Abstract
The prevalence of major depression among patients in primary medical practice is about twice that found in general community surveys [1], [2]. Approximately 60% of people in the community with an affective disorder seek professional assistance, with most (76%) consulting their general practitioner (GP) [3]. Research suggests however, that depression is under-recognized and undertreated in primary medical care [4], [5]. Although some believe that routine screening of patients in primary medical care for depression has little effect on doctor behaviours or patient outcomes [6], other studies suggest that routine brief screening does have positive outcomes, provided that GPs follow-up positive screening results with a more detailed clinical assessment, together with the provision of evidence-based treatments [7–9].
A range of patient, doctor and practice factors contribute to the under-recognition and less than optimal management of depression in general practice. Patient factors include comorbidity and embarrassment; doctor factors include inadequate knowledge and skills; practice factors include inadequate consultation time and insufficient access to specialized mental health resources [4], [10].
An important patient barrier to accurate diagnosis of depression is the presentation of depression and anxiety as somatic complaints. Up to 80% of patients with a diagnosable depressive or anxiety disorder present with somatic complaints [11], which may affect GPs' ability to recognize the origin of the patient's problem as depression or anxiety [12–14]. More recently, a brief screening instrument has been developed that accurately and sensitively detects both the psychological and somatic manifestations of anxiety and depression [7].
As expected, GP knowledge and skills appear to influence the recognition and appropriate management of depression [4], [10]. A large-scale practice audit in Australia found that GPs with a declared interest in mental health and those who had obtained mental health training were more likely to see more patients with depression and more likely to provide appropriate mental health assessment and treatments [15]. In the UK, Marks et al. [16] found that GPs who had a better conceptual understanding of mental illness produced a more accurate diagnosis of the patient's condition. They also noted that GPs with an interest in psychological medicine, with higher levels of empathy and those who asked about social and family problems, more accurately diagnosed psychiatric illness. These results are consistent with further data from the Somatic and Psychological Health Report (SPHERE) national mental health project [17].
The specific beliefs and attitudes that GPs hold toward depression and depressed patients also appear to impact upon their diagnosis and management of this condition. Botega et al. [18] identified three groups of GPs based on their attitudes toward antidepressants/psychotherapy, identification of depression and on their professional unease in relation to depression. The first group of GPs were more likely to distinguish between depression and unhappiness, felt more comfortable and rewarded in working with depressed patients, were less likely to prescribe antidepressant medication and more likely to believe that depressed patients could benefit from psychotherapy. The second group were distinguished by a strong belief that antidepressant medications were more effective than psychotherapy and that depression was an organic disease. These doctors prescribed more medication than doctors in the other two groups. The third group were less likely to find the treatment of depressed patients rewarding and were more likely to refer depressed patients to other health professionals for psychotherapy rather than attempt psychotherapy themselves.
Similarly, Kerr et al. [19] found that low-dose prescribing GPs believed more strongly in psychotherapy and were less likely to believe that antidepressants were useful in general practice and that depression has an organic, biomedical basis. General practitioners who reported greater professional unease in managing depressed patients and a stronger belief in the inevitability of the course of depression, particularly with advancing age, were less likely to follow up depressed patients, a result also found by Ross et al. [20]. However, GPs who have a preference for psychotherapy rather than antidepressant treatment also appear more accurate in diagnosing depression [21]. Despite these outcomes, there is a little research into how mental health training impacts upon these GP attitudes.
Accurate identification and management of depression by GPs may also be the product of the GP's confidence in his/her ability to accurately diagnose and treat depression. For example, Docherty [22] found that lower levels of confidence in the effectiveness of available treatments for depression were associated with poorer recognition of depression. Gerrity et al. [23] also suggested that doctor self-efficacy for recognizing and managing depression was likely to influence the care depressed patients received from their primary care physician.
In summary, there are numerous doctor, patient and practice factors which facilitate or impede GPs' abilities to accurately diagnose and manage depression in their patients. Among the most prominent are GPs' attitudes toward depression, confidence on the part of the GP in relation to this area of practice, symptom somatization on the part of the patient, patient reluctance to accept depression as a diagnosis and to accept appropriate treatment and lack of time and funding for GPs to treat depression. Access to appropriate assessment instruments and appropriate mental health specialist resources may also be significant barriers, particularly for GPs outside the major urban areas. Although data on barriers to the effective management of depression in primary care derived from GP practice audits [e.g. 15] are useful, comparative information from large-scale surveys of what GPs themselves consider to be the significant barriers to the effective assessment and management of depression provides another perspective on the issue. This study involved a direct survey of urban and rural GPs. In particular, our study investigates the impact of prior mental health training on GPs' attitudes to depression, on their confidence in relation to managing depression and on the barriers they identify in the effective management of this condition. It also investigates the impact of these variables on what GPs say about their current clinical practice in relation to the management of depression.
Method
Procedure and participants
In 2001, the Australian Divisions of General Practice (ADGP) endorsed this project and wrote to all Divisions of General Practice requesting their participation. The 52 Divisions that responded were classified as either urban or rural and were requested to forward questionnaire packages (10 for rural and 15 for urban) to a representative sample of their GPs, attempting to balance gender, age and interest in mental health issues. The packages contained an anonymous questionnaire, a reply-paid envelope and a plain language statement which outlined the purpose of the research. The questionnaire took approximately 20 minutes to complete. A total of 420 GPs (69%) completed and returned the surveys. Each GP who returned the questionnaire was paid $40 and his/her Division, $10.
Fifty-three percent of responding GPs were female, 71% were aged between 35 and 54 years and 46% were from rural divisions. More than one-third (37.1%) were Fellows of the Royal Australian College of General Practitioners (RACGP), 11.2% were Fellows of the Australian College of Rural and Remote Medicine (ACRRM), 25.2% had other family medicine qualifications and 41% other postgraduate qualifications. Just under one in 10 (8.3%) had completed or were currently completing postgraduate mental health qualifications. Almost one half had completed some mental health training in the last 5 years (43.6%) and about one third (34.0%) had undertaken training in non-pharmacological treatments for depression and anxiety. General practitioners saw on average 28.7 (SD = 12.4) patients per day with the average consultation length being 16.5 minutes (SD = 6.3).
Measures
Many of the items in the questionnaire (which is available from the authors upon request) were constructed for this study through consultation with a reference group of GPs assembled by ADGP. They provided feedback on the basis of their clinical experience about the relevance and acceptability of the items to GPs. The survey form also included questions that were constructed from previous research [13], [23]. The questionnaire consisted of the following sections.
Section 1 collected demographic data such as GPs' gender, age, years experience in general practice, postgraduate qualifications, and whether they had completed mental health training in the last 5 years.
In the second section, GPs reported on their clinical experience with depressed patients over the previous 6 months (e.g. estimated the percentage of patients who had presented with somatic symptoms of depression and/or anxiety). They also indicated how often they provided specific treatments (e.g. cognitive and/or behaviour therapy and older antidepressant drugs).
The next section consisted of questions about perceived barriers that limit GPs' capacity to care for patients with depression. These included items relating to practice difficulties (e.g. poor reimbursement for time spent managing depression), to the GP him/herself (e.g. incomplete knowledge of available treatments for depression) and to the patient (e.g. patient inability/unwillingness to discuss depressive symptoms).
A further section consisted of eight items where GPs rated their level of confidence (self-efficacy) in relation to the assessment and treatment of depression (e.g. ‘treating depression and/or anxiety with evidencebased psychological treatments’ and ‘liaising with psychiatrists’).
General practitioners' attitudes to depression were assessed by the 20-item Depression Attitude Questionnaire (DAQ) [18], which was designed specifically to assess GPs' conceptualization of depression, their attitudes about the experience of treating depressed patients as well as their views about treatment options [18], [21].
General practitioners' attitudes to depression were also assessed using the Health Attitudes About Depression Scale (HAADS), developed by the SPHERE project. General practitioners rated their level of agreement with 13 statements about the treatment of depression.
The total number of items from the preceding five sections of the survey form was large. Therefore, in order to look for commonalities among items and to reduce the number of variables, item scores in each of the sections were subjected to principal components analysis (PCA) using varimax rotations. That is, PCAs were performed on items in the DAQ and HAADS, on items measuring self-efficacy, items measuring perceived barriers to care, and items measuring current clinical practice. Factor scores were then calculated for each new variable and were used in the statistical analyses.
Data analyses
Multivariate analysis of variance (MANOVA) was conducted with the factor scores from each section of the questionnaire as dependent variables. The independent variables in the first MANOVA s were gender and possession of postgraduate mental health qualifications. The independent variables in the second MANOVA s were practice location and postgraduate mental health qualifications. Independent variables in the third set of MANOVA s were gender, location and completion of any mental health training in the previous 5 years. Statistical significance in the MANOVA s was followed up with ANOVA s on each factor score.
