Integration between general practice and mental health services in Italy: guidelines for consultation-liaison services implementation

Abstract Purpose This paper illustrates some guidelines for the implementation of Consultation-liaison services in contexts where GPs work alone. We present some activity data of our experience in the period 1999–2004 and a critical evaluation of what works and what does not work. Context In Italy single-sited spontaneous initiatives of co-operation and integration between general practice and psychiatry have been implemented in many regions. Recently, the Italian Health Care Government has begun to encourage integration between primary and secondary care for the management of mental health. The Bologna Consultation-liaison Service opened in 1999 in one area. The service was first located in the Community Mental Health Centre and subsequently in a medical non-psychiatric outpatient service. In 2002, the services were implemented in the overall city area, and the Bologna Consultation-liaison Service had its own office in the centre of the town. Data source Data have been collected by reviewing clinical charts. They include clinical (mental status examination, progress notes) and socio-demographic data, assessment scales that measure psychological distress and disability, reports for GPs, and consultation outcome. Conclusion A consultation-liaison service like the one proposed in this paper could contribute to an efficient and fully-integrated collaborative management of common psychiatric disorders, reducing the use of mental health services.


Background
Anxiety and depressive disorders, globally defined as common psychiatric disorders, frequently occur in the general population w1x and are associated with high degrees of subjective distress and disability w2-4x. A recent multicentre survey, including people from 14 countries in the Americas, Europe, the Middle East, Africa and Asia, found that anxiety disorders are the most common mental disorders in almost all countries with prevalence in the range 2.4-18.2% (Interquartile Range 5.8-8.8%). Depressive disorders are the next most common with prevalence in the range 0.8-9.6% (Interquartile Range, 3.6-6.8%) w1x. Considering the number of days out of role (days in the past 12 months in which patients were totally unable to carry out their normal daily activities), people suffering from the more serious disorders reported at least 30 days in the past year (IQR, 32.1-81.4 days). In 1990, unipolar major depression was the fourth cause of disability-adjusted life years (DALYs), an index represented by the sum of life years lost due to premature mortality and years lived with disability adjusted for severity w5x; a projection for 2020 indicates that depression will be the second cause of DALYs after ischemic heart disease w6x. A more in-depth analysis of disability with Medical Outcome Study Short Form-36, found that patients with depression present the widest pattern of impair-This article is published in a peer reviewed section of the International Journal of Integrated Care ment as they are limited in their family and social relationships, have low vitality, have a negative perception of emotional well-being, tend to complain of bodily pain, to have a negative perception of general health and to be limited in carrying out daily activities. Patients with anxiety disorder show a less extensive impairment, with limitations in the mental but not in the physical domain w7x.
As a result, these disorders represent an important public health problem, which involves not only the mental health services, but the general practice as well. In fact, many patients with anxiety and depression are seen and managed in general practice, even in those countries where mental health services are available and efficient w8x. The WHO Collaborative Study on Psychological Problems in General Health Care (PPGHC) reported a depressive disorder prevalence of 12.5% and anxiety disorder prevalence of 10.5% w8x. Similar prevalence values were found in Italy w9x.

Models of integration between primary care and psychiatric services
The crucial role of general practice in the management of these disorders is receiving more attention from health service organisations, and a close co-operation between General Practitioners (GPs) and psychiatrists is underway in many countries, especially in the United Kingdom and in North America w10-12x. In these countries, the GP has a key role in the management of mental illness; as a gatekeeper to secondary care, heyshe plays a central role in the process of care-seeking for people with psychiatric disorders.
In Anglo-Saxon countries, a wide variety of models of joint working between GPs and mental health professionals has evolved: the Community mental health team, the Attached mental health professional and Consultation-liaison w13x. These models are not mutually exclusive and many services combine more than one. The Community mental health team consists of a multidisciplinary staff and usually operates within the remit of sectorised psychiatric services. The major innovations that they have brought to the interface have been the provision of a single point of referral for multi-disciplinary care with pooling and discussion of referrals with GPs. The second model is the Attached mental health professional; in many practices a mental health professional other than a psychiatrist (e.g. primary care nurses, clinical psychologists, social workers and counsellors) works within the primary care staff and provides psychosocial interventions to patients with neurotic illness. In the third model, Consultation-liaison, the psychiatrist usually meets primary care staff and does not see the patient; this model places greater emphasis on developing close links among health care providers and on reducing referrals of milder disorders; there is evidence that shows an enhancement of GPs' skills in the detection and management of mental illness. In all these models of integrated care, mental health specialists (nurses, social workers, clinical psychologists, psychiatrists) visit patients within general practice centres where many physicians and other providers work together. Thus, it is possible to establish professional relationships, see patients jointly with GPs, organise regular face-to-face contacts and schedule meetings with primary care staff.

The Italian context
Anglo-Saxon models are difficult to replicate in our country. In Italy, general practice is guaranteed to each citizen, but a formal primary health care service does not exist. GPs work individually instead of in group practices with little contact with other colleagues and without any kind of functional network. Residents are free to choose the GP, but each GP can have no more than 1500 subjects on his or her list. GPs are not employees of the National Health Service, but work for it on the basis of a nationwide contract and are funded by the National Health District with a fixed allowance per patient irrespective of the number of consultations or other provision of care. Except in emergencies, access to hospital facilities is possible only by referral from GPs. Exceptions for which selfreferral is permitted are specialist psychiatric services, services for dental care, and services provided in gynaecological and paediatric departments.
Mental health care is provided by Community Mental Health Centres (CMHCs) with multi-professional teams (psychiatrists, psychologists, nurses, social workers and occupational therapists), which mainly take care of people with severe and enduring mental health problems. Nevertheless, the CMHCs are a primary level structure, freely accessible for the citizens without GP referral. GPs refer only one third of the patients who ask for a visit to a centre, while another one third is referred by other physicians or social services; the last third does not have any previous contact with health care providers w14x.
Collaborative programmes between general practice and mental health were developed at the end of the 1990s. Before that, the only type of collaboration available was the request for occasional consultation without a shared project, an approach which proved to be unsuccessful. This non-coordinate service organisation implies that patients with common psychiatric disorders can be visited by either GPs or psychiatrists, regardless of the severity and prognosis of their disease. In Italy, awareness of the importance of integration between general practice and psychiatry is growing and some innovative projects have been created. Specifically, epidemiological studies were carried out to evaluate the impact of anxiety and depressive disorders in general practice in our country w9,15,16x. Furthermore, single-sited spontaneous experiences of cooperation and integration between general practice and psychiatry have been implemented in the region of Emilia Romagna w17,18x and in particular in Bologna. These pilot experiences are the premise of the guidelines and of the project that is illustrated in subsequent paragraphs. In 1999, the Italian National Health Care System began to encourage integration between primary and secondary care for the management of mental health, subsequently recommending that the CMHCs should develop primary care Consultation-liaison services w19x. Consistently, some Regional health programmes, for instance those developed in the Emilia-Romagna Region, encouraged the integration between the two agencies and the implementation of consultation services dedicated to general practice w20x.

Guidelines for Consultationliaison service implementation
We illustrate some recommendations for the development of Consultation-liaison services suited to a health care system in which primary care is not supplied by a well-organized service, but by a number of GPs who work independently. In this context, a consultation activity within primary care is not feasible because there are too many doctors to establish close relationships. Moreover, the organisation of meetings with GPs and consultant psychiatrists is quite difficult. For these reasons, the consultation liaison activity is based on Community Mental Health Services and it is often located within CMHCs. Delivered interventions and resources should be clearly defined with the publication of a Service Card to best identify the consultation activity aims.

Aims of Consultation-liaison services
The aims of Consultation-liaison services are: 1) To develop an ongoing collaboration with GPs; 2) To improve the quality of treatment for patients with common psychiatric disorders; 3) To modify the pathways of care, supporting the management of common psychiatric disorders in general practice and focusing the activities of the mental health services towards severe or difficultto-treat cases.

Setting and staff
The Consultation-liaison service is a specialised component of the CMHC and it has to be located within the CMHC or, if possible, in non-psychiatric outpatient facilities. Direct telephone lines may be activated to facilitate contact with GPs.
Consultation-liaison services should have a multidisciplinary staff; usually a consultant psychiatrist, a psychologist, a resident psychiatrist and a nurse; every professional has to receive specific training in this type of joint work and spend, on average, one afternoon a week providing this service. Nurses have a key role, as they should organise the service and manage a crucial system of relationships with other professionals and patients.
Preferably, dedicated staff should manage the Consultation-liaison service: a few professionals chosen among CMHC personnel who dedicate part of their working time to these activities. The selected professionals should acquire expertise and specialised skills with regard to psychiatric disorders that occur in general practice; moreover, they should establish a collaborative and lasting relationship with GPs.

