Atilla Soykan and Bedriye Oncu. Introduction

Transcription

1 Family Practice Vol. 20, No. 5 Oxford University Press 2003, all rights reserved. Doi: /fampra/cmg511, available online at Printed in Great Britain Which GP deals better with depressed patients in primary care in Kastamonu, Turkey: the impacts of interest in psychiatry and continuous medical education Atilla Soykan and Bedriye Oncu Soykan A and Oncu B. Which GP deals better with depressed patients in primary care in Kastamonu, Turkey: the impacts of interest in psychiatry and continuous medical education. Family Practice 2003; 20: Background. Little is known about the reasons behind Turkish GPs limited ability to diagnose and treat major depression. Objective. The aim of this preliminary study is to evaluate the impact of a GP s level of interest in psychiatry and participation in previous continuous medical education (CME) on their ability to recognize and treat major depression. Methods. Thirty-eight GPs from an underdeveloped city in Turkey participated in the study. A survey consisting of questions about their demographic characteristics, self-reported capacity for recognition and antidepressant management of depression, presence of previous CMEs and self-ratings of their interest in psychiatry was given to all GPs. Comparisons were made using hierarchical multiple regression analyses and SPSS software. Results. Almost half of the GPs had participated previously in at least one CME course on depression, and these were significantly more involved with the treatment of depressed patients (P = 0.02). Hierarchical multiple regression analysis indicated that interest in psychiatry was an important factor in predicting the GPs confidence in recognizing and treating depression even after controlling for other variables such as age, gender (P = 0.01) and participation in previous CME (P = 0.05). Conclusion. Our findings suggest that personal characteristics, including a GP s interest in psychiatry, should be considered when planning education and other interventions to improve the detection and treatment of depression in primary care. Keywords. Continuous medical education, depression, interest in psychiatry, primary care. Introduction In Turkey, mental health problems and mental disorders are estimated to affect 17% of the adult population; however, only 13% of those patients receive help for their psychiatric problems. 1 Underdetection of psychiatric problems in primary care is a major factor associated with the low detection and treatment rates in Turkey. In a large-scale epidemiological study, although almost half of the patients with depression were seen by a GP Received 18 November 2002; Revised 8 May 2003; Accepted 19 May Department of Consultation Liaison Psychiatry, Ankara University, School of Medicine, Ankara, Turkey. Correspondence to Atilla Soykan, Yesilyurt Sok. No:23/6, Asagi Ayranci, Ankara, Turkiye; Asoykan@pol.net for their non-psychiatric problems within the previous 12 months, only a quarter were recognized. 1 It is clear that patient characteristics affect the recognition rates of psychiatric disorders in primary care. GPs usually have higher recognition rates for those with high risk for developing psychiatric disorders such as separated/ widowed females, the unemployed and patients with significant social disability. 2 Studies suggest that GPs perceptions of mental disorders may be different from those of psychiatrists, 3,4 and participating in continuing medical education (CME) may improve GPs detection and treatment rates. 3,4 However, the effects of participating in CME do not improve all GPs detection rates uniformly. In one study, the total improvement seen after training about depression was found to be almost entirely attributable to a change in only 40% of participating GPs. 4 This finding may indicate that personal 558

2 Depression: the impacts of GPs interest in psychiatry and CME 559 characteristics such as the level of interest in psychiatry of GPs may also be important in explaining low recognition rates. Some studies suggest that patients value a GP s role as a source of first-contact care and co-ordinator of referrals. 5 On the other hand, some GPs may express less interest in psychiatry and prefer not to be involved with psychiatric cases at all. 6 A GP s prior experience with a psychiatric treatment strategy was found to be a significant factor in predicting future use of treatment, 7 as well as their consideration of themselves as the medical system s first contact for patients with psychosocial problems. 8 In Turkey, once a GP graduates from medical school, there is no obligation for her/him to participate in any educational activities. Moreover, until recently, there have been only limited CME activities available. In the last decade, the Turkish Ministry of Health included depression among the disorders that should be recognized, evaluated and, if appropriate, treated in primary care. Following this change in policy, both the Ministry of Health and pharmaceutical companies began to support a great number of CME programmes on depression and anxiety disorders in primary care. Educational activities were almost exclusively didactic in nature and given by university professors on a voluntary basis. However, the impact of these educational activities has not yet been studied in any part of Turkey. The objective of this preliminary study is to understand better the impact of participating in a CME on the recognition and antidepressant management of the depressed patients by the GPs in a small Turkish city. Additionally, the impact of a GP s level of interest in psychiatry was also assessed. Methods Location and subjects The study was planned with the local branch of the Ministry of Health and took place in September 2001 with the GPs of Kastamonu. Kastamonu is an underdeveloped Turkish city where one psychiatrist serves inhabitants. Thus, the majority of care for psychiatric patients is delivered by GPs. There were multiple reasons for choosing the city of Kastamonu. First, the authors chose a site where referrals to large city hospitals were uncommon and where many patients were treated mainly by local GPs. Secondly, to conduct the study successfully, the authors needed a strong collaboration with the local branch of the Ministry of Health in order to organize the invitation and meeting of GPs without causing major problems in their daily work schedule. Thirdly, the authors chose a site where all the GPs were employed on either a full- or-part time basis by the Ministry of Health. Kastamonu clearly fit these criteria. In the preparation phase of the study, it was decided that inviting 15 GPs on three different occasions could be accomplished with coverage of their practices by other GPs and would not significantly disrupt the care of patients. Of 202 GPs practising in Kastamonu, 45 were randomly selected and invited to participate in the study. Thirty-eight of 45 invited GPs (84.4%) participated in the study, representing 18.8% of all GPs practising in Kastamonu. Verbal consent was obtained from each participant. Instruments Three separate groups of GPs, consisting of 15 in each group, were invited to the study sessions. Sessions were conducted with 13, 12 and 13 participants. Each session started with a questionnaire on GPs demographic characteristics, their perceived capacity for recognition and antidepressant management of depression, and previous CME courses for depression. GPs also rated their interest in psychiatry and psychiatric patients as low, moderate or high. After the questionnaire was completed, a test on anxiety disorders was administered to assess their baseline knowledge on anxiety disorders prior to the educational session. Test questions consisted of six separate cases presenting characteristic DSM-IV symptoms of various anxiety disorder patients. GPs were asked to match the cases with the corresponding diagnosis. The same test was repeated after the 3 h educational session on anxiety disorders in order to see if short-term educational objectives were met. Test results were labelled as true only if all six diagnoses were correct. Statistics All analyses were performed with SPSS software (version 10). Categorical data were analysed with the Pearson chi-square test and, if indicated in the SPSS output, Fisher s exact test P-value was also given. The Pearson correlation test was used for analysis of correlations. Hierarchical multiple regression analysis was performed to test the effects of interest in psychiatry and previous CME on the dependent variables (recognition and antidepressant management of depression). Confidence intervals were 95% (two-sided) for all analyses. Results The sample consisted of 22 male and 16 female GPs (age range years, mean age ± 5.14). Most GPs were born in Kastamonu, graduated from medical school and returned to Kastamonu after graduation. All GPs had been practising medicine in Kastamonu since their graduation from medical school (mean duration 5.15 ± 4.52 years, range 1 22). Of 38 GPs, 36 (94%) believed that he/she could diagnose depression accurately. Surprisingly, only 27 (71%) reported that they had actually been involved with the treatment of a depressed patient, such as prescribing medicine, providing follow-up and/or psychosocial support of a depressed patient. The gender

3 560 TABLE 1 Family Practice an international journal Comparisons between gender, recognition and antidepressant management of depression, interest in psychiatry and replies to test questions on anxiety disorders Participated in Believes that he/she Actually involved Answer to questions previous CME can diagnose with the treatment of about anxiety disorders on depression depression depressive patients Pre-test Post-test Yes (%) No (%) Yes (%) No (%) Yes (%) No (%) True (%) False (%) True (%) Gender Male 12 (32) 10 (26) 20 (53) 2 (5) 14 (37) 8 (21) 13 (34) 9 (24) 22 (58) Female 8 (21) 8 (21) 16 (42) 0 (0) 13 (34) 3 (8) 12 (32) 4 (10) 16 (42) Participated in previous CME on depression Yes (%) 17 (44) 1 (3) *16 (42) 2 (5) 13 (34) 5 (13) 18 (47) No (%) 19 (50) 1 (3) 11 (29) 9 (24) 12 (32) 8 (21) 20 (53) Believes that he/she can diagnose depression (recognition) Yes (%) 25 (66) 11 (29) 24 (63) 12 (31) 36 (94) No (%) 2 (5) 0 (0) 1 (3) 1 (3) 2 (6) Actually involved with the treatment of depressive patients (antidepressant management) No (%) 19 (50) 8 (21) 27 (71) Yes (%) 6 (16) 5 (13) 11 (29) Interested in psychiatry and psychiatric patients High (%) 7 (18) 12 (32) **18 (47) 1 (3) ***17 (45) 2 (5) ****15 (40) 4 (10) 19 (50) Moderate (%) 11 (29) 5 (13) 16 (42) 0 (0) 8 (21) 8 (21) 10 (26) 6 (16) 16 (42) Low (%) 2 (5) 1 (3) 2 (5) 1 (3) 2 (5) 1 (3) 0 (0) 3 (8) 3 (8) *Pearson chi-square, 5.290; df 1; P = 0.02 (two-sided); Fisher s exact test: P = 0.03 (two-sided). **Pearson chi-square, 5.630; df 2; P = 0.06 (two-sided) (significant at trend level). ***Pearson chi-square, 6.611; df 2; P = 0.04 (two-sided). ****Pearson chi-square, 7.308; df 2; P = 0.03 (two-sided). of the GP was not correlated with differences in the other variables (Table 1). Almost half, 18 (47.4%), of the GPs had previously participated in at least one CME on depression. Those who participated in a depression CME activity were significantly more likely to be involved currently with the treatment of depressed patients (chi-square 5.29, df 1, P = 0.02; Fisher s exact test P = 0.03; see Table 1). Previous CME did not predict any significant differences on the true answer rate in the pre-test on anxiety disorders. The level of interest in psychiatry and psychiatric patients was significantly correlated with the treatment behaviour of GPs. The greater the interest, the more likely it was that the GP actually treated (r = 0.33, P = 0.04) depressed patients, and scored higher on the pre-test (r = 0.39, P = 0.01). Additionally, those with low interest in psychiatry differed from high and moderate interest group in terms of their confidence in diagnosing (chi-square 5.630, df 2, P = 0.06; significant at trend level) and treating (chi-square 6.611, df 2, P = 0.04) depressed patients. Moreover, GPs with low interest gave significantly fewer correct answers on the pre-test (chi-square 7.308, df 2, P = 0.03). Yet, the difference between groups disappeared after the education (Table 1). A hierarchical multiple regression analysis was conducted to test for the mediator role of CME between interest in psychiatry and recognition and antidepressant management of depression. At the first step, the set of age, gender and experience explained only 13% of the total variance, and none of these variables made a significant contribution to the prediction of recognition and antidepressant management of depression [F(3,34) = 1.62, P = NS]. After controlling for the variance accounted for by these control variables, interest in psychiatry was entered into the equation at the second step. Addition of interest in psychiatry increased the total explained variance to 28%, with interest in psychiatry being a significant predictor of recognition and antidepressant management of depression [t(33) = 2.64, P = 0.01; F(1,33) = 6.95, P = 0.01]. At the third step, CME was entered into the equation. Although the observed significance of interest in psychiatry decreased somewhat [t(32) = 2.02, P = 0.05], it remained significant, indicating that CME activity only slightly mediated the interest in psychiatry recognition and antidepressant management of depression relationship. It is important to note that after controlling for the variance explained by all of the above variables, addition of CME activity increased the total explained variance only slightly (R 2 change = 1%), with CME having no significant contribution to predicting recognition and antidepressant management of depression [F(1,32) = 0.74, P = NS]. This conservative

4 Depression: the impacts of GPs interest in psychiatry and CME 561 regression analysis supported the view that when compared with CME, interest in psychiatry has a more dominant role on recognition and antidepressant management of depression, and CME only slightly mediated the interest in psychiatry recognition and antidepressant management of depression relationship. Thus, this analysis confirms the main effect of interest in psychiatry on predicting recognition and antidepressant management of depression even after controlling for CME and all other control variables. Discussion GPs behaviour in treating depressed patients is shaped not only by educational activity but also by many other factors. GPs are not necessarily interested in and equally knowledgable about every category of illness such as diabetes, depression or asthma. Moreover, the lack of objective criteria and laboratory tests for diagnosis, the commonality of somatic symptoms masking depression, time constraints on consultation, and lack of appropriate interview skills make the recognition of depression by GPs even more difficult. 9 Yet, depression is a major public health problem in terms of disability, morbidity and economic burden on society, and it must be recognized and treated at the level of primary care. In spite of the small sample size, the results of this preliminary study indicate that when the effects of control variables and interest in psychiatry are controlled statistically, a GP s participation in a previous CME on depression does not significantly predict his/her future behaviour in the recognition and treatment of depressed patients. Yet, those GPs with previous CME activity are significantly more involved with the treatment of depressed patients. Although various kinds of CMEs were effective in increasing the recognition and treatment rates of depression by GPs, 3,4 controversies and negative findings still exist. First, follow-up measures of effective diagnosis and treatment may not be improved with CME activity. Secondly, when present, the observed improvement in recognition and treatment rates may not be long lasting in all GPs. Finally, one must explain the limited efficacy of CME on patient recognition and outcome measures. 