Goldberg DP, Hillier VF. Determinants of the ability of general practitioners to detect psychiatric illness. Psychol Med 1979; 9: Goldberg D, Williams P. A user's guide to the General Health Questionnaire. Windsor: NFER-Nelson, McDowell I, Newell C. Measuring health. Oxford: Oxford University Press, Skuse D, Williams P. Screening for psychological disorders in general practice. Psychol Med 1984; 14: Detection of psychological distress Crossley D, Myers P, Wilkinson G. Assessment of psychological care in general practice. BrMedJ 1992; 305: Boardman AP. The GHQ and the detection of emotional disorder by general practitioners: a replicated study. Br J Psych 1987; 151: Ormel J et al. Outcome of anxiety and depression. Arch Gen Psychiatry 1993; 50: Onnel J et al. Recognition, management, and course of anxiety and depression in general practice. Arch Gen Psychiatry 1991; 48: Brody DS, Lerman CE, Wolfson HG, Caputo GC. Improvement in physicians counselling of patients with mental health problems. Arch Int Med 1990; 150: Goldberg D, Huxley P. Mental illness in the community: the pathway to psychiatric care. London: Tavistock, Crombie DL. Mental illness in primary care. London: Tavistock, Creed F et al. General practitoner referral rates to district psychiatry and psychology services. BrJ Gen Pract 1990; 40: Kerr M, Blizard R, Mann A. General practitioners and psychiatrists: comparison of attitudes to depression using the depression attitude questionnaire. BrJ Gen Pract 1995; 45: Sbiber Aetal. Detection of emotional problems in the primary care clinic. Fam Pract 1990; 7: Royal Colleges of Physicians and Psychiatrists. The psychological care of medical patients. A report of the Royal Colleges of Physicians and Psychiatrists, April Goldberg D, Smith C, Steele J, Spivey L. Training family doctors to recognise psychiatric illness with increased accuracy. Lancet 1980; ii: Mechanic D. Practice orientations among GPs in England and Wales. Med Care 1970; 8: Wilmink FW. Patient, physician, psychiatrist: assessment of mental health problems in primary care. Thesis. Groningen University, Davenport S, Goldberg D, Millar T. Study of correlation between GHQ score, patient behaviour in consultation, and doctor identification of psychological distress. Lancet 1987: ii: Millar T, Goldberg DP. Link between the ability to detect and manage emotional disorders: a study of general practitioner trainees. BrJ Gen Pract 1991; 41: Verhaak PFM. Recognition of psychologic complaints by general practitioners. Med Care 1988; 26: Hjortdahl P. The influence of GP knowledge about their patients on the clinical decision-making process. ScandJ Prim Health Care 1992; 10: Jones LR, Mabe PA, Riley WT. Physician interpretation of illness behaviour. Intl J Psych in Med 1989; 19: Weich S, Lewis G, Donmall R, Mann A. Somatic presentation of psychiatric morbidity in general practice. BrJ Gen Pract 1995; 45: Ahman DG, Bland JM. Diagnostic tests 2: predictive values. Br MedJ 1994; 309: Wilson A. Consultation length in general practice: a review. Br J Gen Pract 1991; 41: Howie JGR, Porter AMD, Heaney DJ, Hopton JL. Long to short consultation ratio: a proxy measure of quality of care for general practice. BrJ Gen Pract 1991; 41: Heaney DJ, Howie JGR, Porter AMD. Factors influencing waiting times and consultation times in general practice. Br J Gen Pract 1991; 41: Howie JGR, PorteT AMD, Heaney DJ. General practitioners, work and stress. In Stress management in General Practice. Occasional Paper 61. London: Royal College of General practitioners, 1993.

6 132 Famfly Practice an international journal 34 Wilson A, McDonald P, Hayes L, Cooney J. Longer booking x Lincoln and Guba. Naturalistic enquiry. California: Sage intervals in general practice: effects on doctors' stress and Publications, California Publications, arousal. BrJ Gat Proa 1991; 41: ^ Gask L, McGrath G, Goldberg D, Millar T. Improving the 33 Hopton JL et al. The need for another look at the patient in psychiatric skills of established general practitioners: general practice satisfaction surveys. Fam Pract 1993; 10: evaluation of group teaching. Med Educ 1987; 21: