Collaboration between general practitioners (GPs) and mental healthcare professionals within the context of reforms in Quebec

Abstract

Background In the context of the high prevalence and impact of mental disorders worldwide, and less than optimal utilisation of services and adequacy of care, strengthening primary mental healthcare should be a leading priority. This article assesses the state of collaboration among general practitioners (GPs), psychiatrists and psychosocial mental healthcare professionals, factors that enable and hinder shared care, and GPs’ perceptions of best practices in the management of mental disorders. A collaboration model is also developed.

Methods The study employs a mixed-method approach, with emphasis on qualitative investigation. Drawing from a previous survey representative of the Quebec GP population, 60 GPs were selected for further investigation.

Results Globally, GPs managed mental healthcare patients in solo practice in parallel or sequential follow-up with mental healthcare professionals. GPs cited psychologists and psychiatrists as their main partners. Numerous hindering factors associated with shared care were found: lack of resources (either professionals or services); long waiting times; lack of training, time and incentives for collaboration; and inappropriate GP payment modes. The ideal practice model includes GPs working in multidisciplinary group practice in their own settings. GPs recommended expanding psychosocial services and shared care to increase overall access and quality of care for these patients.

Conclusion As increasing attention is devoted worldwide to the development of optimal integrated primary care, this article contributes to the discussion on mental healthcare service planning. A culture of collaboration has to be encouraged as comprehensive services and continuity of care are key recovery factors of patients with mental disorders.

Introduction

Mental disorder is a leading cause of morbidity worldwide. Its cost and impact on productivity and individual life expectancy are substantial.¹ Along with addiction, it is a major cause of work absenteeism and now accounts for more lost workdays in the USA and Canada than physical ailments.^2,3 Its prevalence is high, ranging from 4.3% in China to 26.4% in the USA, and almost 11% in Canada;^4,5 but only a third or so of affected individuals seek care.⁶ Mental disorder is also often associated with concomitant disorders: substance abuse or physical problems (most commonly diabetes and cardiovascular disease).⁷

The prevalence and impact of mental disorders have spurred the search for more efficient modes of management, particularly in light of budgetary constraints and rising healthcare costs. Best clinical practices and collaborative models designed to strengthen primary care, patient empowerment and professional co-ordination are promoted, as they yield better health outcomes.⁸ General practitioners (GPs) are the main target of current primary healthcare reforms, being at the entry point to primary care, and responsible for much of care co-ordination and patterns of service use. In light of rising demand, they play a pivotal role in the provision of mental healthcare. International studies show that roughly a third of GPs’ consultations consist of individuals with a detectable mental disorder, representing one of the most important groups to seek GPs’ advice.^9,10 Primary mental healthcare is also viewed as more accessible, less stigmatising and more comprehensive since it manages both physical and mental ailments.¹¹ GPs’ ability to detect, diagnose and treat patients with mental disorders, however, is often considered unsatisfactory.^12,13

To help GPs manage patients with mental disorders more effectively, efforts are being made to develop models of integrated care, for example, stepped care, shared care, patient-centred medical home approach and chronic care. These models include compatible features that may be combined (e.g. best practices, GP group practices and inter-disciplinary work). Stepped care implies a progression, based on patient need profiles, from light to intensive treatment or to low-cost community-based treatment before acceding to high-cost institutional or specialised services.¹⁴ Shared care involves co-ordination among GPs, psychiatrists, psychosocial mental healthcare professionals and the voluntary sector (e.g. peer self-help groups, food banks). Several taxonomies of shared-care models exist.^15,16 The patient-centred medical home approach focuses on the GP–patient relationship and co-ordination and access to care.¹⁷ Leading physician groups in the USA and Canada have endorsed this model as a means of improving primary care.¹⁸ With its focus on improving GPs’ clinical practice and team collaboration, the chronic-care model encompasses holistic care. It revolves around: (1) organisation of services and delivery, (2) patient self-management support, (3) clinical decision support, (4) clinical information systems development, (5) use of community resources (either professionals or services), and (6) community-inclusive healthcare organisation.¹⁹ These integrated care models acknowledge the considerable interdependence among providers. Optimally effective integrated care models include a broad range of management and clinical tools such as case management, clinical protocols, training and patient-centred approaches. They are also supported by major government mental healthcare policy initiatives.^16,20

In numerous industrialised countries, major primary mental healthcare reforms are underway. They are designed to improve access, continuity and quality of care in the management of mental disorders. In Quebec (i.e. the second-most-populated province in Canada, with ∼ 7.9 million inhabitants − 23% of the Canadian population), mental healthcare was restructured starting in 2005 within 95 local healthcare networks, which constitute the core of the Quebec healthcare system, where providers combine primary and specialised care services (at the regional level – 18 regions) to ensure a comprehensive care spectrum.²⁰ Within each of the 95 local healthcare networks, patients and staff were transferred from hospitals to primary care centres referred to as health and social service centres (HSSCs), where mental healthcare teams were implemented. In an effort to co-ordinate mental health resources, single-access points are in operation in the 95 HSSCs serving more than 50,000 inhabitants, where mental healthcare teams can provide evaluation, psychotherapy and follow-up or referral to psychiatric services. In some large hospitals, evaluation liaison modules (ELMs) were established for the purpose of providing rapid psychiatric assessment and treatment recommendations in association with HSSC single-access points. Shared care has also been promoted in all Quebec local networks, with a few psychiatrists (known as ‘responding psychiatrists’) appointed to assist GPs and HSSC mental healthcare teams. GPs were encouraged to transform their practice and enrol in family medicine groups (FMGs), networks clinics (NCs) or HSSCs. Group practices are favoured in these settings. The province is home to 7,199 full-time equivalent GPs (one GP per 1,041 inhabitants).²¹ FMGs involve several GPs working together (n = 8–10 full-time equivalents) with nurses responsible for patient screening, follow-up, referral and patient rostering. NCs are similar to FMGs (n = 7 or more full-time equivalents), except that patients are not rostered with GPs, and nurses act mainly as liaison agents. In HSSCs and hospitals, interdisciplinary teams are on site.

In support of primary mental healthcare reforms, this article aims to assess the state of collaboration between GPs and mental healthcare providers; factors that enable and hinder integrated care, and best-practice models promoted by GPs to improve the management of mental disorders. Based on the study findings, a collaboration model was also developed. The Quebec/Canada mental healthcare system offers an interesting setting to explore these topics, as it has undergone major reforms designed to improve primary mental healthcare (as discussed above). Although based on a specific GP population, this study’s findings should be of wider relevance because primary mental healthcare in most industrialised countries shares similar reform objectives (e.g. optimising GPs’ role, accessibility and continuity of care), and organisational and practice features.²² As increasing attention is devoted worldwide to the development of optimal integrated primary care, this article contributes to the discussion on mental healthcare service planning.

Method

Design and study population

This study employs a mixed-method approach with emphasis on qualitative investigation. It focuses on GPs in five Quebec regions (of 18), including nine local networks (of 95) in urban, semi-urban and rural settings; two of these networks include university-affiliated psychiatric facilities. From a previous survey sample of 398 GPs representative of Quebec’s total GP population of 7,199,^23,24 60 were selected for in-depth qualitative examination of GPs’ collaboration with mental healthcare providers. A sample of 12 GPs in each of the five regions was targeted. Subjects represent a variety of practice settings, including solo or group practices in private clinics, HSSCs, hospitals (acute, psychiatric or long-term), walk-in clinics, FMGs and NCs. Gender representation within the GP sample was also a factor. GPs were randomly listed (considering region, practice setting and gender), and were asked to participate in the research until completion of the sample. Email, postal letter, fax and telephone contacts were used to reach GPs.

In total, 124 GPs were approached to participate in the research; 29 were excluded because they had moved, had retired or were unreachable. Sixty agreed to take part in the research, and 35 refused; a response rate of 63%. Comparisons between respondent and non-respondent GPs were made with regards to sex (χ² = 0.50, degrees of freedom [df] = 1, P = 0.4777) and age (F = 0.10, P = 0.921) yielding non-significant results. The 60- and 398-GP samples were also compared on key parameters: age, gender and fee-for-service income (Table 1) – with no significant difference found. However, the 60-GP sample drew much less income from service fees than Quebec’s GP population as a whole. In addition, the validity of our qualitative approach draws more on the richness and depth of information that we obtained from the subjects and their detailed explanations of the mechanisms by which different factors influenced their behaviours than on statistical representation. Furthermore, our sample size allowed us to reach theoretical saturation. Recruitment occurred from April 2009 to March 2010. Participants signed a consent form approved by a university research ethics board.

Table 1.

Comparison between our 60-GP sample, the previous research based on the 398-GP sample and the GP population in Quebec

	60 GPs (%)	398 GPs (%)	P(χ2)	Quebec GPs* (%)	P (χ2)
Age categories (years)
< 35	1 (1.7)	29 (7.3)	0.350	13.7	0.350
35–44	9 (15.0)	112 (28.1)		27.5
45–54	24 (40.0)	170 (42.7)		35.0
55–64	22 (36.7)	74 (18.6)		18.3
65+	4 (6.7)	13 (3.3)		5.5
Gender distribution
Male	29 (48.3)	194 (48.7)	0.670	55.1	0.322
Female	31 (51.7)	204 (51.3)		44.9
Income level from fee-for-service	54.8 (37.8)	64.9 (39.8)	0.149	74.0	0.005

* 20.

Data collection

A 27-item questionnaire and interview guide were used. Both instruments were tested on three GPs not included in the final sample. The questionnaire, a shorter version of the survey used in the preceding research project,²³ covers four dimensions: (1) GPs’ sociodemographic profile and practice location, (2) continuing medical education, (3) clinical practice features and profile of patients with mental disorders, and (4) comfort level in managing patients with mental disorders. It included categorical or continuous items, with some 5- or 10-point Likert-scale questions. It was self-administered and required 10 minutes to complete. It was used to compare the sample of the first GPs surveyed (the 398 GPs representative of the full Quebec GP population) with the 60 GPs selected for the qualitative investigation. It also provided minimal descriptive information on GPs’ sociodemographic, workplace and patient profiles.

Development of the interview guide was based on a literature review of primary care, the preceding research project and a conceptual model (Figure 1), adopted from Mulvale and collaborators’ study.²⁵ The guide included three sections: (1) GP involvement in mental health, and relationships with mental healthcare providers; (2) contextual and enabling or hindering factors related to GP best-practice collaboration or models promoted; and (3) impact on all these preceding factors on GP collaboration with mental healthcare providers. Interviews (70 minutes) were conducted by one of the three first authors (25% face-to-face; 75% over the phone), recorded and transcribed (respondents’ anonymity was respected).

Figure 1.

Conceptual framework (guiding data treatment) – GPs’ collaboration with mental healthcare providers

Data analysis

For questionnaire data, distribution for categorical variables and mean values for continuous variables were computed with SPSS Statistics 17.0. As for qualitative interview data, transcripts were read by the first three authors and subsequently coded using NVivo 8. The conceptual model (Figure 1) guided data analysis. The researchers ensured coding accuracy and refined the interpretation of results. Data analysis also involved the reduction and synthesis of information, and the production of frequency of themes (i.e. proportion of GPs endorsing a theme). Reports were produced to summarise pertinent results, which were read and discussed by all researchers. In addition, a second-step qualitative analysis was performed using information about GPs’ main practice settings (where GPs spent most of their work hours): (1) solo private clinics (solo practitioners in a single location), (2) group private clinics, (3) HSSCs, and (4) hospitals (including general and psychiatric hospitals). This permitted comparison of GPs’ collaborative practice with regard to their main settings. We also compared GP collaboration strategy for patients diagnosed with common mental disorders (anxiety, depression) and patients with serious disorders (schizophrenia, bipolar disorder) with or without concomitant disorders (physical problems, substance abuse).

Results

GPs’ sociodemographic, workplace and patient profiles

Of the 60-GP sample, 48% were men and 52% women. Mean age was 52 years, with nearly two-thirds of the sample (65%) aged 45–59; accordingly, a majority (73%) possessed 20 years’ experience or more. Approximately half (47%) earned 67–100% of their income from fees for services. Roughly 75% worked in two to four settings. Twenty GPs worked mainly in group private practices, 11 in solo private clinics and hospitals and 18 in HSSCs. Among GPs in private group practices, 50% were also in FMGs and 25% in NCs. Among GPs in solo private clinics or in HSSCs, 27% also were in FMGs. Other pertinent data related to GP practice, extracted from the brief questionnaire, are presented in Table 2.

Table 2.

GP clinical profile and overall characteristics of their patients

Frequency (N = 60)
	All	Common mental disorders	Serious mental disorders
GP clinical profile
GP mean age [Mean (SD)]	52 (7.9)
Number of GP workplaces [Mean (SD)]	2.82 (1.2)
GP level of comfort in mental health on a scale from 0 to 10 (not comfortable to perfectly comfortable) [n (%)]
1–4		–	18 (30.0)
5–7		9 (15.0)	28 (46.7)
8–10		51 (85.0)	14 (23.3)
Consultation length (min) for mental disorder cases according to their main workplace [Mean (SD)]	31.3 (9.1)
Solo practice	24.4 (3.7)
HSSC	31.4 (8.4)
Group practice	36.3 (7.4)
Hospital	33.8 (12.2)
Number of half-day training in mental health during the last 12 months [n (%)]
0–1	14 (23.3)
	2	12 (20.0)
3–4	19 (31.7)
5–20	15 (25.0)
Patient characteristics
Prevalence of patients with mental disorders among GP all patients [n (%)]
0–33	25 (41.7)	42 (70.0)	56 (93.3)*
34–66	23 (38.3)	11 (18.3)	3 (5.0)
67–100	12 (20.0)	7 (11.7)	1 (1.7)
Proportion of mental disorder patients managed by GPs out of all patients [Mean % (SD)]	45.2 (28.0)	33.2 (21.8)
Without concomitant diagnosis		57.6 (23.4)	32.5 (26.6)
With concomitant diagnosis		40.8 (22.8)	59.3 (30.7)
Proportion of patients with mental disorders and concomitant diagnosis managed by GPs [Mean % (SD)]
Patients with mental disorders and substance abuse		18.6 (16.3)	20.1 (22.8)
Patients with mental disorders and intellectual deficiency		7.9 (14.8)	7.2 (9.9)
Patients with mental disorders and somatic problems		53.3 (27.9)	42.6 (32.2)
Patients with mental disorders and at least two other health issues		17.4 (16.32)	20.0 (23.6)

* In this 0–33% category, 85% of GPs managed 0% of these patients, 15% managed between 1 and 10%.

State of relationships between GPs and mental healthcare providers

Patients with mental disorders were generally managed by GPs in solo practice (GPs working on their own whatever the setting) who faced difficult access to mental health resources (see interview quotes in Box 1). Parallel or sequential follow-up and communication with psychiatrists through consultation reports were the norm. GPs identified psychosocial professionals, especially psychologists and psychiatrists, as their main partners in the management of mental disorders. Psychologists were reportedly quoted as playing a key role in helping patients facing challenging life events, thereby preventing accident and emergency department visits and hospitalisations. About half of GPs referred their patients with common mental disorders for psychotherapy either with psychologists in private practice (for patients with private insurance) or the public system (HSSCs). At least two-thirds of GPs considered joint psychotherapy and medication as the best treatment option for managing common mental disorders.

Box 1. Collaboration between GPs and mental healthcare professionals: examples of some qualitative statements.

About state of relationships between GPs and mental healthcare providers

‘In most cases of common mental disorders, I provide follow-up care jointly with a psychologist.’ GP 040 – Hospital*

‘I always insist on prescribing both types of treatment – for common mental disorders, depression. I strongly recommend psychotherapy and pharmacotherapy.’ GP 066 – HSSC**

‘Some patients are more difficult to diagnose, which is why I need the opinion of the psychiatrist in determining the proper treatment. Once patients see the psychiatrist, I receive recommendations that assist me in the treatment of patients whose needs are more complex.’ GP 090 – Solo private clinic

‘It’s not out of bad faith that we don’t work with psychiatrists. It’s simply that we don’t have access to them, so patient reference is done by the emergency department.’ GP 050 – Solo group clinic

‘I certainly do talk often to pharmacists. When I want to change a medication, I seek their opinion to find out if it’s feasible to do so, if one drug should be stopped completely before starting another or if cross-medication is indicated. If a patient’s prescription is about to run out, the pharmacist sends me a fax, so that I can renew it.’ GP 033 – HSSC

About factors that enable collaboration between GPs and mental healthcare providers

‘Since the beginning of my practice, I have always worked in HSSCs. What influenced me in my workplace is that you can give patients the time they need and you have a multidisciplinary team that is close at hand with which you can easily have exchanges.’ GP 021 – HSSC

‘I am not only in private practice. I work in a psychiatric hospital as well, and that may be why I take better care of my psychiatric patients, I treat their MD***, we have team discussions once a week. So obviously, it gives me a lot of experience. You try powerful drugs, something that is more difficult to do in private practice. First, you gain confidence in this kind of medication, and then you gain experience in using them.’ GP 064 – Solo private clinic

‘I practice physical medicine within a psychiatric hospital. I’m surrounded, however, by psychiatrists who make decisions and I’m always aware of how such decisions are taken. This encourages me to take on more patients in my private practice.’ GP 085 – Hospital

‘Informal collaboration is essentially easiest for me because I have a psychiatrist colleague who works with me at the hospital. I can discuss my patients’ cases informally with him from time to time. I find this more useful than going through the formal network.’ GP 081 – Hospital

‘We often use the evaluation liaison module when we have questions about a patient’s proper diagnosis or treatment. They do an assessment and they let us know if we are on the right track or not. So, we’re reassured in that way.’ GP 082 – Group private clinic

About factors that hinder collaboration between GPs and mental healthcare providers

‘I have good relationships with psychiatrists at hospital X because I work there. Access to and communication with psychiatrists is very easy for my patients at hospital X. However, given that psychiatric services are regionalised, as soon as patients fall under the care of another hospital, I lose my contacts.’ GP 028 – Solo group clinic

‘Referrals in psychiatry are difficult, if not impossible. Waiting times are excessively long; telephone consultations for a quick opinion are also not possible.’ GP 049 – Solo private clinic

‘For those without insurance, access to psychotherapy is very difficult. For ten patients in need, only one will be able to receive psychotherapy services in an HSSC or another place that won’t charge a fee.’ GP 070 – Solo private clinic

‘Of course, financial compensation is important. I think that it’s not appropriate for treating patients with MD who require more time, but you’re not being remunerated for that extra time. I believe that a special compensation package should be attached to the management of MD.’ GP 045 – Group private clinic

‘I have never seen any training given by a local psychiatrist, it’s not right. It’s during such training that we should learn to know each other.’ GP 044 – Solo private clinic

‘Single-access points have been implemented at HSSCs but they are experiencing recurring changes and high staff turnover, all of which are not conducive to continuity of care. They are not yet running smoothly, but HSSCs are working on that’ GP 032 – Solo group clinic

About GP best-practice collaboration models and impact on their practice

‘For me the ideal practice is the shared care model.’ GP 030 – HSSC

‘If I have the support of a psychiatrist (e.g. training, telephone consultations, quick office visits), well, I’ll be more interested in providing treatment to mental health patients for both common and serious mental health cases.’ GP 011 – Solo private clinic

‘Ideally, it would help patients to have several therapists in the same physical location.’ GP 041 – Solo group clinic

‘I think we are the access point for a patient. We can provide follow-up care for stable serious mental disorders and we can deal with common mental disorders. We must have rapid access to a psychiatrist who can either see or support the patient during an acute episode or provide follow-up. We must also have access to a multidisciplinary team that will take care of the patient and will not let him or her fall through the cracks. The patient should have access to various professionals on that team: nurse, social worker and psychologist. There should even be a place for the integration of new professionals in the community who could accompany the patient daily in this process of change. I think it would be beneficial.’ GP 065 – HSSC

‘I believe that close collaboration with psychologists would effectively prevent psychiatric consultations.’ GP 015 – Solo group clinic

‘I think that the great majority of GPs agree with the reform in mental health that aims to ensure better frontline management of patients with the support of multidisciplinary teams working with a psychiatrist.’ GP 02 – HSSC

Psychiatrists were required for the few cases of complex common mental disorders, beyond GPs’ expertise (∼ 20% of patients with mental disorders, especially: initial diagnosis of schizophrenia, refractory disorders and mental disorder with substance abuse comorbidity). In addition to the need to establish diagnoses or adjust medication for difficult cases (the reason in ∼ 70% of cases), GPs referred to psychiatrists mainly to: (1) manage crisis situations (suicidal ideation, psychotic episode); (2) access specialised interdisciplinary mental healthcare services (personality disorder clinics, day hospitals); and (3) complete insurance paperwork for patients’ sick leave and return to work. The few patients with serious mental disorders they managed (see Table 2) were followed-up in conjunction with mental health professionals and psychiatrists (e.g. nurses, social workers). For these patients, GPs assessed medication side effects, and in ∼ 10% of cases, treated concomitant physical ailments only. Generally, GPs believed that they did not have sufficient skills to treat these patients alone. Because psychiatrists were so difficult to contact, some GPs established informal links with other GPs who specialised in mental healthcare. The latter had usually practised in HSSCs or psychiatric facilities, showed a keen interest in mental healthcare and saw more patients with mental disorders.

Pharmacists were also mentioned as key partners, but GPs used them no differently than for other medical conditions, and communicated only occasionally with them, by fax or phone, to: (1) renew prescriptions, (2) monitor patients’ compliance to medication, (3) ensure accurate drug dosage, (4) suggest medication changes, and (5) monitor side effects or interaction with other medication. Pharmacists practising in psychiatric facilities were particularly appreciated by GPs for their extensive knowledge of mental healthcare medication. In addition, GPs had little knowledge of the voluntary sector or the detox centres, but believed HSSCs could act as a bridge.

Factors that enable or hinder collaboration between GPs and mental healthcare providers

GPs cited few collaboration-enabling factors; however, they reported one key issue, namely, working in multidisciplinary practice settings, such as in HSSCs. Psychosocial professionals were present in a few private group clinics, but generally GPs were as isolated in these settings as in solo private practice. FMGs and NCs benefited from nurses’ expertise, but the latter focused on physical ailments. Multidisciplinary practice settings were seen as favouring exchanges between professionals, prompt patient referral and joint patient assessment.

In addition, GPs who worked or had practised in HSSCs or hospitals usually enjoyed stronger informal contacts – yet another collaboration-enabling factor. Informal networks were considered more reliable than formal ones (HSSC single-access points, HSSC mental healthcare teams) and permitted: (1) fast and easy access to psychiatrists and psychologists, (2) direct (phone or face-to-face) contact with psychiatrists to discuss cases without having to go through an HSSC single-access point, and (3) rapid response for semi-urgent cases (those that could easily worsen in the absence of psychiatric intervention). Cautiously, GPs avoided overusing informal networks, reserving them for crisis or complex situations. Almost a quarter of GPs claimed they had no informal networks. These included younger physicians who had not had time to establish them, and doctors (mainly at HSSCs) who appreciated their formal networks and saw no reason to look elsewhere.