Six standard multiple regression analyses were also conducted. The independent variables were the factor scores from the DAQ (4 factors), the HAADS (3 factors), the items measuring barriers to care of depressed patients (3 factors), and the items measuring confidence in skills to assess and treat depression (3 factors). The dependent variables were measures of the GPs' current clinical practice. They included three questions that enquired about percentages of patients with depression/ anxiety who presented to the surgery. The remaining three dependent variables asked about GPs' estimates of the percentages of their patients presenting with high prevalence mental disorders whom they referred to other GPs with mental health expertise, psychiatrists and psychologists. In addition, the attitudes and self-efficacy factors were regressed onto the three perceived barriers to care factors and onto two current practice factors (Table 1). In all, 11 regression analyses were conducted.
Multivariate analysis of variance (MANOVA) on principal components analysis factors: significant effects
Results
Attitudes toward depression
Depression Attitudes Questionnaire (DAQ)
A four-factor solution best fitted these data (Table 1). Factor 1, ‘GP helpless’ was best described by the attitude that depression is rather inevitable for some people and there is little GPs can do to assist these people. The second factor, ‘GP effort’, was best described by GPs having difficulty diagnosing and working with depressed people partly because they are ‘heavy going’. Factor 3 was best described by an attitude that psychosocial interventions are likely to be of use for depressed people. Factor 4 was best described by attitudes that depression is a biochemical abnormality that responds well to antidepressants.
MANOVA (Table 1) demonstrated that completion of mental health training was related to attitudes toward depression on the DAQ with those with training both more likely to believe that people with depression could be helped (factor 1) and to feel less strain and stress in assisting these patients (factor 2). Urban and rural GPs did not differ on these attitudinal variables. Gender was however, significantly related to attitudes toward depression. Female GPs were more likely to see depression as a biochemical abnormality that responds well to medication (factor 4).
Health Attitudes About Depression Scale (HAADS)
Principal components analysis found that a three-factor solution best fitted the data (Table 1). Factor 1, ‘GP avoidance’, was best described by an attitude that little can be done for depressed people and that they should not be encouraged to see their GP and discuss depression. An attitude that it is stressful and unrewarding seeing patients with depression and other psychosocial problems, such as alcohol and drug abuse, described factor 2 (GP non-satisfaction). Factor 3 was best described by the attitude that there are insufficient supports for GPs to assess and treat depression.
MANOVA (Table 1), demonstrated that those GPs who had undertaken mental health training in the past 5 years had more positive attitudes to depression, believing that GPs could help people with depression (factor 1) and that it could be satisfying work (factor 2). Urban and rural GPs did not differ on any of these attitudinal variables. However, male GPs were more likely than female GPs to believe that GPs could not help patients with depression.
Confidence in skills for managing depression (self-efficacy)
Table 2 shows the percentages of GPs who said they were mostly or very confident about various aspects of the management of depression. Only about one-quarter to one-third of GPs were confident in managing depression with non-pharmacological evidence-based treatments. A PCA performed on these eight items found a three-factor solution best fitted the data (Table 1). Factor 1 was described by confidence in delivering non-pharmacological interventions. The second factor was described by confidence in recognizing and assessing suicide risk and treating it with medication. The third factor was described by confidence in liaising with psychiatrists and psychologists, presumably for issues of consultation and referral.
General practitioners' confidence in managing depression and identification of barriers to care
MANOVA (Table 1) demonstrated that urban GPs were more confident in delivering non-pharmacological treatments but rural GPs were more confident in liaising with psychiatrists and psychologists. General practitioners with formal mental health qualifications were also more confident in recognizing and assessing suicide risk and treating it with medication and were more confident in delivering non-pharmacological treatments, as were GPs with any mental health training. There was a significant interaction effect between gender and mental health training in general. Female GPs who had not completed any mental health training in the past 5 years most lacked confidence in delivering non-pharmacological interventions.
Barriers to management of patients with depression
Table 2 details the most common perceived barriers to the effective management of depression in their general practice. Principal components analysis of these items resulted in a three-factor solution best fitting the data (Table 1). The first factor, ‘practice variables’, was described by items such as heavy workload, inadequate time and reimbursement. The second factor was described by patient variables such as reluctance to accept a diagnosis and treatment for depression and be treated for depression. The third factor was described by doctor variables, such as incomplete knowledge of diagnostic criteria, assessment tools and treatments for depression, as well as discomfort discussing psychological issues with patients.
MANOVA (Table 1) demonstrated that GPs with postgraduate qualifications were less likely to identify incomplete knowledge about the assessment and treatment of depression as a barrier (factor 3). Similarly, undertaking mental health training in the past 5 years was related to doctor barriers with those having undertaken training also differing on factor 3 (as above). Rural and urban GPs did not differ in their identification of barriers.
The three self-efficacy variables and the total of seven attitudinal variables from the two depression attitude questionnaires were regressed on to each of the three barrier factors in turn. For the practice barriers factor, the overall regression equation was significant (F10,346 = 3.19, p = 0.00), accounting for 5.8% of the total variance. Those GPs who believed that it was stressful and unrewarding seeing depressed patients (factor 2, HAADS) and those who considered working with depressed patients was ‘heavy going’ (factor 2, DAQ) were more likely to identify practice and organizational barriers to the effective management of depression (β = 0.12, t = 2.06, p = 0.04 and β = 0.16, t = 2.58, p = 0.01, respectively).
For the patient barriers factor, the overall regression equation was significant (F10,346 = 5.06, p = 0.00), accounting for 10.2% of the variance. Only the three factors from the HAADS predicted this barrier. That is, patient barriers were more likely to be identified by GPs who: avoided dealing with depressed patients (β = 0.22, t = 3.83, p = 0.00); found working with depressed patients unsatisfying, (β = 0.17, t = 2.94, p = 0.00); and who said there were insufficient supports for the assessment and treatment of depression, (β = 0.12, t = 2.17, p = 0.03). For the doctor barriers factor, the overall regression equation was significant (F10,346 = 13.51, p = 0.00), accounting for 26.0% of the variance. All three self-efficacy factors were related to the perception of doctor barriers. That is, incomplete knowledge of diagnostic criteria, assessment and treatment methods for depression and discomfort in discussing psychological issues were more likely to be identified as barriers by GPs who: lacked confidence in the use of non-pharmacological treatments (β = 0.28, t = 5.72, p = 0.00); lacked confidence in assessing and treating suicide risk (β = 0.22, t = 4.52, p = 0.00); and by those who lacked confidence in liaising with mental health specialists (β = 0.11, t = 2.41, p = 0.02). As well, these barriers were more likely to be identified by GPs who: felt there was little they could do to help depressed patients (factor 1, DAQ; β = 0.16, t = 2.80, p = 0. 01); regarded them as ‘heavy going’ (factor 2, DAQ; β = 0.21, t = 3.96, p = 0.00); and contrary to expectations, by those who considered that it was not stressful and unrewarding seeing depressed patients (factor 2, HAADS; β = − 0.14, t = − 2.61, p = 0.01).
Current clinical practice
Table 3 presents the reported percentage of patients presenting with psychological or somatic symptoms of depression/anxiety and the percentage with a specific diagnosis of depression/anxiety. Significant predictors of percentage of people presenting with psychological symptoms were: greater confidence in the use of non-pharmacological treatments (β = − 0.12, t = − 2.03, p = 0.04); greater identification of practice barriers to care, (β = 0.11, t = 2.08, p = 0.040); and better knowledge of assessment and treatment methods for depression (β = − 0.14, t = − 2.21, p = 0.03).
Current clinical practice by general practitioners regarding mental health patients
Significant predictors of percentages presenting with somatic symptoms of depression/anxiety were confidence in non-pharmacological treatments (β = − 0.19, t = − 3.18, p = 0.00) and confidence in liaising with mental health professionals (β = − 0.15, t = − 2.81, p = 0.01). Significant predictors of percentages with a specific diagnosis of depression/anxiety were better knowledge of assessment and treatments methods (β = 0.14, t = − 2.29, p = 0.02) and on the HAADS, less avoidance of depressed patients (β = − 0.18, t = − 2.93, p = 0.00) and more satisfaction in working with depressed patients (β = − 0.14, t = − 2.32, p = 0.02).
Table 3 also details referral patterns to other health professionals. There were no significant predictors of referral to other GPs with mental health expertise, probably because most doctors do not make such referrals (0.8% of patients). General practitioners who identified more patient barriers (β = 0.16, t = 2.74, p = 0.01) and doctor barriers (β = 0.20, t = 3.23, p = 0.00), were more likely to refer depressed patients to psychiatrists. General practitioners who believed GPs could do little for depressed patients (factor 1, DAQ) also were more likely to refer patients to psychiatrists (β = 0.19, t = 2.84, p = 0.01). Confidence in liaising with mental health professionals predicted frequency of referrals to psychologists (β = − 0.19, t = − 3.39, p = 0.00), as did the belief that working with depressed patients is hard work (factor 2, DAQ; β = 0.13, t = 1.93, p = 0.06).
Table 4 details the frequency with which the respondents said they provided specific treatments for depression. Principal component analysis of these 11 items produced a two-factor solution (i.e. ‘Non-Pharmacological Treatments-NPT’ and ‘Pharmacological Treatments’ [Table 1).
Treatments offered by general practitioners regarding mental health patients
MANOVA (Table 1) demonstrated that GPs with postgraduate qualifications more often delivered NPT and that rural GPs were more likely to prescribe medication for depression than urban GPs. General practitioners with mental health training in the past 5 years were more often delivering NPT. Gender was related to current clinical practice with women more often delivering NPT than men and men more often prescribing medication than women.
The three self-efficacy factors and the seven depression attitude factors were together regressed on to the pharmacological intervention factor but there were no significant predictors (F6,369 = 1.76, p = 0.11). However, together these 10 variables significantly predicted the non-pharmacological intervention factor (F10,336 = 21.17, p = 0.00), accounting for 36.8% of the total variance. Those GPs who more often used psychosocial treatments were: more confident in their use (β = − 0.49, t = − 10.61, p = 0.00); more confident in assessing and treating suicide risk (β = − 0.15, t = − 3.24, p = 0.00); and more confident in liaising with mental health professionals (β = − 0.11, t = − 2.59, p = 0.01). They were also less likely to believe that little could be done for depressed patients and to avoid treating them (factor 1, HAADS; β = − 0.14, t = − 2.73, p = 0.01).
Discussion
Prior mental health training
We found that that prior GP education in assessment and treatment of mental health disorders was related to GPs' positive attitudes about depressed patients. General practitioners who had undertaken this training more strongly believed that depression was not inevitable, that GPs can diagnose and help depressed patients, that it can be satisfying work and were less likely to believe that working with depressed patients should be avoided. These GPs, as well as those with formal postgraduate mental health qualifications, were less likely to identify incomplete knowledge of diagnostic criteria and assessment and treatment methods for depression as barriers to the effective care of these patients. Similarly, GPs with any form of mental health training (including those with formal qualifications) reported they more often used non-pharmacological treatments for depression and anxiety. Furthermore, all forms of mental health training were associated with greater confidence in using these methods. However, only GPs with formal postgraduate mental health qualifications said they felt more confident in diagnosing and treating suicidal patients. Our results therefore elaborate on other literature [4], [15], [19] and in particular expand upon results from the SPHERE studies that GPs who have a declared interest in and have undertaken mental health training, more often provide psychological treatments for their patients [5], [15].
Location of practice
There were no significant differences between rural and urban GPs in their attitudes toward depressed people or in the frequency with which they identified practice, patient and doctor barriers to the effective care of patients with depression. Nevertheless, rural GPs more often prescribed medication for depression and anxiety than urban GPs, who in turn, were more confident in using nonpharmacological treatments. However, rural GPs were more confident in liaising with mental health professionals than urban GPs. This difference may be attributable to the greater use of telemental health in regional Australia; it may reflect closer and better-established professional networks in regional Australia; or it may be a result of these GPs being less confident in their non-pharmacological management of depressed patients.
Gender of the GP
Consistent with results from the SPHERE studies [15], [24], female GPs were more likely to provide mental health treatments to their depressed patients. That is, our data showed that in general, male GPs believed more strongly than female GPs that little can be done for depressed patients and that they should not be encouraged to consult their GP with these problems. Female GPs said they more often used non-pharmacological treatments for depression and anxiety than male GPs, who conversely said they more often prescribed medication than female GPs. This set of results is quite internally consistent. Seemingly conflicting results however, were that female GPs believed more strongly than males that depression is a biochemical abnormality that responds well to medication and female GPs with no mental health training lacked most confidence in delivering non-pharmacological treatments. These results need follow-up for clarification, because there seem to be contradictions between what female GPs believe about depression and what they say they do in practice, particularly subsequent to specific mental health training.