Activities
To assist GPs with the management of patients with anxiety and depression, Consultation-liaison services provide continual and multifaceted clinical support consisting of diagnostic assessment and brief focused therapeutic interventions, and additionally, Consultation-liaison activities.
Diagnostic assessment could be based on a diagnostic system specifically for general practice (e.g. WHO Diagnostic and Management Guidelines for Mental Disorders in Primary Care w21x) because diagnostic classifications used in a psychiatric context, such as the Diagnostic and Statistical Manual of Mental Disorders or International Classification of Disorders are difficult to apply in this setting. In general practice, in fact, the whole spectrum of anxiety and depression symptoms, from psychological distress to mental disorder, is observed, while in specialist practice, more defined and severe mental disorders are usually met.
The assessment should include the establishment of a psychiatric diagnosis, the identification of symptoms or problems as reported by the patient, significant life events, and the possible description of dysfunctional coping behaviours. The evaluation is followed by a treatment plan, which includes pharmacological intervention, counselling, and suggestion of further treatment, when needed. Information is then forwarded to the GP in a typed report designed to be thorough, but concise. Attention has to be given to the protocol's format, which needs to be user-friendly and relatively free of psychiatric jargon. Written communication could be accompanied by telephone communication or interpersonal contact in order to increase understanding and cooperation between psychiatrists and GPs. Occasionally, if the consultant psychiatrists see patients whose condition is deemed too serious to be treated within the above-mentioned system, they should refer these patients to the CMHC. According to the Consultation-liaison service protocol, the consent of the GP is required to allow the transfer.
The Consultation-liaison services could also offer brief and focused therapeutic interventions to support the treatment offered by the GP. These activities are defined as ''shared care'', to highlight the fact that the psychiatrist intervention represents help to the GP's therapeutic plan. When necessary, the psychiatrist could initiate pharmacological treatment and furthermore evaluate the treatment's efficacy and patient's compliance. In other cases, a psychological intervention can be provided by either a psychologist or a resident psychiatrist working under the supervision of the consulting psychiatrist. This treatment should be primarily oriented towards assisting patients with clarification and understanding of their disorder; many psychological interventions are suitable for these settings: problem-solving therapy, brief cognitive-behavioural therapy, brief psychodynamic psychotherapy, counselling, and group psychotherapy.
Finally, Consultation-liaison services can provide liaison activities. Meetings are organised, for the psychiatrist to meet regularly with a team of physicians for the discussion of cases. This can supplement interpersonal meetings, facilitate the discussion of patients post-assessment and advise GPs regarding the clinical management of the patient. This kind of activity also offers a teaching function, providing the GPs with operational skills, more useful than theoretical knowledge. These liaison activities can also be fruitful for the psychiatrists, since they give them the chance to be exposed to the GP's experiences and opinions.

Adjustment to local reality
Emphasis given to the above-mentioned activities (diagnostic consultation, brief interventions, liaison activities, scheduled meetings) and type of organisation could vary considering specific local needs and available resources. This policy derives from the belief that there is not a Consultation-liaison service model working in every context, since we have to consider many cultural, organisational and historical differences and also other factors, such as the cultural orientation of the working group, different individual expertise and training provided to the staff.

The Bologna Consultation-liaison Service
The city of Bologna has about 400,000 inhabitants and is sectored into 5 catchment areas, each serving approximately 80,000 people, with each area having an out-patient psychiatric facility and approximately 50 GPs. The Bologna Consultation-liaison Service opened in 1999 in one area. The service was first located in the CMHC and subsequently in a medical non-psychiatric outpatient service. In 2002, the Service was implemented in the overall city area and the Bologna Consultation-liaison Service had its own office in the centre of the town. Both in 1999 and 2002, the Service was presented to GPs in Bologna by an informative letter and subsequently through a workshop that promoted discussion among GPs and psychiatrists. In 2003, a conference on consultation liaison activities was organised. David Goldberg, one of the major experts in this field, gave a presentation and further Italian consultation-liaison experiences were presented.
Currently, the health care professional staff is composed of 5 consultant psychiatrists, one from every district of the city, 1 psychologist, 2 resident psychiatrists, and 5 nurses. The Consultant psychiatrist provides GPs with consultations (assessment, advises on management, in particular pharmacological treatment, evaluation on needs for referral), sees patients and supervises residents. A psychiatrist, together with a psychologist, runs therapeutic groups for patients with somatoform disorders and for elderly patients suffering from anxiety or depressive symptoms. Psychiatry residents provide control visits and brief psychotherapies. Nurses receive patients, complete sociodemographic forms and administer the evaluation scale (General Health Questionnaire, Work and Social Adjustment Scale), take appointments and revise the database. Moreover, nurses maintain contact with CMHCs to coordinate referral and to make appointments for visits in the mental health sector.
We present a short summary of our experience with some activity data and a critical evaluation of what   works and what does not work. Data have been collected by reviewing clinical charts. They include clinical (mental status examination, progress notes) and sociodemographical data, assessment scales that measure psychological distress (General Health Questionnaire) and disability (Work and Social Adjustment Scale), reports for GPs, and consultation outcome (back-referred to GPs, or triaged to mental health services, or dropped-out). Data are usually inserted in an electronic database. Table 1 shows that the number of patients referred to the Consultation-liaison Service and the number of GPs that referred patients increased in the period 1999-2004. Only a few of these patients were subsequently taken into care in CMHCs while more than 80% were referred back to GPs. Few patients, less than 5%, need a second consultation during the 6 months following the care episode. The majority of patients referred to the service were working middleaged females. Marital status did not appear to be relevant. The patients were suffering from depressive or anxiety disorders with moderate levels of disability ( Table 2). Rates of access to CMHCs or the Consultation-liaison Service of patients with diagnosis of common psychiatric disorders in one district of Bologna are shown in Figure 1. Since Consultation-liaison Service activity has been separated from CMHC activity and has been placed in a different location, the CMHC first visit number has decreased.

Strengths and limitations of the model
Taking into account the Anglo-Saxon model, we had to organise the Consultation-liaison services differently because of the different health service organization. Our model has two main limitations: (a) The risk of not discriminating between Consultation-liaison and routine mental-health activities. GPs had difficulties with the distinction and patients did not appreciate the location within the psychiatric facilities. To avoid these shortcomings, we moved the Consultation-liaison offices first within non-psychiatric medical facilities and subsequently into their own centre, thus operating separately from the CMHCs.
(b) Liaison activities were considered lacking and in need of re-evaluation. In big cities, periodical meetings with GPs and liaison activity are difficult to arrange. Thus, communication between GPs and Consultation-liaison services was mainly via written reports. Direct contact to discuss individual cases was limited to those GPs particularly interested in psychiatry. Initiatives to improve communication with GPs have been planned. Because the telephone consultation service was underutilised, we now have a website under construction which will enable GPs to communicate with the Consultation-liaison service via e-mail.
Despite these limitations, our Consultation-liaison Service has some strength. The primary goal of the pilot centre was to establish a collaborative relationship with GPs. During this time, we were in close contact with a number of the GPs districts and we were able to establish an ongoing collaboration with half of them.
Our results indicate that the service was effective and well-received, since from 1999 to 2004 a growing proportion of GPs referred patients. The majority of GPs expressed appreciation of the service and rated the outcome of most patients as improved. Overall, GPs' satisfaction and clinical evaluation are in accordance with objective data presented in a paper which evaluated first year activities, showing a reduction of primary health care service utilisation following Consultation-liaison intervention w22x. Our results are consistent with more structured research and randomised clinical trials showing that support interventions similar to ours do improve the outcome of patients with the most common psychiatric disorders w23x.

Modification of pathways to care
The objective of the Consultation-liaison Service is to modify the pathways of care, supporting the management of common psychiatric disorders in general practice. We evaluated the impact of the Consultationliaison Service on the CMHC ''workload'' from this prospective. Data on access of patients with anxiety and depressive disorders demonstrate that the CMHC burden is slightly reduced. On the other hand, only a small proportion (10-15%) of Consultation-liaison Service patients is referred to the CMHC, since the large majority of these patients can be properly managed in general practice with a very limited specialist intervention. These data demonstrated that the implementation of a Consultation activity in primary care increases the number of patients who receive psychiatric treatment, as also reported by Jackson and colleagues w11x.

Conclusion
The management of common psychiatric disorders in general practice has received growing attention from the physicians and the National Health Care System, recognising that these disorders are frequent, very disabling and cause significant social and economic costs. In literature, many models of integration between primary care and psychiatric services have been presented, but the majority of them refer to contexts where primary care operates as a gatekeeper and allows a consultation activity provided by mental health professionals who work in close collaboration with GPs. This model of integrated care manages patients with anxiety and depressive disorders effectively. On the other hand, there is little data on consultation services for the management of common psychiatric disorders in countries where primary care is less organised and not integrated with secondary care. For example in Italy, specific institutional projects are lacking and only recently have health policies been drawn up to implement the collaboration between GPs and psychiatrists. We have tried to establish a service not within the primary care but as a go-between for the two agencies; a service that could provide adequate mental health care in a less stigmatising setting. We believe that a Consultationliaison service like the one proposed in this paper could contribute to an efficient and fully-integrated collaborative management of common psychiatric disorders, reducing the use of mental health services. We hope that our experience will contribute to the development of other Consultation-liaison services in countries with a Health System similar to ours, where GPs work alone. Future research could be focused on the economic evaluation of collaborative models at the interface between mental health services and primary care. Moreover, there are little data about the opinions and needs of general practitioners, consultant psychiatrists, and service users regarding collaborative models and strategy development for the integration of mental health care and primary care.