Half of GPs relied on informal networks: psychiatrists they knew personally through work or training programmes. Most also had a list of private psychologists that they recommended to their patients. GPs favoured psychologists trained in cognitive–behavioural therapy who provided short-term and result-oriented therapies. GPs cited patients’ private insurance as a facilitating factor (no need to refer through HSSC single-access points). Access to private psychologist follow-up was immediate when patients had insurance. Private insurance also helped to relieve the financial worries of daily life, thereby further favouring patient recovery. However, it limited the number of psychotherapy sessions. GPs had to transfer patients whose insurance coverage had run out to HSSC single-access points for further psychotherapy sessions. Such occurrences did not favour robust therapeutic alliances and care continuity. GPs found waiting times too long at HSSC single-access points. Some patients even recovered on their own before receiving HSSC services, with GPs providing more assiduous follow-up in the interim.

Training in mental healthcare and two initiatives within formal networks – ELMs and shared care (‘responding psychiatrists’) – were deemed by GPs as effective means of improving collaboration and mental disorder management. Training that featured case discussion was considered doubly beneficial as it fostered networking and knowledge acquisition. GPs were particularly drawn to training offered at local networks, delivered by psychiatrists and involving multidisciplinary teams. Wherever implemented, ELMs and responding psychiatrists were seen as having a positive impact on access to psychiatric care, resulting in prompt psychiatric assessment and recommended treatment, with joint follow-up by GPs and psychiatrists as needed.

On the other hand, GPs cited numerous factors that hindered their management of patients with mental disorders, including: (1) lack of mental healthcare resources; (2) long waiting times for semi-urgent cases for obtaining services, especially psychiatrists and HSSC mental healthcare teams; (3) little or no information on waiting times for access to services; (4) not enough time and too few incentives to cooperate with mental healthcare professionals; (5) methods of remuneration, especially fee-for-service; (6) lack of stability in healthcare structure and provision; (7) GPs’ work settings, particularly solo practices; (8) insufficient mental healthcare or inter-disciplinary training (basic or continuing education) among GPs; and (9) increasing demand for healthcare.

Also, HSSC single-access points – which co-ordinate care with mental healthcare teams and psychiatrists – were not considered to be sufficiently flexible. For instance, referral to specific psychiatrists was not permitted. GPs’ patient assessments were revised, and patients could be turned away by HSSC single-access points if deemed not to require their services. GPs also complained about: (1) insufficient stepped-care planning, if the treatment suggested by psychiatrists for GPs’ patients was ineffective; (2) lack of patient follow-up by psychiatrists; and (3) the impossibility of referring patients to psychiatrists for specific time slots (including evenings and weekends). For GPs’ urgent cases, including patients with psychosis or suicidal ideation, accident and emergency departments were the only resource for rapid access to psychiatric care. When access problems were resolved, GPs judged psychiatric care to be excellent and beneficial both for patient recovery and their own management of mental healthcare cases.

GP best-practice collaboration models and impact on their practices

Ideal care models identified by GPs included the following characteristics: (1) a caseload of patients they felt sufficiently confident to treat, including most common mental disorders and, in the case of some GPs (mostly from HSSC), a few stabilised patients with serious mental disorders; (2) patients with a wide range of ailments, including a few requiring the further services of GPs with specialised knowledge of mental healthcare; and (3) a team of psychosocial professionals and psychiatrists working with them, practising in the GPs’ own settings. Also favoured were adequate training, financial incentives, appropriate modes of compensation (salary or hourly fees) and wide access to care.

GPs insisted on the importance of sharing responsibility for patient care. Depending on patient profile, joint follow-up or shared care should involve psychiatrists, psychologists, nurses and/or social workers. Psychiatrists should be more readily available for difficult cases, crisis situations and destabilised patients. Telephone follow-up (endorsed by 95% of GPs on a weekly basis) and joint consultation with psychiatrists (48% of GPs on a monthly basis at GP settings) were also required (to enhance GPs’ knowledge and skill in treating mental disorders). Irrespective of insurance coverage, psychologists should be universally affordable. In addition, sufficient psychotherapy sessions should be provided to meet the needs of patients with common mental disorders. Short reports from, or phone briefings with psychologists on treatment objectives, approaches and therapeutic planning would help GPs treat their patients more efficiently (especially the more challenging cases). Most GPs (88%) also recommended that nurses specialising in mental healthcare be recruited in their clinics to: (1) prioritise patients; (2) collect useful patient data (life habits, social support) prior to the medical visit; (3) assess initial mental health state; (4) provide psycho-education, medication follow-up and patient support in conjunction with the patients’ family; and (5) help manage more complex cases. Social workers were identified as key partners for serious mental disorders and highly disorganised patients. They – rather than nurses – would handle screening and follow-up of patients requiring more intensive psychosocial intervention. In addition, the voluntary sector would play a greater role, with nurses and social workers acting as brokers for patient referral.

Most GPs (92%) supported reforms designed to extend the role of primary care professionals. Some, however, were sceptical about reforms and doubted whether they could manage increased responsibility and heavier caseloads. Reinforcing multidisciplinary approaches emerged as a solution for bolstering primary mental healthcare, including: (1) reducing the frequency of patient follow-up, thereby lightening GPs’ caseloads; (2) improving patient outcomes through prevention and early detection, and more comprehensive responses to patients’ needs (for both mental disorder and concomitant problems); (3) reducing emergency visits and hospitalisations through improved patient monitoring; and (4) cutting costs associated with the management of mental disorders. HSSC single-access points, mental healthcare teams and shared-care initiatives were seen as major steps forward for reforming primary mental healthcare, but requiring consolidation before they could yield improved patient outcomes. Few management tools were utilised, e.g. clinical protocol, screening diagnosis questionnaire, electronic medical record – but they were expected to be helpful.

Discussion

In the context of current mental healthcare reforms, this study is designed to cast light on GPs’ interaction with mental healthcare providers; factors that enable or hinder collaboration; and appropriate models for improving integrated primary mental healthcare. In continuity with other studies,^9,10 this research highlighted the pivotal role GPs played in managing mental disorders, mainly common mental disorders. GPs treated fewer cases and weren’t as comfortable with severe mental disorder or concomitant substance abuse disorder patients.^12,26 Our findings also showed that the great majority of GPs favoured joint pharmacotherapy and psychotherapy as the most effective treatment for common mental disorders, even though very few contacts were reported between them and psychologists, which points to ‘parallel care’ rather than real integrated care for patients. Studies indicate that most patients with common mental disorders are treated by GPs, without the latter having significant contacts with other mental healthcare providers.¹⁵ In our previous 398-GP survey,²⁷ more than 50% of GPs said that they had no contact with any of the other mental healthcare professionals (e.g. psychiatrist or psychologist). However, referrals are a more common strategy, as shown in our study, as in others. In Valenstein’s study, ∼ 43% of patients with significant depressive symptoms were referred to psychologists.²⁸ In the research by Wang and collaborators, 26% of patients were referred to a psychologist or psychiatrist following a ‘mental health visit’ and 25% were referred following a regular visit to their GP;²⁹ while 10% of patients in Maguire and colleagues’ study were referred to a psychiatrist.³⁰ Globally, overall referrals from GPs to mental healthcare providers varied considerably (being in the highest range in our study) depending on resource and diagnosis included, measure utilised and system studied (e.g. free or limited access to psychiatrist – 26% in Wang et al;²⁹ 58% in Piek et al;³¹ and 23% in Kendrick et al³²).

In light of the study results, the initial conceptual framework (Figure 1) was revised to improve our understanding of the interaction and impact of factors on ‘GPs’ collaboration with mental healthcare providers’. Figure 2 illustrates enabling factors that drive GPs’ collaboration with mental healthcare providers, and best-practice collaboration strategies and models, which are seen by GPs’ as having a positive impact on their practice (e.g. reducing the frequency of patient follow-up, lightening caseloads of patients with mental disorders, extending overall caseload to answer increasing demand for care). The enabling factors summarised all the key components found to be essential to a successful collaborative care model in mental health, extracted from the study and classified in terms of environment or macro-organisational features, practice settings, GPs’ individual characteristics and patient management profiles. Main GPs’ collaborators are also identified including their essential roles.

Figure 2.

Final conceptual framework (based on the study results) – GPs’ collaboration with mental healthcare providers

Highlighted in the literature, as in our study, are recurring issues hindering GPs’ collaboration with mental healthcare providers: long waiting times for accessing psychiatric care or psychotherapy at no charge; limited primary care team practices; and inadequate methods of remuneration and lack of financial incentives for GPs to manage mental disorders.^33,34 In addition, consultation time, serving essentially as a means for GPs to prescribe medication and provide follow-up, was too brief to permit optimal management of mental disorders. GPs in our study were faced with all these challenges, and their recommendation of joint psychotherapy and medication suggested that they recognised the limited effectiveness of a pharmacology-only approach. Some patients resist taking drugs or prefer psychotherapy instead.¹³ Moreover, compliance to medication is generally poor in patients with mental disorders,^14,35 hence the need to strengthen psycho-educational or patient-centred approaches. In countries such as the UK and Australia, efforts have been made to enhance access to free psychotherapy, and co-ordinate GPs and psychologists more effectively. Evidence-based psychological intervention is reported as cost-effective.^36–38

In addition, comprehensive service provision that meets patient needs, and organisational, longitudinal and relational continuity of care have repeatedly been mentioned as key enabling factors for patient recovery – even more so for patients with recurrent or complex mental disorders.^24,39 Non-physicians play a significant role in most successful interventions for chronic illness, in patient screening, medication and support follow-up, or case management as requested.^11,13 As with psychologists and psychiatrists, GPs also recommended that nurses play a key role,¹³ but with some reservations. The serious shortage of nurses in the Quebec healthcare system as in many other jurisdictions would undoubtedly hamper efforts to foster their participation in mental healthcare.^40,41 The literature also underscored that nurses usually reported lacking adequate knowledge and training in mental health, and were more willing to treat serious mental disorder than common mental disorder cases.^42–44

International studies^9,12 have also shown that poor outcomes are associated with a substantial proportion of individuals suffering from mental disorders whose cases are managed in primary care, lending credence to current efforts to promote shared care, multifaceted interventions, and best practices.⁴⁵ Models designed to improve interdisciplinary team practice and care co-ordination are viewed as more effective in the treatment of patients, especially patients with chronic or complex problems such as major depression or schizophrenia.¹⁶ They illustrate approaches and strategies designed to: (1) improve GPs’ ability to manage mental disorders, including patient self-management support; (2) reinforce support and co-ordination between GPs and specialised care professionals; (3) extend biopsychosocial services through greater collaboration with psychosocial teams and more services supplied in the community in support of GPs; (4) transfer patients to specialised care during crisis periods with subsequent follow-up by GPs and other psychosocial practitioners in the community; and (5) organise more efficiently the overall care network for improved healthcare provision and outcomes.

However, implementing strategies to enhance GPs’ management of mental disorders represents a sizable challenge, and models of care such as shared care and patient-centred medical home approach are still globally underdeveloped in most countries as in Quebec.^27,46 Obstacles to such collaboration include the fact that GPs are not usually compensated for time spent communicating with colleagues. For GPs in Quebec and Canada, as in other jurisdictions such as the USA, France and Germany, fee-for-service is the prevalent method of payment. GPs see themselves as autonomous entrepreneurs. Participation in multidisciplinary practices entails more diverse clinical responsibilities – a more fluid state of affairs than is usually encountered in GPs’ solo practices. Guidelines defining joint roles and responsibilities must be clearly established. Other key elements for effective shared care, as per the literature,^47,48 include: the development of electronic medical records, physician leadership, team vision, recognition of diversified expertise requirements, absence of hierarchical distinctions among professionals, adequate space and location; effective management and clinical skills, strong commitment to innovation and patient empowerment, and established clinical relationships (referred to in our study as ‘informal networks’). Investments to improve co-ordination and extend the role of mental healthcare partners would relieve the burden of care for GPs and allow them to enhance their ability to treat patients with mental disorders and expand their caseloads.

Study strengths and limitations

While a number of studies in the last decade have examined GPs’ collaboration with mental healthcare professionals, few have used mixed-method – especially qualitative investigation – and studied patterns of collaboration in regards to mental disorders generally, both common and serious mental disorders.^49–51 This study, however, has some limitations. First, this study may over-represent GPs who are more keenly interested in mental health, thereby accounting for a larger proportion of GPs paid by salary or hourly fees (rather than fee-for-service) in this sample, compared with GPs in Quebec as a whole. Second, GPs more driven by team practice may also be over-represented in this study. Third, the study focused on Quebec; studies must be conducted in other jurisdictions to establish a basis for comparison. The way primary care is organised depends on countries, but most follows the same challenges and trends for reform.^49,52 Finally, no data were collected on the adequacy of GPs’ treatment of patients with mental disorder, considered a major issue.¹²

Conclusion

This article illustrates the dynamics and key enabling factors and strategies to promote collaboration among GPs, psychiatrists and psychosocial professionals. It argues that increased efforts are needed to strengthen collaboration in the field of mental healthcare. The final conceptual framework (Figure 2) highlighted key issues that drive collaboration, and the study showed that most of them are lacking, resulting in GPs working mainly in solo practice. Although still in their infancy, reforms are underway to extend primary healthcare group practice and collaboration among GPs, psychiatrists and mental healthcare teams in most industrialised countries. GPs in our study were strongly in favour of improving team practices. They strongly advocated extending access to psychologists for the treatment of common mental disorders, and to social workers for follow-up of serious mental disorders. Nurses can also play a more pivotal role in patient screening and follow-up, but only if the obstacles of insufficient mental healthcare expertise and personnel shortage are overcome. Support from psychiatrists is needed for GPs’ more difficult cases and for the purposes of knowledge transfer and enhancement of primary mental healthcare expertise. A culture of collaboration has to be encouraged as comprehensive services and continuity of care are key recovery factors of patients with mental disorders. Shared care, an extended role for psychosocial professionals, and more efficient mental primary care organisation should lead to expanded caseloads for GPs and better access to services for patients.