Background

Health care, including mental health care, in England is mainly provided by the national health service (NHS). This is financed from national taxation and administered from the Department of Health in London. Its quality and availability have, in recent decades, become important areas of political debate. As a result, the NHS has been the subject of frequent and high profile government initiatives; some directed at its general structure, others at the detail of patterns of care. Predictably, given its size, it has proved relatively slow to change.

Organisational structure

Local NHS administration is complex [1]. ‘NHS trusts’ are care-providing organizations, which run hospitals, and employ a wide range of staff to provide services in community-based settings. In most cases mental health provider trusts undertake this alone or in combination only with community-based nursing and other services for people with long-term illness. Local ‘commissioning’ organisations, currently (in 2006) called primary care trusts (PCTs), are given the annual health care budget for a geographically defined population and required to organise contracts with NHS trusts for the health care they require. A few PCTs run specialist health care services directly. At present England’s population of just under 50 million is divided between 303 PCTs with a median population of 152,000, (interquartile range 113,000 to 198,000).

It is important from the outset to identify three distinctions. Health care policy and planning, and consequently related statistical data gathering, all assume these distinctions which can cause confusion in comparisons with data from other countries.

  1. 1.

    Learning disability (the group of problems in the F9 section of the 10th revision of the international classification of diseases) is not considered to be a mental illness. Services for learning disabled people are managed by almost completely separate services. This paper only deals with services for people with mental illness.

  2. 2.

    Services are sharply delineated into those for working age adults (aged 18–64), for older adults, and for children and adolescents. This separation is reflected in the policy-making apparatus in the Department of Health where older adult and child and adolescent services are the responsibility of age-defined policy branches, while services for working age adults are managed by a mental health policy group. It is also often seen in statistical collections, which are thus only partly compatible across age groups. This paper deals with services for working age and older adults.

  3. 3.

    Social care is distinguished from health care and organized and funded in a different way [2]. Social care needs include difficulties with mobility, personal care, organising a home, finding satisfactory things to do during the day, and other functions with which a family member, if available, could assist. Where a disabled individual does not have a willing or able family, their care needs are assessed by the social services department, a branch of local government. A social worker may then arrange for the services to be provided by one of a range of organizations run by local government or independently. These services are charged-for where the individuals needing them can afford to pay. This paper attempts to deal with both health and social care.

History

The pattern of care for mentally ill people in England has undergone slow, progressive transformation since the late 1940s. The contributions of policy makers in this process have mostly either been reactions to emerging trends, or consequences of wider policy initiatives aimed at physical health and social care issues. During the 1950s it was observed that the numbers of people becoming long stay patients with schizophrenia or other chronic mental disorders had dropped below replacement level and consequently the resident population of the asylums was beginning to fall sharply [3]. Early in the 1960s Government policy embraced the recognition that the era of the asylums was drawing to a close. In the 1962 NHS hospital plan, acute inpatient psychiatric care was considered a component of the district general hospital [4]. During the 1960s a lot of work was undertaken to characterise key statistical features of the process of mental health care as a community, rather than asylum-based activity. In 1975 a major policy document, Better Services for the Mentally Ill was published [5]. This described the structures appropriate to a fully community-based mental health service. However its publication coincided with a period of economic recession and in her foreword to it, the minister responsible indicated that she saw little immediate scope for progress in the development of new services.

In July 1984 a psychotic 19-year-old woman killed a social worker in a south London hospital [6]. The inquiry into this event raised concerns about the supervision and monitoring of people with severe mental health problems living outside institutional care. During the later 1980s, largely as a result of changes in eligibility for social security benefits, homeless people became a common sight in the streets of major cities. For England this was new. It was evident that a noticeable minority was suffering with severe mental health problems. In this context a new type of government mental health policy, The Care Programme Approach emerged [7]. For the first time this set out to prescribe a framework for clinical practice; its purpose was to mandate a quality standard aimed at minimizing risks to and from chronically mentally ill people living in the community. Two further tragic incidents at the end of 1992 lead to increased emphasis on supervision and security. At the same time, as a part of the UK government’s response to the WHO health for all process, targets for suicide prevention were developed.

The change of government in 1997 led to sharp increases in overall health service funding. Policy about mental health care initially followed the cautious line of the previous government [8, 9]. Subsequently a major new policy paper, the National Service Framework for Mental Health [10] was published. This followed a wide consensus building exercise. Detailed policy implementation guidance, including targets, emerged over the following 18 months [1, 11]. These documents set out a blueprint for a new style of mental health care service. Some details are described in the next section. Responsibility for their implementation was assigned to locally self-determined groupings of NHS, local government and independent sector providers and commissioners, working with service-user and carer representatives in ‘Local Implementation Teams’ (LITs). The number of these has varied over time; in 2003 there were 174 covering a median population of just under 230,000 each (interquartile range 171,000–304,000).

In addition to this work specifically addressing mental health care, the new government introduced two types of measures to help co-ordination between NHS and local authority services for a range of patients with long-term needs including those with mental illness. The first allowed for a range of budget sharing mechanisms [12], the second for the setting up of ‘Care Trusts’, hybrid organisations to perform both health and social care functions [11].

Mental health care provided outside the NHS falls into three broad categories. A small amount of general outpatient and inpatient mental health care is provided for people who prefer to avoid NHS services and can afford (either directly or through private health insurance) to do so. A substantial proportion of the psychodynamic psychotherapy that occurs is outside the NHS. Finally, a significant amount of inpatient care is provided by independent sector hospitals but funded by the NHS. This is discussed in more detail below.

Mental health legislation

The most important comprehensive review of mental health law in England to date was the 1959 Mental Health Act; this was updated in 1982, primarily to strengthen protection for patients, and in 1995 to provide for supervision of a small number of patients outside the hospital [13]. This body of law assumes that most mental health care should take place on the basis of patient consent. Two types of provision are made for compulsory treatment. Doctors and social workers are given powers to detain people in the hospital both in the short-term for assessment and in the longer term for treatment. Except in emergencies, these powers require the agreement of two doctors (who must not work for the same organization), and a social worker, who is seen as providing an important alternative perspective. Criminal courts have similar powers to transfer individuals accused or convicted of criminal offences to hospitals for assessment or treatment. Patients subject to any of these compulsory powers can appeal against them to the Mental Health Review Tribunal, a court that is established under the judiciary not the Department of Health. Where patients are detained for more than 3 years, Tribunal reviews are automatic.

In recent years there has been considerable debate in two areas, and further legislation is expected soon. The first area concerns the treatment and detention of people with ‘dangerous and severe personality disorders’. At issue are plans to allow for preventive detention of individuals whose disorder makes them likely to commit serious criminal offences against other people, but who have not yet done so, and for whom it is at best contentious whether there is any effective treatment. The second concern is the perceived need to make further provision for compulsion in treating a small number of individuals who do not currently need to be in hospital. Generally this would be to provide long-term antipsychotic medication to individuals with established patterns of dangerous relapses and disinclination to comply with treatment.

Scope and statistical sources

The data in this paper are drawn from a number of sources. A wide range of statistical information has been collected about both the provision and activity of all NHS hospital beds since the inception of the NHS. In the middle 1980s a number of sources were added which reported the activity of health service staff working in community settings. During the late 1990s new data collections, making use of the wider availability of information technology, were introduced to monitor a range of government initiatives. These document social care provided to individuals, services provided in local areas (service mapping), and mental health care for individuals [Glover et al. 2003; 14, 15]. Service mapping specifically attempts to cover all facilities for people with mental health problems irrespective of what type of agency provides them. This gives a very broad and detailed view of services on which much of this paper is based. However it is not yet available for older peoples mental health care.

Finally, before presenting the data it is important to note likely deficiencies in its quality. An anonymous reviewer of this paper, argued, clearly from an extensive and detailed knowledge of contemporary NHS mental health care, that many elements of the data presented were not sufficiently reliable to be reported. While I share his concerns about errors and omissions in the data, I disagree with his conclusion. They provide the best evidence currently available about the wide diversity in style and amount of services available around the country. This provides an important counter to the simplistic notion that the pattern of services for a country of around 50 million people, with areas varying widely in wealth, urbanicity, industrialisation, ethnicity and basic demography could be encapsulated in a single model. The system they seek to describe is both diverse and changing.

Routine, comprehensive, national reporting can never include the types of data quality control used in small-scale local research. Completing statistical returns is a small, and uninteresting chore for a wide range of staff whose primary concern is for clinical care. At present in England this process is not directly linked to funding; as long as statistical returns are completed, the actual numbers have few immediate consequences and are not checked in detail. In this context its accuracy is obviously open to question and it is important to use them, if at all, with close attention to their flaws.

Various strategies exist. Occasionally different sources cover the same areas allowing cross checking, (though sometimes apparently different data collections may in reality have the same source data). Sometimes, for example with acute in-patient beds, gaps are evidently implausible and can be taken to indicate missing data. Scrutiny of local values often reveals individual implausible reports; either these may be excluded, or representative figures, such as mid ranging percentiles, can be taken as national proxies. In some cases it is possible to go back to those supplying the data to seek clarification. All these approaches have been used. To avoid suggesting spurious precision, most figures quoted are also rounded. With these provisos, I believe the data offer important insights into the current pattern of mental health care in England.

Unless otherwise specified the data is related to the year April 2003 to March 2004.

Structure (input) data

Mental hospital and other residential provision

Table 1 provides an overview of the range of hospitals and other facilities in which individuals can become resident for short or long periods. It is divided into four sections.

Table 1 Overview of inpatient accommodation by care type and broad age group, England
Full size table

In-patient beds for acute illness

NHS beds form the main provision for in-patient treatment of acute illness. Department of Health estimates of their numbers are collated from annual trust reports of the average numbers of beds available daily [16]. These are reported separately for the age groups described above, with the overall provision for adults and elderly people being about 4.3 per 10,000 of the total population. Provision for people aged over 65 is roughly double that for working age adults in relation to their population numbers.

For working age adults, separate estimates are available from service mapping [17]. These are based on counts of beds available on a specified day. Roughly 10% of areas reported no provision in this category. Since all require it, this must be regarded as missing data, suggesting that the service-mapping estimate should be revised up to around 13,000, similar to the Department of Health estimate. Discrepancies may also have arisen from the detail of the question (point estimate compared with daily average) or from differences in the classification of beds.

Neither service mapping nor routine Department of Health statistics asked about the type of location of these beds in 2003. However the following year this question was introduced into service mapping for working age adult provision. It was disappointingly incompletely answered, with just under 60% of beds and units having their location specified. Of these, about a quarter were reported to be in general hospitals, 70% in specialist psychiatric hospitals and the remaining 5% in community-based units.

Beds for longer stay care

The division between ‘health’ and ‘social’ care complicates the counting of the number of longer stay beds for both working age and older adults. During the late 1980s and early 1990s there was a rapid reduction in health service provision for people needing long-stay care in both age groups. Government funding incentives designed to encourage privatisation of this type of care initiated this. Initially these provided central government funding for people in independent sector accommodation. In the early 1990s, in an attempt to contain costs, funding responsibility was moved back to local government. At the same time a government initiative was launched to re-establish long-term NHS accommodation for the small number of long-term chronically ill adults of working age demonstrably in need of continuing nursing care. However, by this time large numbers of patients had been transferred into independent sector care homes, often with little or no continuing supervision from the psychiatric teams of their district of origin.

Thus in addition to identifying beds provided by the health service, it is important in this context to try to cover this wide range of services. Table 1 shows the Department of Health estimate of 5,520 beds provided by the NHS. This roughly equates to the service mapping estimate of residential rehabilitation and 24-h nurse staffed beds. Once again the service mapping estimate of these facilities is a little lower at 4,200. In addition to this service mapping gives a picture of the wider range of non-NHS provision. Ten different types of supported residential placements are identified in this source; these are set out in Table 2. Those which are staffed 24 h a day have been categorised here as high support, others as low support.

Table 2 Long-term residential beds for working age adults reported in the service mapping, by type of provision
Full size table

From the history sketched out above, it seems likely that in a number of areas, many patients will have effectively become lost to the service, the only continuing involvement of statutory authorities being the payment of bills. It is therefore likely that these data are less complete than data concerning facilities run by the NHS. However detailed inspection, particularly of the figures for hostel provision, suggests that a few of the services reported, while catering for many people with significant mental health problems, do not do so exclusively. This figure probably therefore also includes some overestimation.

Secure provision

Longer term residential provision for individuals who need secure containment is a much more specialised area of work. Historically in England it has been divided into the high secure services, comprising three major institutions outside mainstream NHS management arrangements, and low to medium secure care provided by ordinary local and regional NHS organisations. In recent years this has changed in two ways. Major and intractable problems with the running of the high secure establishments have led to the transfer of their management to local mental health provider organisations. Detailed assessment of the patients in these institutions has also led to the transfer of many for whom high security was no longer considered relevant. At the same time, the speed of increase in the requirement for low to medium secure accommodation has outstripped NHS capacity for growth. This has led to expansion in the proportion provided by the independent sector.

Unfortunately the available statistical sources do not identify secure placements in independent sector beds. However these probably account for a significant proportion of the 5,790 independent sector beds for working age adults as shown in Table 1.

Independent sector beds

It is less easy to be clear about the exact profile of provision in independent sector beds. The only publicly available source of information on these comes from the registration process, which covers all independent sector hospitals. This identifies beds by the broad client group (‘mental illness’ is the only relevant category here) and the age group of the patients. In addition to the secure provision described above, some undertake other highly specialised types of care unlikely to be provided in local general psychiatry services, others provide routine general psychiatric in-patient facilities either for self- or insurance-funded patients or for NHS overspill. Independent sector provision for older people is probably mainly nursing home accommodation for people with dementia.

