Shorter Oxford Textbook of Psychiatry, 6th Ed.

CHAPTER 21. Psychiatric services

Introduction

The history of psychiatric services

The components of a mental health service

Services for psychiatric disorder in primary care

Specialist services for acute psychiatric disorder

Psychiatric services that provide long-term care

Services for people with particular needs

Some difficulties with community care

International service principles

Introduction

The last two chapters dealt with the treatment of individual patients. This chapter is concerned with the provision of psychiatric care for populations. It deals mainly with the needs of and provisions for people aged 18–65 years (‘adults of working age’).

Services for children are described on p. 640, services for the elderly on p. 490, and services for patients with learning difficulties on p. 701. The organization of psychiatric services in any country inevitably depends on the organization of general medical services in that country. This chapter will refer specifically to services in the UK, but the principles embodied in these services apply widely.

The chapter begins with an account of the historical development of psychiatric services. This is followed by descriptions of the commonly available psychiatric services and of the problems encountered with these provisions. The chapter ends with a consideration of some innovations designed to overcome these problems.

The history of psychiatric services

Until the middle of the eighteenth century, there were hardly any special provisions for the mentally ill in Europe, other than in Spain, where some hospitals were present from the Middle Ages (Chamberlain, 1966), reflecting its Islamic influence. Both Christianity and Islam regarded mental illness as the result of supernatural intervention. Within Judaeo-Christian teaching it indicated demonic possession and/or the effects of sin, but within Islamic teaching the intervention was not necessarily malign or the result of wrongdoing. Islamic medicine was more concerned with mental disorder, and the Arab physician Rhazes cared for mental patients in Baghdad in the tenth century.

In Britain until the middle of the eighteenth century, most mentally ill people lived in the community, often with help from Poor Law provisions, or were in prison. In England, the Vagrancy Act of 1744 made the first legal distinction between paupers and lunatics, and made provision for the treatment of the latter. Private ‘madhouses’ (later to be called private asylums) were developed mainly for those who could pay, but accepted some paupers supported by their parishes (Parry-Jones, 1972). At about the same time, a few hospitals or wards were established through private benefaction and public subscription. The Bethel Hospital in Norwich was founded in 1713. In London, the lunatic ward at Guy’s Hospital was established in 1728. In 1751, St Luke’s Hospital was founded as an alternative to the overcrowded Bethlem Hospital which had existed as an institution since 1247, but only latterly became exclusively a ‘madhouse.’

Moral management

At the end of the eighteenth century the Enlightenment led to a more empirical, less theological understanding of madness. This rational approach led to increased public concern in many countries about the poor standards of private and public institutions, including efforts to improve the care of the mentally ill. In Paris in 1793, Philippe Pinel iconically released patients from their chains and introduced other changes to make the care of patients more humane. In England, similar reforming ideas were proposed by William Tuke, a Quaker philanthropist who founded the Retreat in York in 1792. The Retreat provided pleasant surroundings and adequate facilities for occupation and recreation. Treatment was based on ‘moral’ (i.e. social) management consisting of kindness, calm, and activity. Tuke and Pinel both acknowledged that, although mad, patients still could respond like other humans. This contrasted with the previous authoritarian approach aimed at simply controlling patients when medical treatments (usually bleeding and purging) were peripheral to the routine care of the insane. William Tuke’s grandson, Samuel, described their system in A Description of the Retreat (1813), and the Retreat was replicated internationally. It became clear that many mentally ill patients could exert self-control and did not require physical restraint and punishment.

The asylum movement

In the early nineteenth century most mentally ill people received no care and lived as vagrants or as inmates of workhouses and gaols. Public concern about the welfare of the insane in workhouses and gaols and some private madhouses led to the County Asylum Act of 1808, which provided for the building of mental hospitals in each of the English counties. However, it was the Lunatics Act 1845 that required the building of one in every county. At first the new asylums were small and provided good treatment in spacious surroundings. Moral management was championed. The ‘non-restraint movement’, which had started with the work of Gardiner Hill at the Lincoln Asylum in 1837 and was established by John Conolly at the Middlesex County Asylum, Hanwell, in 1845, removed all physical restraints (a practice that has characterized mental healthcare in the UK ever since).

Unfortunately, these early attitudes yielded to a more restrictive approach. More and more patients were transferred from the community and prisons to the new asylums. Initial optimism about the curability of psychiatric disorder faded with the accumulation of chronic cases and an increased focus on organic and hereditary causes. Asylums grew rapidly in size. The original York Retreat had 30 patients, and the first 16 British asylums had about 100 patients each. They averaged 300 patients by 1840, 540 patients by 1870, and 960 by 1900. Some patients with chronic illness were housed in detached annexes or houses in the grounds of the asylum. Other hospitals returned patients to the community either by boarding them out with a family (a form of care that was practised most successfully at Gheel in Belgium), or by returning them to workhouses. The Lunacy Commissioners, whose role was to oversee the care of the mentally ill, were concerned that these arrangements could lead to abuse, and were opposed to them. Nevertheless, nineteenth-century asylums, even when overcrowded, provided a standard of care for the mentally ill that was lacking elsewhere. Thus the mentally ill were protected from exploitation, and were provided with shelter, food, and general healthcare, all without cost.

With the increasing size, overcrowding, and staff shortages, there was less emphasis on moral treatment. A change to a stricter custodial regime was endorsed by the Lunacy Act 1890, which imposed restrictions on admission and discharge from hospital. The bureaucratic and rigid 1890 Lunacy Act had resulted from years of unwarranted committal, but it handicapped British psychiatric development for nearly half a century. For accounts of psychiatric hospitals in the UK and the USA in the nineteenth century, see Jones (1992) and Rothman (1971).

Arrangements for early treatment

The opening of the Maudsley Hospital in 1923 provided an outpatient service and voluntary inpatient treatment in surroundings in which teaching and research were carried out. In the years between the wars the impetus for change increased. The Mental Treatment Act 1930 repealed many of the restrictions on discharge of patients that had been imposed by the Lunacy Act 1890. It allowed county asylums to accept voluntary patients, and changed the term ‘asylum’ to ‘mental hospital’ and the term ‘lunatic’ to voluntary ‘mental patient.’ The 1930 Act also encouraged local authorities to set up outpatient clinics and to establish facilities for aftercare. Therapeutic optimism, already present after von Jauregg’s successful malaria treatment for cerebral syphilis, increased further with the introduction of insulin coma treatment (later abandoned) and electroconvulsive therapy. Efforts were made to improve conditions in hospitals, to unlock previously locked wards, and to encourage occupational activities. Similar changes took place in other countries.

In most countries, these reforms were halted by the Second World War. Psychiatric hospitals became understaffed as doctors and nurses were recruited to the war effort, and became overcrowded as some had to be evacuated for the care of the war injured (Crammer, 1990). In Germany, many psychiatric patients died in a eugenics programme—a shameful period in the history of a country that had been at the forefront of psychiatric progress in the nineteenth and early twentieth centuries (Burleigh, 2000).

Social psychiatry and the beginning of community care

After the Second World War, several influences led to further changes in psychiatric hospitals. Social attitudes had become more sympathetic towards disadvantaged people. Among psychiatrists, wartime experience of treating ‘battle neuroses’ had encouraged interest in the early treatment of mental disorder, and in the use of group treatment and social rehabilitation. In the UK, the National Health Service led to a general reorganization of medical services, including psychiatry. The introduction of chlorpromazine in 1952 made it easier to manage the disturbed behaviour of patients with psychosis. It became possible to open wards that had been locked (although some hospitals had done so before such drugs became available), to engage patients in social activities, and to discharge more of them into the community.

Despite these changes, services continued to be concentrated at single sites, often remote from centres of population. In the USA, Goffman (1961) vividly exposed the detrimental effects of such ‘total institutions.’ He described how their impersonal, inflexible, and authoritarian regimes eroded identity and through this ‘institutionalization’ generated chronicity. In the UK, Wing and Brown (1970) demonstrated how large mental hospitals characterized by ‘social poverty’ led to ‘clinical poverty.’ Vigorous social rehabilitation was introduced to improve conditions in hospital and to reduce the effects of years of institutional living. Occupational and industrial therapies were used to prepare chronically disabled patients for the move from hospital to sheltered accommodation or to ordinary housing (Bennett, 1983). Many long-stay patients were responsive to these vigorous new methods. There was optimism that newly admitted patients could also be helped in these ways.

Away from the hospitals, day units were established to provide continuing treatment and rehabilitation, and hostels were opened to provide sheltered accommodation. As a result of all these changes, the numbers of patients in psychiatric hospitals fell substantially in the UK and in other countries. The changes started later in the USA, with the introduction of Medicaid in 1965, but then progressed very rapidly. Despite these changes, services were still based in large mental hospitals that were often far from patients’ homes. The provision of community facilities for newly discharged patients often lagged behind need.

Hospital closure

After initial successes in discharging many institutionalized patients, it was anticipated that asylums could be closed and replaced by small psychiatric units in general hospitals, with support from community facilities. In most countries, the programme of hospital closure took place gradually. A notable exception was Italy, which at first lagged behind most other countries but later made rapid changes. In 1978, the Italian Parliament passed Law 180, which prohibited admissions to mental hospitals forthwith. It aimed to abolish the mental hospitals altogether over a period of 3 years and replace them with a comprehensive system of community care. Psychiatric ‘diagnostic’ units limited to admissions of 7 days were to be set up in general hospitals and community services to be developed for each catchment area. The scheme was based on the work of Franco Basaglia in hospitals in north-east Italy, and on the proposals of the professional and political movement that he founded. This movement—Psichiatria Democratica—combined a left-wing political view that patients in psychiatric hospitals were the victims of oppression by the capitalist system with the conviction that severe mental illness was induced more by social conditions than by biological causes. Basaglia’s charismatic personality, qualities of leadership (and a wife who was an Italian Senator) helped him to succeed with his reforms. Others found them more difficult, and their impact varied. Where the reforms were financed adequately and were implemented by enthusiastic staff (particularly in the north east), the new provisions were successful. However, in Rome and the south the new facilities were often inadequate, and problems were encountered (Fioritti et al., 1997). The state mental hospital system came to an end in 1989 (Burti, 2001).

In the UK and elsewhere the pace of change was slower, but similar problems arose. Many patients needed ongoing intensive support, and many required repeated readmissions to hospital—so-called ‘revolving-door patients.’ Rehabilitation services had to adjust their expectations and provide continuing care. Some discharged patients attended day hospitals for years without showing further improvement (Gath et al., 1973). The early expectations of de-institutionalization had clearly been over-optimistic. The policy of ‘community care’ was introduced to develop more adequate community provision.

The rise of community care

As hospitals closed, community psychiatric services acquired three responsibilities. The first was to provide treatment for those individuals with severe mental illness who would previously have remained in hospital for many years. The second was to assist primary care services in the detection, prevention, and early treatment of the less severe psychiatric disorders. The third was to treat severe acute psychiatric disorder as far as possible without lengthy admission to hospital, and as near as possible to the patient’s home. Services were to be comprehensive, to deliver continuity of care, and to be provided by multidisciplinary teams.

The framework for this ‘sectorized’ service was laid by the 1959 Mental Health Act in the UK. This Act required hospitals to provide outpatient follow-up for their own discharged patients, and also required social services involvement in both compulsory admissions and community support. A highly localized ‘sectorized’ service was therefore a practical necessity for effective joint working.

These general principles were applied rather differently in the UK and in the USA. In the UK, emphasis was placed initially on provision for patients discharged from long-term hospital care. In the USA, more emphasis was given to the prevention and early treatment of mental disorder as a way of avoiding admission. President Kennedy’s Community Mental Health Centers Act (1963) established community mental health centres (CMHCs) staffed from several disciplines. These centres offered psychological and social care, emphasizing crisis intervention and the treatment of acute, often relatively minor, psychiatric disorders. They had no outreach facility, so patients with chronic and severe mental illness (who seldom actively seek care) became neglected. CMHCs evolved a rather ‘anti-medical’ approach, which led to difficulties in recruiting and retaining psychiatrists (Talbott et al., 1987). This further limited their ability to care for severely ill psychotic patients. Dissatisfaction with the centres grew as people discharged from long-term hospital care found their way into private hospitals or prisons, or joined the homeless population of large cities (Goldman and Morrisey, 1985).

In the UK and elsewhere, some commonly agreed principles concerning community services developed from these early experiences:

• Minimizing inpatient care. Hospital admissions were to be brief, and as far as possible patients were to be admitted to psychiatric units in general hospitals rather than to psychiatric hospitals. Whenever practicable, people were to be treated as outpatients or day patients.

• Providing rehabilitation early on. The aim of this was to protect and possibly improve residual functioning and prevent further deterioration.

• Multidisciplinary teams. Care was to be provided by teams, usually consisting of psychiatrists, community nurses, clinical psychologists, and social workers, often working in collaboration with members of voluntary groups.

• Legal reform. New laws were introduced (e.g. the 1983 Mental Health Act and its 2007 amendment in England and Wales) to limit the uses of compulsory treatment, to encourage alternatives to inpatient care, and to strengthen the rights of the individual.

As experience increased after the reforms had begun, the following additional features were introduced:

• Care packages based on an assessment of each patient’s needs.

• Case management by a named, clinically trained care worker who led and coordinated the work of others involved in care.

• User involvement. Service users were increasingly involved in planning both their own treatment and the services for the population.

• Outreach to take services to vulnerable people who might otherwise find it difficult to engage with the care that is offered, and to arrange follow-up.

• Risk assessment carried out regularly, and more formally as part of the care plan.

These developments have contributed to the current pattern of services.

The components of a mental health service

From the 1960s to the mid-1980s, mental health services in the UK and indeed most parts of the world evolved slowly, based on professional consensus. In the 1980s, interest in evidence-based medicine spread to mental health services research. This approach was international and increased the pace of change dramatically. The fundamental components of most care, namely inpatient wards, day hospitals, and community mental health teams (CMHTs), had not been subject to research, but had established themselves without it and have proved durable. The more recent highly specialized teams (assertive outreach teams, crisis resolution/home treatment teams, and early-onset (psychosis) services) have been more policy and research driven. It is probably premature to comment on their future role. What follows is a description of the UK services, which addresses research and international evidence where this is relevant or helpful. The NHS is entering a phase of potentially profound change, and the predictability and uniformity of services are likely to change markedly during the lifetime of this book.

Inpatient wards

No comprehensive service can survive without access to 24-hour nursing supervision for acute episodes of severe illness. Surprisingly, inpatient wards are often overlooked in descriptions of comprehensive services. They serve patients at risk from neglect or suicide and those lacking insight. Wards usually accommodate 10–20 patients. It is rarely possible to effectively staff and run stand-alone units of less than three to four such wards (30–60 beds). The famed Italian Law 180 restricted wards to 15 patients, and this is a common international goal (Burns, 1998). Ward size is a trade-off between privacy and domesticity on the one hand and effective supervision on the other. Single rooms are preferable, as they afford maximum privacy and, although initially expensive, they improve flexibility and have been shown to significantly reduce conflict.

Single-sex accommodation

Concern for the safety and privacy of female patients has led to government requirements for single-sex wards. This reversal of the development that occurred in the 1960s and 1970s is non-negotiable, even if it may have been overtaken by the increasing availability of single ensuite accommodation in new units.

Smaller, more flexible units such as ‘crisis houses’ offering 24-hour care are a useful complement to inpatient wards. However, they are not a replacement for them, and where this has been attempted it has usually resulted in an unacceptably high rate of transfer to the psychiatric intensive care units (PICUs). Ward design and management are increasingly crucial as improved community care concentrates involuntary and disturbed inpatients in them.

Continuity of responsibility

Responsibility by the same clinical team for both community and inpatient care is unusual except in the UK and Italy, where it has delivered strikingly shorter stays and improved bed management. Continuity is difficult in dispersed populations, and increasingly even in cities as each team has fewer inpatients. A separate inpatient team ensures better inpatient standards, but with the risk of diverging approaches (e.g. a ‘medical model’ inpatient team and a ‘psychotherapeutic’ community team). It has recently become common in the UK, for reasons that are unclear but which may include the problems of wards having to cope with several admitting teams. Both the move to single-sex accommodation and specialized teams result in each team often having only one or two patients on a ward. Diffused responsibility and organizing multiple ward rounds have promoted the change, but not without cost (Burns, 2010). It appears to be in conflict with the expectation of continuity of clinical responsibility in the 1983 Mental Health Act, and particularly in the 2007 amendment permitting compulsory treatment in the community through the new provision of Supervised Community Treatment Orders (SCTOs, which are invariably referred to as Community Treatment Orders, or CTOs; see Chapter 4). The question arises as to whether the inpatient consultant or the community consultant should make the decision.

Longer inpatient care

Acute inpatient wards admit patients for weeks or a couple of months, with rapid discharge anticipated. Some patients require longer, more secure care service because of illness severity or for legal reasons. Modern rehabilitation wards are thus generally restricted to patients with persistently unacceptable behaviour.

Diagnosis-specific wards

Diagnosis- or disorder-specific wards are relatively rare in the UK. In Scandinavia and Central Europe, separate wards for alcohol and substance abuse are long established. Wards for specific, specialized problems such as anorexia nervosa or resistant schizophrenia offer highly specific regimes. These are generally regional, and in addition to acute admission wards, not an alternative. The current focus on organizing services along ‘care pathways’ may result in more specialized inpatient services, but it is too early to know whether this will happen.

Day care

Day care is provided in day hospitals and day centres, and there is not a clear distinction between them. Day hospitals are generally provided by health services, include medical and nursing staff, and can offer treatments (e.g. the prescription and monitoring of medication, psychotherapies). Day centres are provided by social services or voluntary organizations. They rarely provide specific treatments or employ clinically trained staff. However, services vary according to local context. A drop-in day centre may provide psychiatric assessment and treatment in areas of high social mobility and homelessness. Broadly speaking, day centres provide long-term social support and day hospitals provide focused interventions and treatments (Catty et al., 2005).

Acute day hospitals in Europe and partial hospitalization services in the USA have been energetically proposed as acceptable and economical alternatives to inpatient care (Marshall, 2003), but have had little impact. Day hospitals never achieved their anticipated prominence in the UK, having perhaps been overtaken as CMHTs have become more comprehensive. They serve specific groups well (e.g. mothers with small children, eating disorders, or personality problems). Day care is difficult to organize in rural settings.

Supported accommodation and residential care

The extent of inpatient care that is needed depends on local access to other supervised accommodation. Patients remain well outside hospital only with adequate support and stable, affordable accommodation. Supervision may be needed to ensure self-care, continued medication, and to anticipate and defuse crises. It can be provided by voluntary agencies, social services, or health services. Voluntary agencies tend to be more efficient at providing long-term residential care (Knapp et al., 1999), but reluctant to accept risky patients. A mixed economy works best.

Outpatient clinics

Psychiatrists assess patients, advise them and their referrers, and provide treatments. In state-funded systems ‘office-based’ practice is rare, as most professionals work in outpatient clinics or mental health teams. Psychiatrists and psychologists may still operate independently within outpatient clinics, but with access to enhanced resources and second opinions. Outpatient clinics may operate either alongside CMHTs, or as part of them. They work better for severe illness when fully integrated with CMHTs (Wright et al., 2004). Clinics provide an efficient format for assessments and monitoring of treatment progress.

Multidisciplinary community mental health teams (CMHTs)

Most community mental health services consist of varied forms of multidisciplinary CMHT. Nurses have long worked outside hospitals (e.g. midwives, health visitors), and have made psychiatric home visits since the 1950s. Social workers, psychologists, and occupational therapists are increasingly found in CMHTs. The staffing of these teams varies internationally, but they all hold regular meetings to assess and review the management of patients. These reviews incorporate their varied professional perspectives and allocate tasks according to staff skills and patient needs. They were developed in France and the UK in the delivery of sectorized psychiatry, refined in Italy, and further elaborated in North America and Australia.

The generic-sector CMHT (‘the CMHT’)

Who it is for

The CMHT is the fundamental building block of modern community mental health services. It originated as mental hospital catchment areas (which often covered a whole city or county) were divided into sectors of 50 000–100 000 inhabitants to permit ongoing care. It provides assessment and care for patients who have been discharged from psychiatric units and for outpatients who require more support than can be provided by primary care. Current sector size in Western Europe ranges from 20 000–50 000 members of the population, determined both by resources (the size shrinking as investment increases) and by the local configuration. As more specialized teams are established, the CMHT’s remit may be narrowed and its sector size increased.

CMHTs prioritize individuals with severe mental illnesses (SMI), such as psychoses and severe affective disorders. However, diagnosis is not all—complications due to social adversity, personality difficulties, or substance abuse can make secondary mental healthcare necessary even for apparently ‘minor’ disorders. Several tools have been developed for clarifying this threshold (Slade et al., 2000), but are of limited use, and most teams rely on clinical assessments. In countries with limited private care, such as the UK, CMHTs also treat mild and transient disorders.

Staffing and management

CMHTs range from 5 or 6 to over 15 full-time staff. They emphasize skill sharing and a degree of generic working, and have evolved an informal, democratic style (Burns, 2004), with senior psychiatrists initially providing clinical leadership. With increased staff numbers and treatment complexity, ‘team managers’ now coordinate the workload. There is little consensus on their role, which can range from the purely administrative to determining clinical priorities and supervising staff. Clinically active team managers usually have reduced caseloads. If clinical leadership and team management are separated (which is common if there is a strong medical presence), the roles need to be well defined and relationships clarified.

Assessments

The key to good care is accurate assessment, and CMHTs vary in how they provide this. In general, psychiatrists are responsible for an initial assessment (often in an out-patient clinic) and involve the team members in treatment. Increasingly, other team members have taken a role in assessments, either individually or jointly with the psychiatrist. This issue generates strong opinions, but there has been surprisingly little research into it. One study suggests that joint assessments with medical involvement may be highly efficient (Burns et al., 1993a). With highly developed primary care, non-medical assessments may be effective, but otherwise medical time should prioritize assessments. For the most severely ill patients, home-based assessments pay considerable dividends (Stein and Test, 1980; Burns et al., 1993a).

Case management

Most CMHT staff act as clinical case managers (Intagliata, 1982; Holloway et al., 1995), taking lead responsibility for the delivery, coordination, and review of care for an agreed number of patients. Early forms of case management involved individuals with limited mental health training, working in isolation, and the results were poor (Marshall et al., 2001). Currently clinical case managers provide direct care to build a trusting relationship and provide treatment and continuity for patients with complex needs. They also utilize the full range of team resources. The caseloads of staff members are explicitly limited (usually 15 to 20), and reviews are recorded and systematic. In the UK this has been formalized as the Care Programme Approach (Department of Health, 1990), which requires a named case manager (called a care-coordinator) and a document called a care plan, which records the patient’s needs or problems, the proposed interventions for them, and who is responsible for each of them, plus an agreed date for review.

Care plans are often supplemented with a risk assessment and a contingency (crisis) plan. This should not need more than two pages, but vigilance is required to prevent it from becoming weighed down with administrative details that obscure its clinical purpose. Such simple structured paperwork can be adapted to any service. It fulfils a vital coordinating role in complex care, and serves as a natural focus for clinical reviews. The value of this document (as with the risk assessment and contingency plan) lies in its brevity. Too much information is as risky as too little, and the level of detail needs to be clinically (not managerially) determined.

Team meetings

CMHTs always have one, and often two, regular meetings per week for both clinical and administrative business. They last between 1 and 2 hours, and the level of structure varies.

Allocation of referrals. Deciding who will assess new patients need occupy only 10 minutes at the beginning of the routine meeting, or can simply be delegated to a senior member to decide. Referrals can be allocated according to who is first available, or by matching the clinical problem against available skill and training. It is remarkable how many CMHTs fail to have an agreed timetable for new assessments (as is routine in outpatient clinics). Much time is saved if each member has agreed regular assessment ‘slots.’ There is then no need to ‘coordinate diaries’ for joint assessments, and patients can be directly booked into available slots if the member is absent at allocation. Time spent discussing allocations (and particularly having a ‘referral meeting’) before assessment is unprofitable, and most well-functioning teams delegate the task.

Patient reviews. Reviews are needed for new patients, routine monitoring, crises, and discharge. They can range from simply reporting the problem and proposed treatment in uncomplicated cases, through to detailed, structured, multidisciplinary case conferences. These may include other services (e.g. general practitioner, housing, child protection). New patient reviews are particularly important for providing a broad, experienced overview, particularly if assessments are distributed across the team. They also ensure rational and fair allocation to caseloads. Routine monitoring is often overlooked, yet is probably the most important review for team efficiency. It shapes and redirects treatment and identifies patients who are ready for discharge. The burden on individual staff members is regularly monitored. Routine monitoring is a statutory requirement of the Care Programme Approach, and is good practice in all case management. Crisis reviews are unscheduled but allow case managers to seek advice when they are unsure. Discharge reviews provide an excellent opportunity for audit and learning within the team.

Managing waiting lists and caseloads. Effective CMHTs need to guarantee prompt access. Routine assessments should take place within 2–4 weeks. Striving for a shorter waiting time is rarely productive, and waiting times of much over 4 weeks have a rapidly rising rate of failed appointments (Burns et al., 1993a). Urgent assessments (which include most psychotic episodes) need to be seen within a week, and usually within a couple of days. Emergency assessments are for those associated with immediate risk (e.g. hostile behaviour or suicidal intent) and ideally need to be seen the same day.

A practical approach to managing waiting times is to count the number of referrals in the preceding year and allocate routine assessment slots for about 120% of that rate. Thus if a team that had 300 referrals in the previous year allocates 7 slots a week, there will be one extra available each week for urgent assessments and emergencies. Easy access to routine assessments reduces pressure for urgent and emergency referrals, and is much more efficient than emergency rotas.

Communication and liaison

Team meetings ensure internal communication, but CMHTs also need good links with the wide network of professional colleagues. Most routine communication takes place by letter, phone, or during individual clinical care. More structured liaison is advisable with primary care and general hospitals. Hospital links may be between specific CMHTs and wards, or CMHTs may provide input to patients from their sectors. Dedicated liaison psychiatry and psychosomatic services are common in well-resourced services.

General practice liaison. GP liaison systems originated in Balint groups but are now more likely to involve shared care or co-location of CMHTs in GP health centres (Burns and Bale, 1997). An effective and sustainable system involves regular (usually monthly) timetabled meetings between the two teams, or a ‘link’ CMHT member attending the GP health centre. It is important always to be clear about responsibilities; blurring of boundaries is risky.

Liaison with other agencies. The same principles apply to liaison with other agencies (social services, housing, and charitable and voluntary sector providers). Whether regular meetings are cost-effective will depend on the volume of shared work. Professional confidentiality and information sharing are more sensitive.

Mental health services research

The last 30 years have witnessed a change in how mental health services are planned and evaluated. A new academic activity, mental health services research (MHSR), has come into existence which has attempted to subject services and systems to the same evidence-based medicine approach to rigorous testing (e.g. controlled studies, randomized controlled trials, meta-analyses) that has been used previously with treatments. This has had two profound effects. First it has internationalized thinking about optimal services. The language of science means that findings from another part of the world can be applied locally. Secondly, it has shifted the locus of control in service development away from clinicians and towards academics and policy makers.

There are undoubted benefits to MHSR. For example, thinking and communication are markedly clearer, and some ineffective practices (e.g. traditional ‘brokerage’ case management) have been identified and abandoned. However, there are also drawbacks, particularly the risk of over-interpretation of results and their application without recognition of the modifying effects of the local context. This is briefly outlined in the cases of Assertive Community Treatment (ACT) teams. Three fundamental conceptual problems have characterized MHSR:

• a tendency to equate a service with a treatment, rather than with the platform for delivering treatment

• the assumption that the ‘control’ services in randomized controlled trials are inert and consistent, when in fact they are active and variable ‘comparator’ services

• a tendency to ignore the ‘pioneer’ effect—shiny new demonstration services with charismatic leaders have an inherent advantage.

MHSR methodology is improving, but these pitfalls in interpreting studies always need to be considered.

Mental Health National Service Framework (NSF) 1999

Mental health services in England and Wales were radically reformed following the publication of the National Service Framework in 1999 (Department of Health, 1999) and the NHS Plan in 2000 (Department of Health, 2000). These drew heavily on MHSR and pioneering services in North Birmingham and Australia. They initially recommended the complete replacement of the UK’s sectorized generic CMHTs with four ‘functional’ teams—assertive outreach, crisis resolution and home treatment, early intervention and primary care liaison teams. The sheer impracticality of such a massive reorganization (for which there simply were not adequate trained staff) and challenges to the strength of the evidence modified the proposal, and CMHTs were retained and primary care liaison teams abandoned.

An ambitious programme to establish the first three types of functional teams was put in place, starting with the creation of 300 assertive outreach teams (AOTS). These three teams are described below. The research evidence will be briefly addressed but, as demonstrated by that on AOTs, can change radically over time.

Assertive outreach teams (AOTs)

The most extensively replicated and researched specialist CMHT is the AOT. This is based on the assertive community treatment (ACT) model, whose landmark study by Stein and Test was published in 1980. This demonstrated improved clinical and social outcomes with substantially reduced hospitalization at slightly lower overall costs. Two Cochrane reviews with meta-analyses, one of case management showing an increase in hospitalization, and one of ACT showing a reduction (Marshall and Lockwood, 1998; Marshall et al., 2001), were instrumental in the adoption of ACT in the UK. AOTs (see Box 21.1) are costly, with one full-time case manager for about 10 patients. Consequently they are reserved for the most difficult (‘hard to engage’ or ‘revolving-door’) psychotic patients, who have frequent, often dangerous, relapses.

The AOT approach is based on proactive outreach—visiting patients at home and persisting with visits even when patients are reluctant. It exploits teamworking, with daily meetings and several members working with most patients, rather than exclusive individual relationships. This is needed for reasons of safety (it may not be safe to visit some patients alone) and also because of the complexity of patients’ needs. The approach is very practical (e.g. it involves taking patients shopping, sorting out their accommodation, and delivering their medicines daily if necessary) and goes well beyond traditional professional boundaries. Despite strong convictions that AOTs must closely follow Stein and Test’s original practice (assessed by ‘model fidelity’ measures) (Teague et al., 1998)) or outcomes will be poor (McHugo et al., 1999), the evidence for this is questionable (Fiander et al., 2003; Burns et al., 2005). Local clinical adjustments are both sensible and justified.

Box 21.1 Core components of assertive community treatment (ACT)

• Assertive follow-up

• Small caseloads (1:10–1:15)

• Regular (daily) team meetings

• Frequent contact (weekly to daily)

• In vivo practice (treatment in home and neighbourhood)

• Emphasis on engagement and medication

• Support for family and carers

• Provision of services using all team members

• Crisis stabilization 24 hours a day, 7 days a week

Who and what are AOTs for

Assertive outreach research has been conducted exclusively with psychotic patients, generally those with poor medication compliance and often with comorbid alcohol or drug abuse and offending behaviour. AOTs generally take only patients who cannot be stabilized despite CMHT support.

Enthusiasm for AOTs has often led to a preoccupation with their structure and process rather than with what they do. Despite the extensive early evidence of a reduction in hospitalization (Marshall and Lockwood, 1998), this has never been achieved against a CMHT control (T. Burns et al., 2002), and never in the UK, despite careful attention to faithful implementation (Killaspy et al., 2006, 2009). Careful examination of AOT studies (Burns et al., 2007b) confirmed that AOT/ACT only reduced hospitalization when bed management was poor, and this was not generally the case in the UK. More strikingly, the intensive staffing levels in AOT conferred no extra benefit. Despite very promising beginnings, there is no clear evidence that AOTs are routinely superior to CMHTs, and many of them are now being disbanded.

Crisis teams

Crisis intervention therapy in the early 1970s built on Caplan’s work (Caplan, 1964), but this psychodynamic focus has faded. Crisis teams have been most prominent where local services are poorly developed (they may be the onlycommunity services) or social mobility is high (e.g. centres of major cities), with many transient and homeless patients for whom continuity of care cannot be established.

Crisis teams are characterized by rapid response and accessibility. The duration of crises (and hence the window for intervention) has generated endless unprofitable debate. Most teams will see patients immediately, certainly on the same day. Their clinical aims (i.e. the assessment and management of acute episodes of mental illness) and staffing (mainly nurses and social workers) are similar to those of CMHTs. Given the acuteness of the problems, medical input is probably essential, despite attempts in some services to dispense with it. Crisis teams may be based in mental health services alongside CMHTs, or in the emergency rooms of general hospitals with 24-hour availability. Important functions include distinguishing acute mental illness from severe personal crises (e.g. involving self-harm), and accessing beds for both voluntary and involuntary admissions. Hospital-based crisis teams are often run in conjunction with liaison services (see Chapter 15).

Crisis resolution/home treatment (CR/HT) teams

CR/HT teams, which were introduced by the National Service Framework, were developed in Australia (Hoult, 1986). Their primary purpose is to avoid unnecessary admissions and to reduce the duration of inpatient care. They also are, in part, a response to increased consumer demand for access out of hours. They draw heavily on assertive outreach practice with limited, shared caseloads, flexible working, extended access, and an emphasis on outreach. Reduction in hospitalization is proposed to offset much of their cost (Smyth and Hoult, 2000). They aim to treat patients who would otherwise be in hospital, and to focus on the severely mentally ill. They offer intensive visiting (usually daily for a limited period) and considerable practical support, and they work with patients’ social networks. Most aim to limit themselves to 6 weeks of involvement. Such intensive teamworking requires highly effective communication, and the teams meet daily (in some cases twice daily). Information transfer is burdensome, and liaison with CMHTs is complex, requiring absolute clarity about local arrangements for clinical responsibility.

Variations in practice and sustainability

Crisis teams are particularly difficult to research rigorously, as the informed consent required for randomized controlled trials is almost impossible to obtain. However, there is some evidence that CR/HT teams may reduce the need for hospital care (Hoult, 1986; Johnson et al., 2005), although this is not clear-cut (Glover et al., 2006). As with AOTs, the potential for impact will depend on local provision. Although the UK model is well specified (including who it should and should not care for) (see Box 21.2), in reality practice varies considerably. Unless local services are particularly undeveloped, a full 24-hour service is very wasteful, and an on-call facility to the emergency room and police station at night usually suffices. In practice, contact frequencies are generally lower, patients stay with the service longer than anticipated, and restricting the service to patients who would otherwise be in hospital is rarely achieved. CR/HT teams are inevitably referred individuals who are recurrently in crisis (often with alcohol and relationship problems), who cannot be refused care but would not ‘otherwise be in hospital.’

Box 21.2 Remit of UK crisis resolution/ home treatment teams

‘Commonly adults (16 to 65 years old) with severe mental illness (schizophrenia, manic depressive disorders, severe depressive disorder) with an acute psychiatric crisis of such severity that, without the involvement of the CR/HT team, hospitalization would be necessary.

The service is not usually appropriate for individuals with:

• mild anxiety disorders

• primary diagnosis of alcohol or other substance abuse

• brain damage or other organic disorders including dementia

• learning disabilities

• exclusive diagnosis of personality disorder

• recent history of self-harm but not suffering from a psychotic or serious depressive illness

• crisis related solely to relationship issues.’

Medical input and the team’s independence from CMHTs both vary markedly. Some teams operate completely parallel to CMHTs, some share extensively (often acting as an adjunct to CMHT care—for example, making evening medication drops), and some insist on CMHT involvement. Local circumstances (and staff availability) will ultimately determine the configuration. However, clarity about responsibility and leadership is crucial. Patient evaluation of CR/HT teams varies. Many appreciate the promptness of response and the level of support. Others complain about intrusiveness and having to explain themselves to a series of different staff who are often insufficiently trained to make the necessary decisions.

Traditional crisis services may have a relatively limited lifespan. A World Health Organization review found that many early crisis services had ceased to exist within a decade (Cooper, 1979). Many had successfully improved local access and consolidated as a more general service (including, but not restricted to, crisis work). Others had become overwhelmed with inappropriate referrals or patients who could not be referred on, and eventually closed. A similar pattern has been observed (but not documented) in crisis houses.

Crisis houses and respite care

Excessive formality with regard to hospital care may prolong admissions and dissuade patients from seeking help. ‘Crisis houses’ allow admission with a minimum of formality, and often with reduced supervision. They are usually small (4 to 8 beds), in a domestic setting, and take patients for a period of days, or occasionally a week or two. They are favoured for vulnerable women and early intervention services. They do not replace inpatient care, but are generally very welcome for a minority of patients. However, without careful targeting and integration into local services they can become chaotic (and potentially dangerous) or blocked.

Early intervention services (EIS)

Early intervention services are the third functional team introduced by the National Service Framework. As discussed in Chapter 11, a long duration of untreated psychosis (DUP) in schizophrenia is known to be associated with poorer outcomes (Marshall et al., 2005). EIS teams have been developed mainly from Australian and UK models (Birchwood et al., 1998; Edwards et al., 2000). They aim to shorten DUP, protect current social functioning and networks, and improve outcome by providing prompt effective treatment. Despite a detailed prescription, UK teams vary considerably and prioritize different features of the approach. Some minimize the importance of diagnosis in favour of easy access, while others restrict themselves to schizophrenia in order to maximize skills. The duration of involvement ranges from 18 to 36 months, with proposals of up to 60 months. Some have a narrow age band (a ‘youth service’), while others take all first-episode patients irrespective of age (Burns, 2004). Three quite different activities are reported in the EIS literature, which may or may not be part of any given team (see Box 21.3).

Case management

The core of EIS is a specialized CMHT that case manages first-episode psychosis patients. Teams prioritize the protection of social networks and functioning (by keeping the patient at college or work, and emphasizing family interventions, etc.). They assume that there will be a return to premorbid functioning, and hospitalization is avoided (crisis and respite houses are preferred) and welfare benefits are minimized. An assertive outreach approach with practical and shared case management is used during the acute phase, with less intensive follow-up thereafter. ‘Zoning’ is commonly used, with ‘acute’ patients seen at least weekly, recovering patients seen weekly to monthly, and a third category for those who are still open to the team but not in active treatment. Teams differ in terms of whether they insist on keeping patients registered when they have recovered and do not want contact. They may or may not provide inpatient care.

Box 21.3 Components of early intervention teams

• Case management—ongoing care of identified patients

• Early identification—awareness-raising campaigns for psychoses

• High-risk and prodromal patient identification and treatment

EIS case managers aim to prevent patients and their families from lowering their expectations prematurely. As mentioned above, many avoid obtaining benefits for patients, encouraging as rapid a return to work or college as possible. Self-management of illness is strongly endorsed, with an emphasis on learning relapse signatures and individually appropriate responses.

Early recognition and high-risk intervention

Some EIS teams are actively involved in psychosis public awareness campaigns and lecturing in schools and colleges (McGorry and Jackson, 1999). A minority attempt to identify ‘ultra-high-risk’ patients with prodromal symptoms of schizophrenia and offer them treatment (both psychological and pharmacological) in order to prevent progression to psychosis. There is some evidence to support this approach (McGorry et al., 2002), but it remains highly controversial and is generally restricted to research teams.

Experience with EIS teams was much less than with AOTs when they were introduced in the National Service Framework, and they included confusing and contradictory prescriptions. Even more than with AOTs, the difficulties of making them work in dispersed rural and semi-rural settings have posed real challenges.

A settled model has not yet evolved for EIS teams. Indeed all three of the National Service Framework functional teams are still evolving rapidly and may change significantly (or even disappear) in the lifetime of this edition of this book. Understanding how to work within them or refer to them requires an understanding of local practice, and avoiding relying too much on published policy documents.

The evidence for the efficacy of EIS interventions is covered in Chapter 11.

Forensic and rehabilitation teams

Community-focused services face particular difficulties with treatment-resistant patients, especially those with socially unacceptable or offending behaviour. Such patients do not fit well into open wards. Specialized forensic teams provide care where offending behaviour and danger to others predominate. They provide secure inpatient care (in addition to their traditional roles with prisons and the courts), which is increasing internationally (Priebe et al., 2005). Some forensic teams provide community services (intensive case management of dangerous patients), but integrating them with general services can be controversial. The focus on risk assessment and extensive joint visiting are the main ways in which they differ from AOTs.

Rehabilitation services have reached out from long-stay wards to more community placements. A significant number of patients remain profoundly disabled despite receiving optimum treatments. Rehabilitation teams focus on the long-term management of disability, rather than on symptoms and episode-based care. The threshold from general services to rehabilitation services varies enormously. Rehabilitation teams generally serve patients who cannot survive without supervised accommodation even between relapses. These include a shrinking group of long-stay patients, but also increasingly a very disturbed ‘new long-stay’ population with comorbid substance abuse and behavioural disturbances. Inpatient provision is needed for this group, and a range of hostel care can promote stability, but is difficult to sustain (Hyde et al., 1987). ‘Hospital hostels’, where patients can live in hostel-type domestic accommodation, yet be subject to the MHA, have been successfully established by some rehabilitation services.

Diagnosis-specific teams

Highly specialized teams for specific disorders (e.g. eating disorders, personality disorders, bipolar disorder) focus on and concentrate specific skills. They provide specialized treatments, usually at regional level. They are strongly supported both by their professionals and by the families of sufferers. However, the opportunity costs of establishing such teams in a health service with finite resources require careful consideration.

Other components of a mental health service

The main components (teams and facilities) of specialized mental health services have been outlined above. It will be clear from that description that some of these components (e.g. inpatient wards, generic-sector CMHTs) are fairly ubiquitous and durable, while others (e.g. most of the specialized teams) are more variable in their availability and less consistent in practice. The way in which the functions and obligations of mental health services are distributed between these different teams will vary. The most important of these components will now be outlined and described. They can be summarized as follows:

1primary mental healthcare teams which support primary care and conduct initial assessments

2outpatient care

3specialist psychological treatments which are often integrated into the specialist or primary mental healthcare teams

4crisis/home-based treatment teams which treat acute mental disorder in the patient’s home

5Accident and Emergency liaison services, which are often part of the crisis service, that assess and treat people who present to Accident and Emergency departments

6general hospital liaison services for medically ill people who have mental health problems

7assertive outreach teams for severely mentally ill people who are difficult to engage or high users of services

8early intervention services for people during their first episode of psychosis and the 3 years thereafter

9employment services, including supported employment which helps people to find paid employment and supports them in it through job coaching, and services that link to training and education

10vocational rehabilitation, which includes sheltered workshops, transitional employment, and the Clubhouse model

11day hospitals and day care services, including drop-in centres

12self-help and service user groups and advocacy services

13inpatient services

14forensic services, with accommodation for high, medium, and low levels of security, and court-diversion schemes

15other residential services, including short-term accommodation outside hospital for acutely ill people or those in other forms of crisis

16. services for specific groups, such as people with eating disorders, and mother and baby units

17social and welfare services, including help with obtaining benefits, advocacy, community support workers, and home help/meals

18services for special groups, consisting of outreach services to marginalized groups, such as the homeless and refugees.

Services for psychiatric disorder in primary care

Classification of psychiatric disorders in primary care

ICD-10 and DSM-IV were developed for use in psychiatry. They are too detailed for routine use in primary care, and their fine distinctions are seldom helpful when selecting treatment in this setting. The World Health Organization has therefore developed a simpler classification for use in primary care. Each diagnosis is linked to a plan of management.

Identification of psychiatric disorders in primary care

The first stage of providing services for psychiatric disorders in primary care is to detect them. This is not a simple matter, as many patients present with both physical and psychological problems. There is evidence for considerable undetected psychiatric disorder among those attending their general practitioner (see Figure 21.1). In addition, some people with a psychiatric disorder complain only of physical symptoms. These can be of a coincidental minor physical illness or part of the symptoms of the psychiatric disorder (e.g. palpitations in an anxiety disorder, or tiredness in a depressive disorder). Some describe physical rather than psychiatric symptoms because they fear that the doctor may not respond sympathetically to a psychiatric illness. Recent studies have indicated that it is not just the patient who is responsible for this focus. Often patients mention both psychiatric and physical symptoms but the doctor focuses on the physical ones. Other patients are unaware that their physical symptoms have a psychological origin. The term conspicuous morbidity refers to the cases that are detected, and the term hidden morbidity refers to the rest. Hidden morbidity is generally less severe than conspicuous morbidity.

How effectively general practitioners identify undeclared psychiatric disorder depends on the following:

• their ability to gain the patient’s confidence, thus enabling them to disclose the psychiatric problems of which they are aware

• their skill in assessing whether physical symptoms are caused by physical or psychiatric illness; this requires a good knowledge of physical medicine as well as psychiatry.

Disorders that are treated in primary care

Most psychiatric disorders in primary care attenders can be treated successfully by the general practitioner or another member of the practice team. Examples include most adjustment disorders, the less severe anxiety and depressive disorders, somatization, and some cases of alcohol misuse. For a more detailed account, see Goldberg et al. (2009).

Images

Figure 21.1 Pathways to care in the UK.

Patients presenting with physical symptoms. In primary care, many patients with psychiatric disorder present with physical symptoms. Such patients are often called somatizers. Many of them will reveal associated psychiatric symptoms and accept a psychiatric diagnosis when interviewed appropriately—these are termed facultative somatizers. Others deny any psychiatric symptoms and reject a psychiatric diagnosis, however skilfully they are examined—these are called pure somatizers.

Distressed high users attend frequently with a variety of complaints. Most are female and middle-aged (Koch et al., 2007). Around 80% have a present or past psychiatric disorder, usually depression, somatization, or generalized anxiety, and around 60% have a concurrent physical illness (Katon et al., 1990). Since most refuse to be referred to a psychiatric team, the general practitioner has to manage them. One approach is to help to make links between symptoms and stressful life events, leaving out psychological processes, and to discuss how these stresses might be reduced.

Disorders that are referred from primary care to the psychiatric services

Table 21.1 shows that on average about one in four of the patients with psychiatric disorder who are identified by general practitioners is treated by the psychiatric services. This referred group includes patients with severe depressive disorders, bipolar disorders, schizophrenia, and dementia. General practitioners are more likely to refer patients with other disorders when:

• the diagnosis is uncertain

• the condition is severe

Table 21.1 Pathways to care with rates of psychiatric disorder among adults at each level of care*

Images

• there is a significant suicide risk

• the condition is chronic

• the necessary treatment cannot be provided by a member of the primary care team

• previous treatment in primary care has been unsuccessful

• psychiatric services are accessible and responsive

• the patient is willing to attend.

Treatments provided by the primary care team for acute disorders

Acute problems are often treated with counselling, used alone or combined with medication. Some practices also provide simple behavioural treatments. Many primary care teams include a counsellor. The availability of a counsellor has not been shown to reduce the prescribing of psychotropic drugs in the practice (Mynors-Wallis et al., 1995; Sibbald et al., 1996b).

Improving access to psychological therapies (IAPT)

In response to complaints about the variability of psychotherapy and psychological treatment provision nationally, a nationwide provision has been introduced—improving access to psychological therapies (IAPT). IAPT is a highly prescribed process, but local flexibility has been encouraged with regard to who provides it and where (e.g. specialist mental health services, primary care, voluntary bodies). IAPT has two components—a stepped-care treatment of anxiety and depressive disorders, and a supported employment service (see below). The supported employment service aims to help patients to return to work when their anxiety and depression symptoms have receded (reducing unemployment due to psychiatric disorders was the other political stimulus for IAPT).

The first step in IAPT is self-help using telephone and Internet programmes. A form of brief cognitive–behaviour therapy (CBT) is offered to those with residual symptoms, and referral on to specialized services only takes place after both self-help attempts and CBT. A striking feature of IAPT is the level of structure—therapists use detailed manuals, and symptomatic status is recorded after every session using the Patient Health Questionnaire (PHQ-9) (Kroenke et al., 2001) for depressive symptoms and the Generalized Anxiety Disorder Assessment (GAD-7) (Spitzer et al., 2006) for anxiety symptoms.

The IAPT programme has only recently been rolled out nationally after piloting, and it is too early to know exactly what impact it may have. In particular, the balance of resources between simple self-help programmes and more intensive, short CBT treatment is still being explored. Some services emphasize active promotion of self-help, often relying on easy onward referral for CBT. Others have devoted more resources to providing nonspecialist-service brief CBT. The IAPT programme also includes a commitment to vocational rehabilitation, most often in the form of supported employment (see p. 619). The format of this provision is highly variable and as yet unevaluated. The ease of access and clear commitment to psychological treatments have won widespread approval. Those who provide the service are positive about its outcome, but to date there has been no independent research evidence of its impact.

Treatments provided by the primary care team for chronic disorders

The respective roles of the general practitioner and the community team should be clearly defined in relation to any patient with chronic mental disorder, and reviewed regularly. For example, with regard to the care of some patients with chronic schizophrenia, the general practitioner might be responsible for their physical health, assess their general progress, administer and encourage compliance with medication, and support the family. Kendrick et al. (1995) found that even with additional training, general practitioners are not very reliable at making structured assessments of patients with long-term psychiatric illness. However, subsequent research demonstrated that this was possible, and a dedicated Quality and Outcome Framework (QOF) payment is made to practices that maintain a register of their patients with severe mental illness and provide them with annual structured health checks. Dedicated QOF payments may not survive NHS reorganization, but such physical monitoring has become increasingly important because of the high rates of metabolic syndrome associated with the atypical antipsychotics. It is generally better that the psychiatric team assesses these patients and agrees a plan with the general practitioner. This plan will include elements provided by the psychiatric team and by the primary care team, as in the above example.

Work in primary care by the psychiatric team

There are several ways in which a psychiatric team can work with the primary care team, and these have changed over time (Burns and Bale, 1997).

Advising and training general practitioners and their staff (Balint groups)

In this style of working, the psychiatrist and other members of the team do not see patients, but give advice based on the general practitioner’s assessment of patients. This approach, the first used, stems from psychoanalytically inspired ‘Balint’ groups (Balint, 1964). They are valued by a small group of highly motivated GPs, but are mainly of historical interest.

Assessing and referring

The psychiatrist assesses patients when the GP is uncertain about diagnosis or treatment. He may do this on his own or jointly with the general practitioner. Patients who are identified as needing specialist treatment are then referred to a psychiatric outpatient clinic in the usual way. This approach is now very rarely used, and has receded as liaison with GPs has improved.

Assessing and treating (‘shifted out patients’)

The psychiatrist works mainly in primary care, seeing most patients at the primary care centre or at home, rather than in the hospital outpatient clinic. Clinical psychologists and psychiatric nurses can also work in primary care, providing assessment, counselling, or behavioural treatment. Patients no longer need to visit a psychiatric clinic, but there may be little contact between the psychiatric and primary care teams. This approach, which was popular in the 1980s, has been shown to be generally ineffective (Wooff and Goldberg, 1988), as the focus shifts and skills are lost. There is slightly more positive evidence for social workers located in primary care.

Shared care

This approach is consistent with the goal of ‘seamless care.’ It has had enthusiastic advocates (Essex et al., 1990) and several well-publicized demonstration services. However, these demonstration services rarely survive the departure of their advocates, and shared care has not caught on generally. The term is ambiguous, covering everything from GPs monitoring lithium levels to completely shared responsibility. In the care of severely mentally ill patients, absolute clarity of responsibility has repeatedly been shown to be crucial (hence the Care Programme Approach), and the risk of confusion does not seem to be worth taking.

Liaison meetings

Probably the most useful and time-efficient method is to arrange regular (usually monthly) face-to-face meetings between members of the psychiatric team and the primary care team (Burns and Bale, 1997). This can range from meetings of the entire teams to relying on identified ‘link’ members. The important point is that it is face to face and takes place routinely—not just in relation to crises. Joint patients are discussed, together with possible referrals. Trust is established along with a more realistic appraisal of what each team can do. Such liaison meetings are increasingly required by commissioners.

Specialist services for acute psychiatric disorder

Patients who are referred to specialist care

Patients who are treated by the psychiatric services are a subgroup of people with mental disorders. In some countries patients can go directly to a specialist, so that those treated by the psychiatric services may not be very different from those treated in primary care. In countries such as the UK, where the general practitioner acts as the ‘gatekeeper’ to specialist services, the number and types of patient reaching the psychiatric services depend on:

• the willingness of general practitioners to treat psychiatric disorder

• the treatment skills and resources of the primary care team

• patients’ willingness to attend for specialist psychiatric advice

• the general practitioner’s criteria for referral to the psychiatric services

• the psychiatric services’ criteria for accepting referrals.

In the UK, most patients in contact with the psychiatric services have severe and chronic anxiety disorders, severe mood disorder, schizophrenia and other psychoses, or dementia. Patients who are particularly likely to be cared for by specialists are those who are suicidal, those who are potentially dangerous to others, and those with dual diagnoses of mental disorder and substance misuse or personality disorder.

Provisions for acute specialist care

Specialist care of acute psychiatric disorder requires community teams, supported by outpatient, day patient, and inpatient provisions.

Generic versus specialized services

Community teams have traditionally been generic, dealing with all kinds of adult psychiatric problems. More recent specialized separate teams for crisis management and continuing care are described above. The potential advantages of each of these arrangements are summarized in Table 21.2.

Table 21.2 Generic vs. specialist services

Images

Day hospitals

Day care is used in four main ways:

1Acute day hospital care for patients with acute illness who would otherwise be admitted to hospital. Such care is generally more appropriate for those who can be with their families in the evening and at weekends. Suitable conditions include depressive disorders of moderate intensity and without substantial risk of suicide, anxiety and obsessional disorders, and some eating disorders. Nine trials of acute day hospital treatment (involving 1568 people) found it feasible for 25% of those currently admitted to inpatient care, with comparable rates of improvement and readmission. Four out of five trials found that day hospital care was cheaper than inpatient care (Marshall, 2003; Marshall et al., 2003). However, such acute day hospitals are very different from routine day hospitals, and have failed to generalize. They seem to need either a charismatic leader or a research programme to sustain them.

2Transitional day care to allow earlier discharge from hospital.

3Day treatment programmes that provide more intensive treatment for people who are not responding to outpatient care. There is evidence for the value of this approach in personality disorder (Bateman and Fonagy, 2001), and it is increasingly being used for severe eating disorders.

4Day care centres for the support of long-term service users. Such care needs to be planned carefully with an active treatment programme that is specific to each patient’s needs, otherwise attenders may become dependent and ‘stuck.’

Day hospitals and day centres vary considerably and overlap in their clientele and activities (Catty et al., 2005). Increasingly, CR/HT teams are based in and integrated with day hospitals.

Inpatient units

The number of beds required for acute psychiatric disorder is difficult to determine exactly, as it depends on:

• the level of morbidity, including substance misuse, in the population served by the unit

• the willingness and ability of families to care for acutely ill relatives, which in turn depends on the quality of accommodation and the availability of other family members

• the extent and availability of crisis services

• facilities for treatment of acute psychiatric disorder outside hospital, such as well-staffed hostels

• facilities for early discharge of patients from hospital after the acute phase of the disorder; these facilities resemble those under the previous two points, although it is generally easier to discharge early than to avoid admission.

The design of inpatient units for acutely ill patients should strike a balance between the patients’ needs for privacy and the staff’s requirement to observe them. They need to be sufficiently large to ensure adequate staffing for emergencies, yet avoid an anonymous or ‘institutional’ feel. Three 15- to 20-bedded wards is considered to be the minimum necessary for safe effective care, and units of over 100 patients are considered probably too large. Ward design should minimize the possibility of suicide by hanging. There is a need for secure areas for the most disturbed patients, areas where patients can be alone, and areas where they can interact with others. There should be provision for occupational therapy, the practice of domestic skills, and recreation. Outdoor space is desirable. Single-sex accommodation is now mandatory.

The siting of acute inpatient units. Inpatient care for acute psychiatric disorders is generally provided as part of a general hospital complex. This siting reduces stigma and provides easy access to general medical services when required. The disadvantages of such siting include the difficulties of providing adequate space for occupational activities and of creating an informal environment suitable for psychiatric care in a hospital designed primarily for the different needs of physically ill patients. Many of these problems can be overcome if the psychiatric unit occupies a separate building within the general hospital complex.

Collaboration with community teams. Patients are at increased risk, particularly for suicide, when they move from inpatient to community care. About 25% of suicides among psychiatric patients in England and Wales occur within 3 months of discharge from inpatient psychiatric care, with a peak in the first 1–2 weeks. Most commissioners require contact in the week following discharge.

Psychiatric services that provide long-term care

Characteristics of patients who require long-term care

Diagnosis

With the exception of the elderly (who are considered in Chapter 18), most psychiatric patients who require long-term care have schizophrenia, chronic affective disorders, presenile dementia, or personality disorders associated with aggressive behaviour or substance misuse. Patients who need care in hospital for more than 1 year, other than those in forensic services, are sometimes referred to as the ‘new long-stay’ (in contrast with the ‘old long-stay’, who had been resident in hospital for many years before hospital closure programmes were initiated).

Problems

There are several ways of classifying the problems of patients who need long-term psychiatric care. They can be divided into seven groups, three of which are contained in the World Health Organization’s classification of disablement (see p. 23), on the basis of the following problems:

• symptoms such as persistent hallucinations or suicidal ideas

• unacceptable behaviours, such as shouting obscenities, and threatening or carrying out violent acts

• impairments that interfere with the functioning of a psychological or physical system—for example, poor memory or lack of drive

• disabilities that interfere with the activities of the whole person, such as inability to dress

• handicaps which are social disadvantages directly resulting from disability—for example, inability to work, or to care for children

• other social disadvantages which are not directly related to disability, but which are a consequence of the stigmatizing attitudes of other people—for example, unemployment, poverty, and homelessness

• adverse personal reactions to illness and social disadvantage—for example, low self-esteem, hopelessness, denial of illness, or the misuse of drugs.

Wing and Furlong (1986) proposed a list of patient characteristics that make it difficult to treat the individual in the community, and that list is still useful (see Table 21.3). Patients with severe and persistent problems of this kind may need care in a well-staffed hostel, or in a hospital that can provide appropriate rehabilitation and security.

Table 21.3 Patient characteristics that make community care difficult

Risk of harm to self and others

Unpredictable behaviour and liability to relapse

Substance misuse

Poor motivation and poor capacity for self-management or for performance of social roles

Lack of insight into the need for treatment

Low public acceptability

Requirements of a community service providing long-term care

If patients with psychiatric disorder are to be treated in the community, it is necessary to provide all the elements of care that they would have received in hospital. The following seven provisions are required to replace long-term care in hospital:

1. suitable and well-supported carers

2. appropriate accommodation

3. suitable occupation

4. arrangements that enlist the patient’s collaboration with treatment

5. regular reassessment, including assessment of physical health

6. effective collaboration among carers

7. continuity of care and rapid response to crises.

Complicated and expensive arrangements are required to make these seven elements available as readily in the community as they are in hospital. Lack of these arrangements may leave patients homeless, without constructive occupation, inadequately treated, and without a carer. Failure of community care may also leave carers unsupported and family life disrupted. When community care began, it was concerned mainly with patients who had become institutionalized and compliant after many years in hospital. These patients could often be managed in the community without much difficulty. Younger patients who have spent less time in hospital are often less compliant and more difficult to manage.

The carers

When people with severe mental illness live at home, family and friends are the main carers. They provide much of the help that would have been given by nurses had the patient remained in hospital. For example, they may have to encourage the patient to get up in the morning, maintain personal hygiene, eat regular meals, and occupy him- or herself constructively. Carers also encourage collaboration with treatment. If patients have many problem behaviours, prolonged involvement in their care is stressful. Carers may then need support and advice, and sometimes periods of respite.

Volunteer carers play an important part in many systems of community care. Trained volunteers can help to support patients and their families, and some charitable organizations employ professional carers, such as hostel staff.

Community psychiatric nurses play an essential part in community care by supporting patients and carers, and by evaluating patients, supervising drug therapy, and encouraging social interaction. Monitoring is particularly important for patients whose mental state is unstable and who may act in a threatening or dangerous way when they are more disturbed.

Accommodation

Patients who have been discharged from hospital have obvious needs for food and shelter. Many live with their families, and some care for themselves in rented accommodation. Others need more help, which can be provided in three ways.

In lodgings. Some people are willing to receive patients with mental disorders as lodgers and to provide them with extra care. This practice works well in some countries, but has not been adopted widely in the UK.

In group homes. Some patients are able to live in group homes (i.e. houses in which four or five patients live together). The houses may be owned by social or health services or by a charity. The residents are often people with schizophrenia who have social handicaps but few positive symptoms. They are chosen as being able, with others, to perform the essential tasks of running the house, with each person using their remaining abilities. People who live in group homes receive regular support and supervision, usually from a community nurse or social worker, who ensures that the arrangements are working well and encourages residents to take on as much responsibility as possible.

In hostels. Much long-term residential care is provided in hostels. For some people, hostels are half-way houses from which they move on to more independent living. Others need to remain in hostels for a period of years. Although most hostel residents live fairly independent lives, a few of the most disabled require additional care. Levels of supervision vary, staff may sleep at night, or they may remain awake as in a hospital ward.

Occupation

Most people with severe and long-term psychiatric disorder would like to work, and some can work normally. Others require assistance, which can be of several kinds:

• Vocational rehabilitation is the longest established approach. Extensive training is provided to develop the skills and confidence needed for competitive employment. Training may take place in sheltered workshops or work crews, or in a transitional job in which the mentally ill person works for a while in a paid post ‘owned’ by the rehabilitation scheme. Unfortunately, few patients progress to open employment.

• The Clubhouse model is a specialized form of vocational rehabilitation that is based around a building which is jointly administered and maintained by people with mental illness (Beard et al., 1987). Originating in the USA, it has established itself internationally but is not widespread.

• Supported employment was originally developed for people with learning disabilities. Individuals are placed in competitive work without any extended preparation, but with support on the job (Bond et al., 1997). It has been shown in several randomized controlled trials (Bond et al., 2008) to be markedly more effective than vocational rehabilitation in obtaining open employment. Most of the studies have been conducted in the USA, but a European replication (Burns et al., 2007) confirmed equal efficacy (over 50% of psychotic patients became employed). It is government policy in the UK and a component of the IAPT programme, but provision remains patchy.

• Special sheltered work in which people can work productively but more slowly than would be acceptable elsewhere. Such work includes horticulture and the making of craft items.

• Voluntary activities. Some patients who cannot meet the requirements for sheltered work can undertake voluntary activities.

• Occupational therapy. Those who are more severely handicapped need occupational therapy in order to avoid boredom, under-stimulation, and lack of social contacts. These occupational activities are often provided in a day hospital or a day centre.

Reassessment

Patients who are living outside hospital require the same regular reassessments that they would have received in a long-stay hospital. Regular reappraisal of the patient’s mental state and of collaboration with treatment is usually performed by the key worker, who is often a community nurse, with planned but less frequent reassessment by a psychiatrist. Some patients with chronic disorders forget appointments, and it is important to have a recall system and take prompt steps to re-establish contact. It is important to review physical health as well as the psychiatric disorder, because patients with long-term psychiatric disorder may not seek help for physical illness, or may not accept the help that is offered. Physical assessment is therefore an important part of the general practitioner’s contribution to the care of these patients.

Risk assessment. It is more difficult to anticipate threatening or dangerous behaviour in community care than in hospital. An important part of reassessment is a regular evaluation of the degree of risk to the patient him- or herself, and to other people. Such assessment involves static factors such as the patient’s past behaviours, and dynamic risk factors such as the current level of substance misuse or of psychosocial stressors (see p. 51). Risk assessments lead to a risk management plan that is an important component of the overall care plan. The scientific value of structured risk assessment has been disputed (Wootton et al., 2008). However, such assessments are of value in demonstrating that clinical decisions have been taken after due consideration of the possible adverse consequences.

Continuity of care

Community care staff need to gain the confidence of their patients and to know them well. These aims cannot be achieved if there are frequent changes in staff. Continuity of care is important, and staff should be extra vigilant when care has passed to a new worker.

Response to crisis

Community care staff need to respond quickly to crises. Families and the staff of voluntary organizations accept patients more readily if they know that help will be available quickly in an emergency. Furthermore, readmission to hospital may be avoided by prompt action. A crisis plan that has been agreed in advance with the patient and carers is helpful. Staffing levels need to be adequate for a quick response, preferably by staff who know the patient or else by an emergency team.

Working with the family and volunteers

Community care is costly, and in most countries public funds are limited, with the result that arrangements often depend on contributions by families and voluntary groups. It is important that these families and voluntary groups are involved in the planning of services, and that there is agreement about their responsibilities and those of professional staff. Without this agreement, family members may believe that they are required to take on over-demanding tasks, and professionals may be concerned that volunteers are taking on tasks beyond their capabilities. It is good practice to involve families and voluntary groups in the evaluation of services.

Other components of a community service

Rehabilitation and recovery

In psychiatry, the term rehabilitation denotes procedures for helping patients to reach and maintain their best level of functioning. This help may be provided in an inpatient unit, a day hospital, or a rehabilitation centre. The procedures used in rehabilitation are medical, psychological, occupational, social, and residential.

Medical procedures. Most patients in rehabilitation programmes require medication to control symptoms of schizophrenia or of chronic affective disorder.

Psychological procedures. Psychological methods include supportive therapy, behavioural programmes, and social skills training (see Chapter 20).

Occupational procedures. Work helps to structure the day and provides an opportunity for interaction with other people. Good results can be a source of selfesteem, and payment is a further incentive. For most people the aim is to return to ordinary work; those who cannot achieve this are trained for activities such as gardening, crafts, and cooking, all of which can provide a sense of achievement and a constructive use of time. Where possible, training takes place in ordinary community facilities in order to avoid the segregation and potential stigmatization associated with special activities.

Social procedures. Whenever practicable, patients should be encouraged to join social groups that are attended by healthy people.

Inpatient care

People with long-term mental illness may require inpatient care for acute treatment at times of relapse, for intensive rehabilitation, and very occasionally for long stay. The basic requirements for an inpatient unit for patients with long-term illness are broadly similar to those for patients with acute illness (see p. 616), except that accommodation can be more domestic. Nevertheless, because the pace and intensity of treatment are usually slower for people with long-term illness than for those with acute disorders, it is generally better if possible to separate the care of the two groups. Patients with severely disturbed behaviour need secure areas adequately provided with outdoor space, and occupational and social activities either in the same building or nearby.

Services for people with particular needs

Clinical work and research have identified several groups of patients whose needs may not be fully met in the usual psychiatric services, and who require special attention.

Members of ethnic minorities and migrants

Members of ethnic minorities have special needs related to their culture. Many also have needs related to poverty and other social disadvantage, and some are recent migrants or refugees. Members of ethnic minorities are also more subject to discrimination. The differences between two ethnic minorities may be greater than those between the first minority group and the general population. There is therefore no single best pattern of service applicable to members of all ethnic minorities. However, the following general points are widely relevant.

Use of the services

Members of ethnic minorities are less likely than other people to use services provided for the majority of the population. They are less likely to consult general practitioners when they have a psychiatric disorder, to accept referral to psychiatric services, and to comply with psychiatric treatment. When members of ethnic minorities ask for help, professional staff are less likely to identify psychiatric disorder, and less able to explain illness and treatment in terms that take into account the beliefs and cultural background of the patient. These and other problems, discussed below, are not due to a limited command of English but, more importantly, to cultural differences.

In the UK, these problems have been studied in particular among people of Asian and Afro-Caribbean origin. People of Asian origin consult their general practitioners more frequently about most conditions than do members of the general population, but consult them less about psychiatric symptoms (Murray and Williams, 1986; Gillam et al., 1989), and often present with physical symptoms. In some ethnic minorities, referral to a psychiatrist is avoided because it can affect marriage prospects.

Under-recognition of psychiatric disorder

There are two reasons why general practitioners and psychiatric staff may fail to recognize mental disorder in members of ethnic minorities. First, there may be problems of communication, which could partly be overcome by the provision of doctors who are members of the ethnic minority or by providing interpreters. Secondly, the presentation of psychiatric disorders may differ between members of ethnic minorities and the general population.

Provision of services

Two approaches have been proposed. Either special services can be provided or efforts can be made to render all services culturally competent. The advantages of dedicated services are obvious, and their disadvantages are as follows:

• They may reinforce isolation from the mainstream population.

• They are practical and cost-effective only when the ethnic-minority population is substantial and relatively homogeneous.

Homeless mentally ill people

When hospitals were closed, it was feared that many discharged patients would become homeless, but when closures were planned carefully, few of them became so (Leff, 1993; Harrison et al., 1994). Nevertheless, surveys have found high rates of chronic psychiatric disorder among the residents of hostels for the homeless (Fazel et al., 2008). The most common disorders were alcohol dependence (8–58%) and drug dependence (45–54%). Psychotic illness and major depression each ranged from 3% to 42%. Alcohol dependence has increased over recent decades.

Young people with persisting psychiatric disorder

Most psychiatric patients can be treated by means of a short stay in a psychiatric unit followed by intensive community care. However, there is an important group, composed mainly of young men with unremitting psychotic illness, who need prolonged care. Such people do not fit well into an acute admission ward, and are helped more by treatment in a less stressful environment. Many can be cared for in well-staffed hostels (‘hospital hostels’), provided that they can be admitted to an acute psychiatric unit if their disorder relapses. When planning community services, it has to be recognized that many members of this group are likely to require intensive care throughout their lives.

Early intervention services have been introduced for people with first episodes of psychosis. They are described on p. 294.

Patients with challenging behaviour

Most of these patients have schizophrenia, often accompanied by personality disorder or the misuse of alcohol or drugs. A few have brain damage. In the past, this small group remained in hospital for many years, where periods of aggressive behaviour could be identified early and treated quickly. Most are now cared for in the community, where they require intensive supervision, which is expensive to provide and may not be readily accepted by them. Even with intensive supervision, problem behaviours are often difficult to predict.

Since aggressive behaviour is alarming and may endanger other people, the public tends to judge the effectiveness of a psychiatric service on the basis of its ability to care for this small group. These patients are not managed well in a general psychiatric unit, where their aggressive or unpredictable behaviour may alarm other patients, and special units are needed. Some require long-term care in a sheltered and well-staffed hostel.

Doctors with psychiatric problems

Although doctors have tried to reduce the stigma associated with psychiatric disorder, many of them do not seek help if they develop such a disorder themselves. It is helpful to provide special arrangements to enable psychiatrically ill doctors (and other mental health staff) to obtain treatment away from their place of work. If staff are to seek treatment appropriately, a greater acceptance of psychiatric disorder is required within the medical profession, with appropriate arrangements for aftercare and a gradual (and in some cases supervised) return to work. This issue should be discussed openly during medical education and training. When psychiatric disorder is chronic or recurrent, difficult problems are encountered concerning fitness to practise.

Refugees

Refugees have the general problems of members of ethnic minority groups (described above), together with specific problems consequent upon the experiences that led them to seek refuge in another country. These experiences include persecution, physical injury, torture, or rape, the witnessing of the injury, torture, or rape of loved ones, and bereavement. The consequences include general medical as well as psychiatric conditions. The latter are mainly post-traumatic stress disorder and depressive disorder (described on p. 157 and p. 206, respectively), with culture-specific variations such as a predominance of physical symptoms in depressive disorders.

Refugees need integrated medical and psychiatric care from a team that is aware of cultural factors. Integrated care is important because general and psychiatric disorders occur together, and because refugees often come from cultures in which distress and psychiatric disorder are generally manifested by physical symptoms. In addition, aid workers are sometimes reluctant to refer refugees to a solely psychiatric service which they think inappropriate for what appears to be a normal response to overwhelmingly stressful circumstances.

Staff who provide services for refugees should be experienced in the treatment of post-traumatic conditions, and should be aided by interpreters. Female members of staff are needed to help female refugees, especially those who have experienced rape. Other staff should have experience in treating children and adolescents. For a review of the psychiatric problems of refugees, see Mollica et al. (2009), and for a review of such problems among child and adolescent refugees, see Fazel et al. (2005).

Some difficulties with community care

The burden on relatives

If members of the family are to take responsibility for patients, by housing them, encouraging adaptive behaviour, supervising their medication, and reporting signs of relapse, they need to be well informed, adequately supported, and able to obtain help in an emergency. Such support is time-consuming, and is rarely available unless community care is well resourced and the needs of the family are given high priority. Many carers report poor communication from psychiatric services (e.g. about discharge plans). Staff experience continuing conflict over their duty of confidentiality to the patient and their wish to support and inform the relatives. There is no easy answer to this dilemma. Carers also report difficulty in coping with negative symptoms and socially embarrassing or aggressive behaviour. Care plans should always consider the possible effects of the patient’s illness on any children in the home.

Problems with the distribution of resources

When resources are limited, there is a conflict between the needs of patients with acute disorders and those of patients with chronic psychiatric disorders. The conflict is most evident in primary care, where the former group of patients is generally more demanding of care and more responsive to it.

Problems with the coordination of services

In most countries, long-stay hospital care for patients with chronic psychiatric disorder is provided by a single agency (a hospital authority). Community care requires coordinated action by several agencies, each of which usually has other responsibilities—for example, social services departments have responsibilities for mentally healthy children and elderly people as well as for psychiatric patients. This broad responsibility generates impossible problems with regard to the allocation of resources. Regular negotiation is the only possible response.

International service principles

In countries that have few specialist psychiatric resources, it is essential to decide on priorities. The World Health Organization (1984) has identified four priorities:

1. rapid response to psychiatric emergencies

2. provisions for chronic severe psychiatric disorder

3. care for psychiatric disorders associated with general medical illness

4. care for any high-risk groups that are found in the country (e.g. drug misusers).

The World Health Organization advocates the training of auxiliary workers who can supplement the efforts of fully trained staff. Thus in countries with few trained psychiatrists, resources may be used most effectively by improving the skills of nurses who can perform the first-line management of psychiatric disorders, supported by general practitioners.

Further reading

Gelder MG, Andreasen NC, López-Ibor JJ Jr and Geddes JR (eds) (2009). Part 7: Social psychiatry and service provision. In: The New Oxford Textbook of Psychiatry. Oxford University Press, Oxford. (The 10 chapters in this part of the textbook review the whole field of community psychiatry, including psychiatry in primary care.)

Goldberg D and Huxley P (1992). Common Mental Disorders: a biosocial model. Routledge, London. (This book is still valuable for its account of the development and utility of the ‘filter’ model which relates epidemiology to the provision of services.)

Thornicroft G and Szmukler G (2001) Textbook of Community Psychiatry. Oxford University Press, Oxford. (A comprehensive work of reference.)