10 This study suggests that the level of interest in psychiatry is a factor that may be associated with the limited efficacy of CME on depression. Our study suggests that interest in psychiatry is a major contributing factor to a GP s ability to recognize depression and become involved with antidepressant management of the depressed patient. After controlling for the effects of sex, age, experience and even CME, interest in psychiatry appears to be a significant variable in predicting the GP s recognition and treatment of depressed patients. Additionally, although none of the GPs had attended a CME on anxiety disorders previously, among GPs with a high interest in psychiatry, baseline knowledge of anxiety disorders was significantly better compared with those with low interest. On the contrary, the low interest group did significantly worse in the pretest assessment, were less confident in their ability to diagnose depression and were reluctant to be involved with treatment of depressed patients. No comparisons with the literature could be drawn as our search of major databases did not reveal any studies on the impact of a GP s interest in psychiatry on the treatment of depression. It is noteworthy that the majority of GPs (94.7%) believed that they can diagnose patients with depression. In contrast, even after extensive training, the actual recognition rate was 39% in the Hampshire study. 10 The implications of the discrepancy between the actual and the perceived rate of recognition observed in this study need to be studied further. In our sample, although almost all GPs believe that they can diagnose depression, only three-quarters actually attempt to treat depression. During the preparation phase of the study, we were not expecting such a dichotomy, and we did not add any questions to clarify what happens to those patients that they diagnose but do not treat. This interesting finding should be assessed in future studies. Small sample size and relying on GPs perceptions rather than objective measures are two major limitations of this preliminary study. A study with a larger sample size would definitely increase the value of the hierarchical multiple regression analysis performed in this study. Adding objective outcome measures and including more questions about possible variables affecting the GPs attitudes towards psychiatry, such as having psychiatric patients in close relatives or a personal history of psychiatric problems, and the presence of violence or suicide acts of patients while under their treatment, would also be needed for more thorough evaluations. Detailed determinants of interest in psychiatry should be addressed in future studies. It may be that developing strategies to address interest in psychiatry are just as important as the actual content of a depression CME. In conclusion, we found that interest in psychiatry is an important factor in GPs recognizing and treating depression in Kastamonu, Turkey. Larger scale studies are needed to confirm the generalizability of this finding. Acknowledgements We thank Alex Sabo, MD, for his editorial help for the context and language of this study, and Tulin Gencoz, PhD, for her statistical work. References 1 Kılıc C. The incidence, disability, related factors of mental disorders in adult population and the results of the use of mental services. In Erol N, Kılıc C, Ulusoy M, Kececi M, Simsek Z (eds), Mental

5 562 Family Practice an international journal Health Profile of Turkiye. Ankara: Eksen Tanıtım Ltd Sti., 1998: Balestrieri M, Bisoffi G, Tansella M et al. Identification of depression by medical and surgical general hospital physicians. Gen Hosp Psychiatry 2002; 24: Hannaford PC, Thompson C, Simpson M. Evaluation of an educational programme to improve the recognition of psychological illness by general practitioners. Br J Gen Pract 1996; 46: Goldberg DP, Steele JJ, Smith C et al. Training family doctors to recognize psychiatric illness with increased accuracy. Lancet 1980; 2: Grumbach K, Selby JV, Damberg C et al. Resolving the gatekeeper conundrum: what patients value in primary care and referrals to specialists. J Am Med Assoc 1999; 282: Turton P, Tylee A, Kerry S. Mental health training needs in general practice. Primary Care Psychiatry 1995; 1: Daniels ML, Linn LS, Ward N et al. A study of physician preferences in the management of depression in the general medical setting. Gen Hosp Psychiatry 1986; 8: Rabinowitz J, Feldman D, Gross R et al. Which primary care physicians treat depression. Psychiatry Serv 1998; 49: Eisenberg L. Treating depression and anxiety in primary care: closing the gap between knowledge and practice. N Engl J Med 1992; 326: Thompson C, Kinmonth AL, Stevens L et al. Effects of a clinicalpractice guideline and practice-based education on detection and outcome of depression in primary care; Hampshire Depression Project randomised controlled trial. Lancet 2000; 355: