GP group profiles and involvement in mental health carejep_1597

Transcription

1 Journal of Evaluation in Clinical Practice ISSN GP group profiles and involvement in mental health carejep_1597 Marie-Josée Fleury PhD, 1 Jean-Marie Bamvita MD PhD, 2 Lambert Farand MD PhD, 3 Denise Aubé MD MSc FRCPC, 4 Louise Fournier PhD 5 and Alain Lesage MD MPhil 6 1 Associate Professor, Department of Psychiatry, McGill University, Douglas Mental Health University Institute Research Centre (DMHUIRC), Montréal; Québec, Canada 2 Research Associate, Douglas Mental Health University Institute Research Centre (DMHUIRC), Montréal, Québec, Canada 3 Associate Professor, Department of Health Administration, Faculty of Medicine, Université de Montréal, Montréal, Québec, Canada 4 Clinical Professor, Department of Social and Preventive Medicine, Laval University, National Public Health Institute of Québec, GRIOSE-SM (CSSS de la Vieille-Capitale et CHRG), Québec, Canada 5 Associate Professor, Department of Social and Preventive Medicine, Université de Montréal, and Researcher, CRCHUM, Montréal, Québec, Canada 6 Professor, Department of Psychiatry, Université de Montréal, Centre de Recherche Fernand-Seguin, Hôpital Louis-H. Lafontaine, Montréal, Québec, Canada Keywords chronic disease management, general practitioner profile in mental health, mental health care system, primary mental health care, shared care Correspondence Dr Marie-Josée Fleury Department of Psychiatry McGill University Douglas Mental Health University Institute Research Centre (DMHUIRC) 6875 LaSalle Boulevard Montréal, QC H4H 1R3 Canada flemar@douglas.mcgill.ca Accepted for publication: 10 September 2010 doi: /j x Abstract Rationale and objectives Mental health is one of the leading causes of morbidity worldwide. Its impact in terms of cost and loss of productivity is considerable. Improving the efficiency of mental health care system has thus been a high priority for decision makers. In the context of current reforms that privilege the reinforcement of primary mental health care and integration of services, this article brings new lights on the role of general practitioners (GPs) in managing mental health, and shared-care initiatives developed to deal with more complex cases. The study presents a typology of GPs providing mental health care, by identifying clusters of GP profiles associated with the management of patients with common or serious mental disorders (CMD or SMD). Methods GPs in Quebec (n = 398) were surveyed on their practice, and sociodemographic data were collected. Results Cluster analysis generated five GP profiles, including three that were closely tied to mental health care (labelled, respectively: group practice GPs, traditional pro-active GPs and collaborative-minded GPs), and two not very implicated in mental health (named: diversified and low-implicated GPs, and money-making GPs). Conclusion The study confirmed the central role played by GPs in the treatment of patients with CMD and their relative lack of involvement in the care of patients with SMD. Study results support current efforts to strengthen collaboration among primary care providers and mental health specialists, reinforce GP training, and favour multi-modal clinical and collaborative strategies in mental health care. Introduction The World Health Organization estimates that by 2030 mental disorders may be the leading cause of morbidity in industrialized countries and second-leading worldwide after HIV/AIDS [1]. Clearly, mental disorders constitute a major public health issue. In industrialized countries, their prevalence ranges from 4.3% to 26.4% annually [2], and their economic burden (treatment cost and loss of productivity) is considerable. In Canada, they rank among the most costly illnesses [3]. General practitioners (GPs) represent the first point of contact for patients with mental disorders. In the course of a single year, about 80% of the population in industrialized countries consult a GP: roughly a third of these patients suffer from a detectable mental disorder [4]. In Québec, 92% of these patients are diagnosed with common mental disorders (CMD) (e.g. depression and anxiety) and 8% with serious mental disorders (SMD) (e.g. schizophrenia and bipolar disorders) [5]. In Canada, among patients with mental disorders, 45% consult a GP while about 25% consult other health care practitioners [6]. This preference for GPs is due to proximity, greater accessibility, less stigmatization and a holistic approach in which physical problems are also managed [7]. General practitioners treat patients with CMD more frequently than patients with SM [8]. Contrary to most patients with CMD, patients with SMD face major functional disability including concurrent substance abuse and physical problems (e.g. diabetes, obesity and hypertension) and require substantial, long-term help Blackwell Publishing Ltd, Journal of Evaluation in Clinical Practice 18 (2012)

2 M.-J. Fleury et al. GP profiles and involvement in mental health [9]. SMD are less prevalent (2 3% of the population) than CMD, but they generate close to half of total mental health care costs [10]. GPs greater confidence in treating patients with CMD than with SMD is due to their belief that SMD is too complex for routine primary care. Case seriousness and complexity aside, lack of support from psychiatrists and insufficient mental health care training are the leading causes given for transferring patients to specialized care. In addition, time constraints, fear of aggression, anticipated problems between crises, and patient-compliance difficulties are reasons provided for transferring patients to specialists [11,12]. The burden of SMD has prompted countries to improve their mental health care system by strengthening communitybased services and primary care [13]. In mental health care, the benefits of strengthening primary care by controlling and minimizing the need for specialized care are strongly recognized [14]. Enhancing primary care by implementing shared-care models, involving greater coordination between GPs, psychiatrists and mental health psychosocial professionals, has been recommended [15,16]. Enhanced efficiency and patient outcomes are cited as benefits of a strong, integrated primary care system [14,15]. Continuing medical education (CME) and other multi-modal incentives (e.g. clinical guidelines, payment incentives and strong policy support) have also been found to enhance GPs ability to detect, diagnose and treat patients with mental disorders [17]. Despite current health care reforms designed to boost primary care and service integration, most GPs report they have little contact with mental health care professionals [18]. A number of factors contribute to solo practice, common among GPs, including (i) long waiting time for psychiatric expertise; (ii) poor perception among GPs of inter-organizational collaboration and mental health care system quality; and (iii) high overall demand for medical services [12]. In Canada, particularly in Québec, this situation is aggravated by a major shortage of GPs. Reportedly, 25% of the population in Québec is without a GP; these patients use walk-in clinics and emergency rooms for medical services [19]. To enhance the role of GPs in the treatment of patients with mental disorders, a better understanding is needed of the GP profiles that enable or hinder their involvement in such treatment. Are there clusters of GP profiles that incite them to take on more or fewer patients with CMD or SMD? We have found no previous study that identified different GP profiles dealing with mental disorders. For example, do GPs who are older, work mainly in solo private practice, have frequent contact with mental health care professionals, and see fewer patients overall take on more patients with mental disorders? Conversely, do GPs who are younger, practice mainly in walk-in clinics, attend relatively few CME sessions, and see more patients overall take on fewer patients with mental disorders on an ongoing (rather than a one-time) basis? Increasing attention is devoted worldwide to optimizing integrated primary care models; accordingly, this study presents a typology of GPs mental health care practice. While the study focuses on Québec and Canada, its findings are of relevance to many industrialized countries, which share similar reform objectives as well as organizational and practice features. By identifying GP profiles that enable or hinder their involvement in mental health care, the study may contribute to improvements in service planning. Background: Québec/Canada primary care and GP mental health partners In Canada, health care is a provincial jurisdiction and has been regionalized over the past two decades. Under the Canada Health Act (1984), all residents are entitled to free inpatient or outpatient medically required care at the point of delivery. Patients receive treatment at publicly funded facilities or are seen by GPs or private specialists in the community who charge their provincial health plan for their services. The ratio of GPs per inhabitants in Canada is among the lowest in OECD countries [20]. Québec has a population of 7.5 million and 7199 full-time GPs (1 GP per 1041 inhabitants); however, it has approximately 10% more GPs than most Canadian provinces [21]. In Canada, GPs are mainly paid through fees for services, but also partly through salary or hourly fees. In a recent survey, 23% of GPs reported working in solo practice, 51% in group practice, and 24% in multidisciplinary team practice [22]. In Québec, family practice groups (FMGs), networks clinics (NCs) and community health centres (known as CLSCs) favour multidisciplinary team practice. FMGs and NCs are new primary care settings, launched in 2001 to improve care continuity, particularly for patients with chronic diseases. FMGs involve several GPs working closely together with nurses responsible for patient screening, follow-up, referral and patient registration. NCs are similar to FMGs, except that patients are not registered with GPs and nurses act mainly as liaison agents. In 2008, close to 20% of population in Québec was registered in FMGs [23]. In community health centres, GPs can coordinate with mental health services (i.e. psychosocial services or mental health teams, comprising diversified psychosocial professionals specializing in mental health problems). About 50% of GPs also work in hospitals and close to 40% reported working in walk-in clinics [24]. One study [25] estimated that walk-in clinics were a regular source of care for 60% of patients in some Québec regions. Other than GPs, Québec mental health services include: (1) psychiatrists (about 1000, of whom 45% practice in Greater Montréal) or psychiatric teams working in hospitals or out-patients clinics; (2) community health centres (170 CLSCs); (3) psychologists in private clinics (about 8000); and (4) the voluntary sector (i.e. community-based agencies and crisis centres, about 400). Since 2005, Québec has launched reforms designed to reinforce primary care and service integration among GPs, psychiatrists and psychosocial services (i.e. shared care initiatives) [26]. Methods This cross-sectional study focuses on GPs in Québec. Selected territories represent urban, semi-urban, and rural areas and encompass university-affiliated and more remote centres. Diverse practice settings were included: solo or group practice in private clinics, community health centres, family medicine groups, network clinics, hospitals and walk-in clinics. The sample was extracted from a list provided by the Québec Federation of General Practitioners (FMOQ), the union that represents GPs in the province, in consultation with an advisory committee consisting of health care decision makers. All 1415 practising GPs in the selected territories were considered potential participants in the 2010 Blackwell Publishing Ltd 397

3 GP profiles and involvement in mental health M.-J. Fleury et al. study. The research was approved by a university research ethics board. Each participant signed a consent form. Data collection A self-administered 143-item questionnaire was designed by the research team and validated by a multidisciplinary group of twenty experts (researchers, GPs and psychiatrists). Wherever possible, RAMQ (Régie de l assurance maladie du Québec) data from 2006, including GPs fee for service reimbursement data, were compared with questionnaire results. The questionnaire covered six domains: (1) GP socio-demographic and attitudinal profile; (2) patient characteristics; (3) clinical practice features; (4) GP collaboration with other medical or psychosocial mental health care professionals; (5) GP perception of mental health service quality; and (6) GP opinion regarding supportive strategies for better care integration. The questionnaire categorized mental disorders as: (i) CMD, including depression and/or anxiety, adaptation disorders, personality disorders and substance abuse; or (ii) SMD, excluding the above disorders and including three examples, schizophrenia, bipolar disorder and delusional disorder. The questionnaire, which required 30 minutes for completion, consisted mainly of categorical or continuous items, with some 5-point Likert Scale questions (1 = strongly disagree to 5 = strongly agree). It was mailed to GPs from September 2005 to February 2007 along with supporting letters from the Québec College of Physicians and FMOQ. There were three follow-ups: letters; telephone calls from nurses; and personal contact by medical administrators. Details on the questionnaire can be found in another publication [27]. Analyses and definition of variables Initially, descriptive analyses were performed. Frequency distribution for categorical variables and mean values for continuous variables were computed. GP profiles were generated using cluster analysis. Twelve variables were selected for cluster analysis, based on their potential to illustrate GP practice. A literature review on GP practice and shared care was carried out to establish a baseline for variable selection. The research advisory committee conducted a review of variables, based on management change-enabling criteria. With regard to GPs socio-demographic and attitudinal profile, four variables were considered (gender, age, income and CME). One variable related to patient characteristics was included: proportion of medical visits for mental disorders. Five variables represented clinical practice features: number of practice settings; main practice setting; patients with CMD taken on; patients with SMD taken on; and number of patients seen per week. Finally, two variables were extracted with regard to GPs collaboration with mental health care partners: number of contacts with mental health professionals per year and importance attributed to inter-professional collaboration in mental health care. The GP age was divided into five categories in accordance with Canadian data banks: 35, 36 45, 46 55, and >65. Income referred to the proportion of GPs salary from fees for services: 0 24%, 25 49%, 50 74% and %. Classification by proportionality was also used to reflect the distribution of patients taken on by GPs for CMD or SMD. The term taking on refers to more than seeing patients on a one-time basis: it implies follow-up over time for the same or subsequent condition, including medical tests (physical and/or mental health), medication, side effect monitoring, psychotherapy or any kind of psychosocial support. Taking on patients means that GPs accept them as part of their clientele. The percentage of patients taken on by GPs for CMD and SMD was based on their answer to the following question: Among patients seen for mental disorders in your medical practice every week, what is the proportion of patients with CMD (or with SMD) whom you follow on a continuous basis? CME was measured based on the number of half-day mental health sessions attended in the previous year. The proportion of patients visiting for mental disorders was computed as the ratio of the number of mental health-related visits (diagnostics or medical act) to total patient visits. This measure is based on a typical week and includes patients managed on an ongoing basis and seen on a one-time basis only. The number of practice settings was determined from GPs account of hours spent at each setting. Main practice settings was defined as the site where GPs spent 50% of their time or more. The number of total patients seen by GPs was based on a typical week, regardless of the patient diagnosis or medical act dispensed. Contacts referred to direct communications in the past 12 months between GPs and mental health care providers, including phone calls, patient follow-up and shared-care activities. Finally, the perceived importance among GPs of inter-professional collaboration was based on their responses on a scale regarding mental health providers in Québec, including psychiatrists, psychiatric teams, community health centres, psychologists in private practice, community-based agencies and crisis centres. For cluster analysis, inter-subject distance was measured with the Log-likelihood method. Schwartz Bayesian criteria were used to classify participants. Qualitative variables were multi-categorical, except for gender. Continuous variables were standardized before clustering. Profiles were automatically generated with spss Statistics 17.0 (SPSS Inc., Copyright IBM Corporation 2010 IBM Corporation, Route 100 Somers, NY 10589, USA). Variables were entered in descending order of importance. Categorical variables were entered first (patients with CMD and SMD taken on, gender, age, main practice setting and income), followed by continuous variables (proportion of patients visiting for a mental disorder, number of patients seen per week, importance attributed to interprofessional collaboration in mental health, CME, number of practice settings and contacts). Importance attributed to interprofessional collaboration in mental health care was not retained in the final model as no sufficient difference was found between groups. Results Sample The sample comprised 398 GPs, for a response rate of 41%. Our sample was compared with non-respondent GPs and overall GP population in Quebec and Canada considering important parameters (sex, age, clinical practice settings, etc.); no significant difference was found in any of these comparisons. Further details of the sample s characteristics are available in another publication [27]. GP socio-demographic and practice features General practitioners were at 49% men and at 51% women. Mean age was 48 years, with about three-quarters aged years Blackwell Publishing Ltd

4 M.-J. Fleury et al. GP profiles and involvement in mental health Table 1 General distribution of participating general practitioners (GPs) 1. GP socio-demographic characteristics Frequency % Gender, n (%) Male Female Age categories, n (%) All, mean (SD) years old years old years old years old >65 years old Income from fee for services, n (%) All, mean (SD) % % % % Patient characteristics Proportion of patients visiting for mental disorders, mean (SD) Patients with CMD taken on, n (%) All, mean (SD) % % % % Patients with SMD taken on, n (%) All, mean (SD) % % % % Clinical practice features Main practice setting, n (%) Solo private practices Medical group clinics Community health centres Family medicine groups Network clinics Walk-in clinics and emergency rooms (n = 29) Hospitals Number of practice settings, mean (SD) Number of patients seen per week, mean (SD) Number of CME sessions attended in MH in the 12 past months, mean (SD) Collaboration between GPs and other professionals Contacts with mental health professionals per year, mean (SD) Importance attributed to inter-professional collaboration, mean (SD) CMD, common mental disorders; CME, continued medical education; MH, mental health; SMD, serious mental disorders; SD standard deviation. (Table 1). More than half of GPs earned % of their income from fees for services; only 15% did not earn fee-for-service income. Generally, GPs reported practising in two settings. Walk-in clinics and emergency rooms were the main setting for about 30% of GPs. On average, GPs saw 90 patients per week, including 25% with mental disorders (standard deviation: 19). Among patients with mental disorders, a vast majority was diagnosed with CMD (60% with depression and/or anxiety). GPs reported taking on almost all patients with CMD, but only a minority of patients with SMD. To enhance their mental health care expertise, on average GPs attended 3 days of CME sessions in the past 12 months. Globally, GPs had little direct contact with mental health care providers: 55% had contacts with psychologists; 51% with community health centres; 48% with psychiatric services; and 6% with community-based agencies and crisis centres. Up to 14% of GPs had no contact at all with other professionals. On a point-scale ranging from 21 to 64, the mean score for the importance attributed to inter-professional collaboration between GPs and other mental health care professionals was 50 points. GP profiles in mental health care Five distinct GP profiles dealing with mental disorders were identified. From the sample of 398 GPs, 76 (19%) were automatically discarded by the classification program. The remaining 322 participants were clustered into five profiles (Table 2). Profile 1 (n = 76, 24%) includes mainly mid-aged female GPs who receive fees for services and practice in medical group clinics and family medicine groups. Many patients see them for mental disorders. GPs in this profile take on the second-largest proportion of patients with either CMD or SMD. They also attended the second-highest number of CME session days in the previous year. Finally, they placed third with regard to the number of patients 2010 Blackwell Publishing Ltd 399

5 GP profiles and involvement in mental health M.-J. Fleury et al. Table 2 Cluster analysis of Québec general practitioners (GPs) GP characteristics Profile 1 (n = 76) Profile 2 (n = 76) Profile 3 (n = 67) Profile 4 (n = 55) Profile 5 (n = 48) Total (n = 322) Gender, n (%) Male 21 (13.2) 76 (47.8) 6 (3.8) 21 (13.2) 35 (22.0) 159 (100.0) Female 55 (33.7) 0 (0.0) 61 (37.4) 34 (20.9) 13 (8.0) 163 (100.0) Age categories (years), n (%) 35 1 (3.4) 2 (6.9) 12 (41.4) 2 (6.9) 12 (41.4) 29 (100.0) (11.9) 10 (9.2) 54 (49.5) 16 (14.7) 16 (14.7) 109 (100.0) (41.8) 32 (23.9) 0 (0.0) 30 (22.4) 16 (11.9) 134 (100.0) (4.5) 32 (72.7) 1 (2.3) 6 (13.6) 3 (6.8) 44 (100.0) >65 4 (66.7) 0 (0.0) 0 (0.0) 1 (16.7) 1 (16.7) 6 (100.0) Income from fee for services, n (%) 0 24% 0 (0.0) 6 (7.7) 12 (15.4) 53 (67.9) 7 (9.0) 78 (100.0) 25 49% 7 (21.9) 5 (15.6) 12 (37.5) 2 (6.3) 6 (18.8) 32 (100.0) 50 74% 11 (36.7) 7 (23.3) 10 (33.3) 0 (0.0) 2 (6.7) 30 (100.0) % 58 (31.9) 58 (31.9) 33 (18.1) 0 (0.0) 33 (18.1) 182 (100.0) Main practice setting, n (%) Solo private practices 14 (34.1) 19 (46.3) 8 (19.5) 0 (0.0) 0 (0.0) 41 (100.0) Medical group clinics 31 (59.6) 18 (34.6) 0 (0.0) 1 (1.9) 2 (3.8) 52 (100.0) Community health centres 0 (0.0) 0 (0.0) 1 (3.8) 25 (96.2) 0 (0.0) 26 (100.0) Family medicine groups 18 (32.7) 13 (23.6) 16 (29.1) 5 (9.1) 3 (5.5) 55 (100.0) Network clinics 3 (17.6) 9 (52.9) 3 (17.6) 0 (0.0) 2 (11.8) 17 (100.0) Walk-in clinics (and emergency rooms) 3 (3.1) 17 (17.5) 30 (30.9) 7 (7.2) 40 (41.2) 97 (100.0) Hospitals 7 (20.6) 0 (0.0) 9 (26.5) 17 (50.0) 1 (2.9) 34 (100.0) Patients with CMD taken on, n (%) 0 24% 0 (0.0) 0 (0.0) 0 (0.0) 6 (18.2) 27 (81.8) 33 (100.0) 25 49% 1 (20.0) 0 (0.0) 2 (40.0) 2 (40.0) 0 (0.0) 5 (100.0) 50 74% 6 (33.3) 0 (0.0) 7 (38.9) 4 (22.2) 1 (5.6) 18 (100.0) % 69 (25.9) 76 (28.6) 58 (21.8) 43 (16.2) 20 (7.5) 266 (100.0) Patients with SMD taken on, n (%) 0 24% 32 (18.7) 25 (14.6) 40 (23.4) 26 (15.2) 48 (28.1) 171 (100.0) 25 49% 14 (53.8) 5 (19.2) 3 (11.5) 4 (15.4) 0 (0.0) 26 (100.0) 50 74% 8 (21.6) 16 (43.2) 9 (24.3) 4 (10.8) 0 (0.0) 37 (100.0) % 22 (25.0) 30 (34.1) 15 (17.0) 21 (23.9) 0 (0.0) 88 (100.0) Continuous variables, mean (SD) Number of patients seen per week 98.4 (30.3) (33.0) 83.4 (37.0) 54.1 (32.2) (42.8) 93.2 (40.7) Proportion of patients visiting for mental disorders (MH) 23.6 (16.1) 23.2 (11.2) 22.6 (15.1) 41.6 (30.3) 17.5 (11.4) 25.5 (19.2) Contacts with other professionals per year 0.7 (0.7) 1.1 (1.1) 0.5 (0.6) 1.6 (1.2) 1.1 (1.3) 1.0 (1.0) Number of CME sessions attended in 3.5 (2.8) 2.5 (2.0) 1.9 (1.3) 4.7 (3.7) 2.5 (3.5) 3.0 (2.9) MH in the 12 past months Number of practice settings 2.0 (0.7) 2.1 (0.8) 2.3 (0.9) 1.8 (0.7) 1.6 (0.6) 2.0 (0.8) Typology of GP profiles general profile Private group-practice GPs actively involved in mental health (with moderate patient volume) Private practice GPs actively involved in mental health (with high patient volume) GP labelled profile Group practice GPs Traditional proactive GPs Diversified practice GPs (with low patient volume) Diversified and low-implicated GPs Multidisciplinary-team practice GPs with high consultation in mental health (but low patient volume) Collaborative-minded GPs Mainly walk-in clinic GPs less actively involved in mental health care (with high patient volume) Money-making GPs CMD, common mental disorders; CME, continued medical education; MH, mental health; SMD, serious mental disorders; SD standard deviation Blackwell Publishing Ltd

6 M.-J. Fleury et al. GP profiles and involvement in mental health seen per week, number of practice settings and frequency of contact with other professionals. This profile, encompassing mainly GPs in private group practices actively involved in mental health care with moderate patient volume, was named: group practice GPs. Profile 2 (n = 76, 24%) comprises mainly older male GPs earning fees for services and working mainly in network clinics and solo private practices. This profile reported the highest volume of patients seen per week and proportion of patients with CMD or SMD taken on. It places second with regard to the number of practice settings and frequency of contact with other professionals, and ranks third for the proportion of patients with mental disorders and number of CME sessions attended in the previous year. Profile 2, including chiefly GPs in private practices actively involved in mental health care with high patient volume, was labelled: traditional pro-active GPs. Profile 3 (n = 67, 21%) represents mainly young female GPs with balanced sources of income (fees for services and salary). They have the highest number of practice settings without any setting being predominant. They place third as to the proportion of patients with CMD taken on. They rank fourth with regard to: the number of patients seen per week for any reason; proportion of patient visits specifically for mental disorders; proportion of patients taken on for SMD; CME attendance in the previous year; and frequency of contact with other professionals. This profile, encompassing principally GPs in diversified practices with low patient volume, was denominated: diversified and low-implicated GPs. Profile 4 (n = 55, 17%) consists mainly of mid-aged female GPs earning over 50% of their income from salary and working in community health centres and hospitals. Their clientele includes the greatest proportion of patients with mental disorders. They rank first for frequency of contact with other professionals and number of CME sessions attended in the previous year. They rank third for patients with SMD taken on. They place fourth with regard to patients with CMD taken on and number of practice settings. They see the lowest number of patients per week. This profile, including chiefly GPs in multidisciplinary team practices with intensive mental health care consultation but low patient volume, was labelled: collaborative-minded GPs. Profile 5 (n = 48, 15%) encompasses mainly young male GPs who receive fees for services and practice in walk-in clinics. They have the lowest number of practice settings, proportion of patients visiting for mental health reasons, and patients taken on for CMD and SMD; yet, they rank second for the number of patients seen per week for any reason and frequency of contact with other professionals. They place third for CME attendance in the previous year. This profile, encompassing primarily GPs in walk-in clinics who are less actively involved in mental health care but have high patient volume, was named: money-making GPs. Discussion This study was designed to present a typology of GPs practising in mental health care by identifying clusters of GP profiles associated with the management of patients with mental disorders (CMD or SMD). It is based on a sample of 398 GPs, surveyed in and representative of the GP population in Québec regarding overall socio-demographic and clinical practice features. As stated in previous studies [16], GPs reportedly play a pivotal role in managing patients with CMD. In our sample, only a minority of GPs took on patients with SMD. As noted elsewhere [27], GPs are not generally at ease taking on patients with SMD. Increasingly, GPs practice in physician-group clinics or multidisciplinary team practices, as endorsed by current reforms [28]; however, in Québec many GPs practice mainly in solo and have little contact with mental health care providers. Solo practices and walk-in clinics, which are key sites for GPs in Québec, are not conducive to treating patients with complex conditions such as SMD. To our knowledge, this is the first study to identify distinct profiles of GP practice in mental health care. Five such profiles, representing 81% of GPs in our sample (322/398), were found: (1) private-group-practice GPs and (2) private-practice GPs, both actively involved in mental health care; (3) diversified-practice and low-implicated GPs ; (4) multidisciplinary-team-practice GPs who actively consult with regard to mental health care (labelled: collaborative-minded GPs ); and (5) walk-in-clinic GPs less actively involved in mental health care (labelled: money-making GPs ). These profiles take into account eleven variables, selected on the basis of their potential to illustrate GP practice and of management change-enabling issues in the health care system (i.e. gender, age, income, CME, number of and main practice settings, patient volume seen per week, proportion of patients with mental disorders, patients with CMD or SMD taken on, and contacts with mental health care providers). Three profiles (1, 3 and 4) were more closely associated with female GPs; and two (2 and 5) with male GPs. Two profiles were more closely associated with older GPs (1 and 2); and two with younger GPs (3 and 5); with no age-specific association for profile 4. Profiles 1 and 2, encompassing older male and female GP groups, were actively involved in mental health care, especially with regard to complex cases. Profiles 1 and 2 were the largest group, representing close to half of the study s sample. Conversely, profiles 3 and 5, including younger male and female GPs, were not very actively involved in mental health care. Profile 4 GPs, predominantly mid-aged and female, reported the largest proportion of mental health care cases among their patients, but this profile exhibited only average propensity to take on patients with CMD or SMD. As this group tends to work in multidisciplinary team practices (community health centres and hospitals) and attends CME sessions more assiduously, we hypothesized that profile 4 physicians play a consulting role for GPs who are less at ease with mental health care and, specifically, shared-care models. In all, most GP profiles are involved in mental health care (profiles 1, 2 and, somewhat less, 4; 64%), which is consistent with the literature focusing on the central role of GPs in mental health care. As other studies have found [29], our results show that male GPs usually have higher patient volumes than do their female counterparts. Family reasons aside, this difference may be due to the fact that more men practice in walk-in clinics, as the profile 5 shows [30]. Given the association between older GPs (either male or female) and a higher proportion of patients with mental disorders taken on (especially in the case of SMD), the explanation may be that seniority and expertise enable GPs to treat patients with complex clinical profiles [31]. Patients with mental health problems, particularly SMD, may require more care, time, and consultations and be more difficult to treat [32]. Often, they have concurrent diagnoses (e.g. substance abuse) and interrelated physical or social problems [33]. GPs generation practice style may also 2010 Blackwell Publishing Ltd 401

7 GP profiles and involvement in mental health M.-J. Fleury et al. be at play, as it is hypothesized that younger clinicians may be less willing to make a long-term commitment to patients, especially in complex cases. These patients may also be seen as less profitable to younger GPs in the midst of establishing their practice. Group-practice models and shared-care initiatives constitute superior settings for treating patients with complex conditions [15,16]. GPs in our sample who treated patients with mental disorders did so in such settings, including family medicine groups, network clinics, private clinic group practices and community health centres. They also saw these patients in solo practice, which is generally associated with ongoing care continuity and patient satisfaction [34]. Conversely, GPs at walk-in clinics and diversified practices (profiles 3 and 5), including mainly younger male and female GPs, were associated with the lowest proportion of patients seen for mental disorders and the lowest proportion of patients with SMD taken on. For newly certified GPs, walk-in clinics or diversified practices may represent a means to acquire sufficient knowledge and experience to take on more complex cases subsequently. Walk-in clinics are also known as favouring access to care rather than continuity of care [35]. Although it is closely related with practice settings, mode of payment was not associated with GPs involvement in mental health care in our sample. Fee for services, however, is linked to greater patient volume and productivity, compared with other modes of payment, and favours short-term and multiple consultations. The literature offers conflicting results on the impact of GPs modes of payment: they may have an impact on clinical practice, but less so than other incentives [36]. Training and contacts are more difficult to interpret with respect to GPs practice in mental health care. In our study, GPs who attended more CME sessions in the previous year tended to be younger (profiles 4 and 5) or actively involved in mental health care (profiles 1, 2 and 4). As previous studies have demonstrated [17,37], training enhances GPs ability to deal with patients with mental disorders generally in the long run, when knowledge, confidence and clinical tools mesh into multi-modal strategies. A high incidence of contact between GPs and other professionals may be indicative either of shared-care initiatives (profiles 2 and 4) or patient transfer (profile 5 mainly walk-in clinic GPs). Profile 3 ( diversified and low-implicated GPs ) ranked fifth with regard to frequency of contact with other professionals and CME sessions attended and fourth with regard to the proportion of patients visiting for a mental disorder and the taking on patients with SMD. Conversely, GPs in profile 4 ( collaborative-minded GPs ) reported the highest frequency of contact with other professionals, the greatest number of CME sessions in the previous year, and the largest proportion of patients seen for mental disorders. Strength and limitations This study is of value as it is the first to our knowledge to provide a typology of GPs practice in mental health care, based on a representative sample in Québec. As little research has been published on GPs management of mental disorders, this study may be useful to decision makers involved in implementing reforms. It does, however, present some limitations. First, it is cross-sectional, which precludes causal inferences. Second, the response rate was relatively low, but not lower than in other studies on GPs. Third, as data were collected from self-reporting GPs, results must be viewed as an approximation of actual GP practice. Fourth, our findings are based on a sample of GPs in Québec; other studies are thus needed to know in which extend Québec GP profiles can be founded elsewhere. Finally, no data were collected on GPs adequacy in treating patients with mental disorders (not considered in the cluster model), although this is considered a major issue [16]. Conclusion Following a survey of 398 GPs in Québec, five GP profiles related to mental health care were identified, labelled, respectively: (1) group practice GPs; (2) traditional pro-active GPs; (3) diversified and low-implicated GPs; (4) collaborative-minded GPs; and finally (5) money-making GPs. As three out of five profiles (1, 2 and 4) play a central role in the treatment of mental disorders, our findings confirm current knowledge regarding GPs involvement in mental health care, especially with respect to CMD. However, no profile stood out as being more actively involved with patients who have serious disorders. Profile 5, the money-making profile encompassing GPs working mainly in walk-in clinics, was reportedly the least actively involved in mental health care. As suggested by profiles 1, 2 and 4, a number of characteristics favour GPs involvement in mental health care: seniority in primary care, previous mental health care CME attendance, and group practice settings (most assuredly not walk-in clinics). Unfortunately, in Québec, GPs tend to work in solo practices and shared care is still underdeveloped. Profile 4, which regroups GPs in multidisciplinary team practices, seems to play more of expert-consultant role in mental health care. This profile may provide a base for more shared-care models, which GPs in our study strongly support. In light of our results, increased collaboration among primary care providers and mental health care specialists and further training and use of multi-modal clinical or collaboration strategies in mental health care should be promoted to help GPs meet current needs and facilitate the implementation of chronic diseasemanagement and shared-care models. These recommendations are relevant for Québec, other Canadian provinces and OECD countries that share similar health care features and reform objectives, namely to enhance integrated primary care and promote innovative system-change strategies. Acknowledgements This study was funded by the Canadian Institute of Health Research (CIHR), Fonds de la recherche en santé du Québec (FRSQ) and other decision-making partners. We would like to thank all the grant agencies and our partners (especially Drs Carrier, Trudeau, Rodrigue, Marcoux, Lambert, and Ms Gérome and Mr La Roche), the general practitioners who took part in the study, our research coordinator, Youcef Ouadahi, and other co-investigators of this research (Drs Tremblay, Lamarche and Lussier and Mr Poirier). References 1. Mathers, C. & Loncar, D. (2006) Projections of global mortality and burden of disease from 2002 to PloS Medicine, 3 (11), e The WHO World Mental Health Survey Consortium (2004) Prevalence, severity, and unmet need for treatment of mental disorders in Blackwell Publishing Ltd

8 M.-J. Fleury et al. GP profiles and involvement in mental health the World Health Organisation Mental Health Surveys. Journal of the American Medical Association, 291 (21), Lim, K.-L., Jacobs, P., Ohinmaa, A., Schopflocher, D. & Dewa, C. S. (2008) Une nouvelle mesure, fondée sur la population, du fardeau économique de la maladie mentale au Canada. Maladie Chronique au Canada, 28 (3), Watson, D. E., Heppner, P., Roos, N. P., Reid, R. J. & Katz, A. (2005) Population-based use of mental health services and patterns of delivery among family physicians, 1992 to Canadian Journal of Psychiatry, 50 (7), Ouadahi, Y., Lesage, A., Rodrigue, J. & Fleury, M.-J. (2009) Les problèmes de santé mentale sont-ils détectés par les omnipraticiens? Regard sur la perspective des omnipraticiens selon les banques de données administratives. Santé mentale au Québec, numéro spécial sur les soins primaires et la santé mentale, 34 (1), Gagné, M.-A. (2005) Avancement des objectifs des soins de santé mentale axés sur la collaboration. Mississauga, Ontario: Initiative canadienne de collaboration en santé mentale. 7. Rothman, A. A. & Wagner, E. H. (2003) Chronic illness management: what is the role of primary care? Annals of Internal Medicine, 138 (3), Fleury, M.-J., Bamvita, J.-M. & Tremblay, J. (2009) Variables associated with general practitioners taking on serious mental disorder patients. BMC Family Practice, 10, Nelson, G. (2006) Mental health policy in Canada. Canadian Social Policy: Issues and Perspectives, 4th edn (ed. A. Westhues), pp Waterloo: Wilfrid Laurier University Press. 