Barriers to the care of depressed patients
General practitioners who believed that the treatment of depressed patients was stressful, unrewarding and involved considerable effort, more often identified practice or organizational barriers. Those who avoided the treatment of depressed patients, who derived little satisfaction from it and who felt there were insufficient supports for such treatment, identified greater patient barriers. Unsurprisingly, those who lacked self-confidence in how to manage depressed patients, who correspondingly felt there was little GPs could offer these patients and who felt their management involved considerable effort were more likely to identify doctor barriers.
Current clinical practice
Contrary to other literature [e.g. 11,13], GPs in the present study did not identify an overly high percentage of their depressed patients as presenting primarily with somatic symptoms. This result also contrasts with that of Davenport et al. [15] although their incidence figures are probably related to the use of an instrument specifically developed to detect patients with somatic symptoms of depression and anxiety.
General practitioners who reported better knowledge of and confidence in the use of assessment and nonpharmacological treatments estimated they saw more depressed patients, which is consistent with other studies [21], [22]. Furthermore, these GPs were less likely to believe that depressed patients should be discouraged from consulting their GP and they considered that working with these patients was rewarding and satisfying, which again is similar to the results of other studies [21]. General practitioners who identified more practice barriers to care of depressed patients reported they saw more of these patients, perhaps because those who see more depressed patients are more sensitized to these types of barriers.
As might be expected, the reported use of nonpharmacological treatments was related to confidence in their use, to confidence in diagnosing and treating suicide risk and to confidence in liaising with mental health specialists. General practitioners who used these treatments were also less likely to believe that GPs should avoid the treatment of depressed patients.
General practitioners who believed there was little they could do to assist depressed patients, who lacked knowledge and comfort in the assessment and treatment of depression and who considered such patients resistant to treatment, were more likely to refer them to psychiatrists [c.f. 18]. Those GPs who believed that the management of depressed patients is hard work and who were confident liaising with mental health specialists, were more likely to refer patients to psychologists.
Limitations of the study
Our findings need to be understood within the limitations of the study. The survey instrument was developed for this study and has not been validated. The study sample contains a greater proportion of female practitioners than the national workforce and a greater proportion of practitioners in the 35–54 age-group, but the average number of hours worked by the GPs in the study sample was similar to that of the national workforce [25]. Although we requested Divisions of General Practice to distribute questionnaires to GPs across the spectrum of interest in mental health issues, we are unsure of the representativeness of the sample in this domain. Consequently, the generalizability of the results to the wider GP population may be somewhat limited.
Conclusions
The present study extended the findings of other reported studies, by examining the relationships of specific mental health training, practice location and GP gender to GP attitudes toward depression, confidence in managing depressed patients, identification of barriers to care and their reported current clinical practice.
Prior mental health training for GPs appears related to their attitudes toward depressed patients, the degree to which they identify barriers to the care of depressed patients, and their confidence in assessing and managing depression. In turn, their attitudes toward depression are related to their diagnostic rates for the common mental disorders. However, many different forms and types of mental health training were reported by the GPs, ranging from that leading to a formal qualification, to weekend courses, to supervision of cases. Further research should investigate the cause–effect relationships of different types and amounts of training to observable changes in GP behaviour and measurable changes in patient outcomes, particularly since studies have suggested that training programs may not improve patient management. For example, mental health education programs for GPs in the UK have not had significant effects on patient outcomes [26] despite reports by the GPs that their detection rates and management effectiveness had improved [27]. By contrast, Naismith et al. [17] found that the intensive SPHERE program resulted in an increase in diagnostic rates of common mental disorders, confidence in management skills, provision of mental health treatments and emphasis on use of NPT. This program involved intensive support for GPs, together with ongoing education and skills-based training, which proved to be more effective than clinical practice audits.
Rural GPs may benefit from training in assessment and treatment of depression, consistent with data from Davenport et al. [15] that diagnostic rates for patients presenting with psychological symptoms are higher in urban practices. Involvement in training may impact on female GPs' confidence in delivering NPT, but also on the practice of male GPs. However, cause–effect has to be determined.
Finally, other data show that better-educated patients are provided with more appropriate treatments, probably because they demand more from their doctors [15]. As mental health literacy improves [28], a goal of beyondblue, patients in general practice will demand evidence-based treatments for common mental disorders. Our study suggests that education, incorporating skills training [17], may influence management.
Footnotes
Acknowledgements
We thank ADGP and staff at individual Divisions and GPs who completed the questionnaires for their support. Funding was provided by beyondblue: the National Depression Initiative.