10. VGQ. (2003).Rapport annuel du vérificateur général du Québec à l Assemblée Nationale pour l année Québec: Gouvernement du Québec. 11. Hodges, B., Inch, C. & Silver, I. (2001) Improving the psychiatric knowledge, skills, and attitudes of primary care physicians, : a review. American Journal of Psychiatry, 158 (10), Kates, N., Fugere, C. & Farrar, S. (2004) Family physician satisfaction with mental health services: findings from a community survey. Canadian Psychiatric Association Bulletin, 36 (2), WHO (2001) The World Health Report 2001 Mental Health: New Understanding, New Hope. Geneva: World Health Organization. 14. Starfield, B. (1998) Primary care visits and health policy. Canadian Medical Association Journal, 159 (7), Craven, M. A. & Bland, R. (2006) Better practices in collaborative mental health care: an analysis of the evidence base. Canadian Journal of Psychiatry, 51 (6 Suppl. 1), 7S 72S. 16. Walters, P., Tylee, A. & Goldberg, D. (2008) Psychiatry in primary care. In Essential Psychiatry (eds R. M. Murray, P. McGuffin, K. S. Kendler, S. Wessely & D. J. Castle), pp Cambridge: Cambridge University Press. 17. Williams, J. W. Jr, Gerrity, M., Holsinger, T., Dobscha, S., Gaynes, B. & Dietrich, A. (2007) Systematic review of multifaceted interventions to improve depression care. General Hospital Psychiatry, 29 (2), Fleury, M.-J., Bamvita, J.-M., Tremblay, J. & Lesage, A. (in press) Extent and determinants of general practitioner referrals and contacts with mental health Care Providers. Canadian Journal of Community Mental Health. 19. Gladu, F.-P. (2007) La pénurie réelle ou ressentie de médecins de famille au Québec. Peux-t-on y remédier? Canadian Family Physician, 53 (11), Eurostat (Online Database) and OECD Health Data (2008) (for Canada, United States and Japan). Available at: health/healthcare/docs/echi_63_88_pra_en.pdf (last accessed 15 November 2010). 21. Savard, I. & Rodrigue, J. (2007) Des omnipraticiens à la grandeur du Québec. Évolution des effectifs et des profils de pratique. Données de à Québec: Direction de la planification et de la régionalisation FMOQ. 22. Collège des Médecins de Famille du Canada (2008) Soutenir les effectifs futurs en médecine familiale au Canada. En fait-on assez aujourd hui pour se préparer pour demain? Mississauga, Ontario: College des médecins de famille du Canada. 23. MSSS (2009) Évaluation de l implantation et des effets des premiers groupes de médecine de famille au Québec. Québec: Ministère de la Santé et des Services Sociaux. 24. Collège des Médecins de Famille du Canada (2004) Sondage nationale 2004 auprès des médecins. Ottawa: Collège des médecins de famille du Canada, Association Médicale Canadienne, Collège royal des médecins et chirurgiens du Canada. 25. Haggerty, J., Pineault, R., Beaulieu, M. D., Brunelle, Y., Goulet, F., Rodrigue, J. & Gauthier, J. (2004) Continuité et accessibilité des soins de première ligne au Québec: Barrières et facteurs facilitants. Ottawa: Fondation canadienne de la recherche sur les services de santé (FCRSS). 26. Fleury, M. J. (2009) Présentation. Santé mentale en première ligne. Revue santé Mentale au Québec, numéro spécial sur les soins primaires et la santé mentale, 34 (1), Fleury, M. J., Bamvita, J.-M., Farand, L. & Tremblay, J. (2008) Variables associated with general practitioners taking on patients with common mental disorders. Mental Health in Family Medicine, 5 (3), Beaulieu, M. D., Rioux, M., Rocher, G., Samson, L. & Boucher, L. (2008) Family practice: Professional identity in transition. A case study of family medicine in Canada. Social Science & Medicine, 67 (7), Contandriopoulos, A. P. & Fournier, M. A. (2007) Féminisation De la profession médicale et transformation de la pratique au Québec. Montréal: GRIS, Université de Montréal. 30. Williams, A. (2002) Changing geographies of care: employing the concept of therapeutic landscapes as a framework in examining home space. Social Science & Medicine, 55 (1), Kravitz, R., Franks, P., Feldman, M., Meredith, L. S., Hinton, L., Franz, C., Duberstein, P. & Epstein, R. M. (2006) What drives referral from primary care physicians to mental Health specialists? A randomized trial using actors portraying depressive symptoms. Journal of General Internal Medicine, 21 (6), Ballester, D. A., Filippon, A. P., Braga, C. & Andreoli, S. B. (2005) The general practitioner and mental health problems: challenges and strategies for medical education. Sao Paulo Medical Journal, 123 (2), Younes, N., Gasquet, I., Gaudebout, P., Chaillet, M. P., Kovess, V., Fallissard, B. & Hardy Bayle, M. C. (2005) General practitioners opinion on their practice in mental health and their collaboration with mental health professionals. BMC Family Practice, 6 (1), Pineault, R., Levesque, J. F., Roberge, D., Hamel, M. & Couture, A. (2008) L accessibilité et la continuité des services de santé: une étude sur la première ligne au Québec. Ottawa: Instituts de Recherche en Santé du Canada (IRSC) et Fondation Canadienne de la Recherche sur les Services de Santé (FCRSS). 35. Barbeau, D., Perron, S., Ouellet, V. & Morency, P. (2001) Un regard critique sur les services de consultation sans rendez-vous. Le Médecin du Québec, 36 (1), Shortt, S. E. D. & Stanton, S. (2001) Does Changing the Way Doctors are Paid Change the Way They Practice? Evidence from an Ontario Academic Health Science Centre. Ottawa: Canadian Health Services Research Foundation. 37. Fournier, L., Aubé, D., Roberge, P., Lessard, L., Duhoux, A., Caulet, M. & Poirier, L. R. (2008) Vers une première ligne forte en santé mentale: Messages clés de la littérature scientifique. Québec: Institut national de santé publique du Québec Blackwell Publishing Ltd 403