Other acute inpatient facilities

In recent years there has been considerable awareness of the undesirability of the environment commonly found in NHS inpatient psychiatric units. This has enhanced the drive to find alternatives, particularly for women patients. In addition to the types of facility described above, a number of innovative projects have developed around the country. Service mapping identified 43 for which bed numbers were provided and which appeared properly classified [17]. Between them, these provided just over 270 beds, an average of 6.3 per project. Various descriptions including crisis houses, safe houses, or alternatives to admission were given for these facilities. In some cases it was specified that they were available only to current long-term clients, in others that they were available to anyone for whom admission might be considered appropriate. A few indicated that they were specifically for individuals with mental health problems whose housing situation had become impossible in the short term.

Ambulatory care facilities

Policy and organisation

Through most of the period since the late 1940s, mental health care in the NHS was structured similarly to specialist medical care. Patients would be referred by a general practitioner to a consultant psychiatrist and seen first in an outpatient clinic, or very occasionally on a home visit. Urgent admissions to hospital, or access to other more specialised resources, (for example psychological treatments) would also be arranged through contacts between a general practitioner and a hospital doctor.

Since the early 1990s, community mental health teams (CMHTs) have increasingly taken over the roles of accepting referrals and providing both short- and longer-term care in many places. The mental health care blueprint set out in the policy guidance described above, both formalised and extended this [1]. Care is to become focused on a network of interrelated types of community-based teams. CMHTs comprising groups of nurses of varying levels of seniority, and one or more doctors, social workers, psychologists, occupational therapists and possibly others such as carer support workers or counsellors, are to be the central element. These teams will accept referrals from general practitioners, social workers, other professionals or individuals seeking help for themselves. New patients will be allocated to an appropriate team member for assessment and treatment. If necessary, further discussion at a team review will lead to involvement of other professionals. Most of the work of these teams is made up of individual ambulatory patient contacts, the patients usually visiting the professional at the team base. Occasionally patients would be visited in their own home for specific purposes. Some group work would also commonly occur at the team base.

Three further types of team are prescribed as core components of local mental health service. Crisis resolution teams (or acute home treatment teams as they would more commonly be described outside the UK) are intended to manage situations where individuals would formerly be likely to have been hospitalised. They are intended to minimize admission through intensive home based support and possibly the assistance of day hospitals and other community facilities. Assertive outreach teams are intended to provide long-term care for a small number of particularly difficult patients whose problems are compounded by either personal disorganisation or a disinclination to accept treatment. Comprising mainly nurses, these work with a very high staff to patient ratio. Patients are visited frequently and staff work flexibly, often assisting with a wide range of health and social tasks. The aim is to develop greater social stability and treatment compliance. Psychosis early intervention teams were intended to provide early diagnosis and treatment as well as supportive assistance to young people making the necessary life adjustments associated with the onset of this type of illness.

The task of mapping all these developments is complex. Predictably, they have developed more quickly and completely in some places than others. There are differences of opinion about the appropriateness of the blueprint in some (particularly rural) contexts as well as variations in the available funding support. In a national mapping exercise, definitions may be interpreted differently in different places and local services always tries to ensure that local anomalous or possibly innovative arrangements are not missed. Hence services may be reported in categories not wholly appropriate. In preparing the data set out below, service mapping information has been newly scrutinized for these effects, hence in some cases numbers of teams reported are less than previously published [17].

Outpatient clinics

A broad view of outpatient clinic provision is provided by Department of Health hospital activity statistics [16]. These show that nationally 540 people per 100,000 had a first outpatient attendance within the year and that, on average, there were 7.2 attendances for each first attendance. First attendance rates increased with age, (children and adolescents 260, working age adults 570, older adults 810). Around 1% of attendances were identified as specialist forensic consultations and 2.5% as psychotherapy. Forensic clients and older adults attended less frequently (5.8 and 4.6 attendances per first attendance respectively) while psychotherapy patients attended more frequently (13.0 attendances per first attendance).

The richest detail about this type of provision is the service mapping data, available only for working age adult services [17]. Around 120 out of 174 (69%) LITs reported some outpatient clinic provision. Between them, they reported 350 clinics, 33 of them (9.4%) with specialist functions, the remainder general psychiatric outpatients. Specialist clinics reported included seven for people with eating disorders, five rehabilitation clinics for people with long-term illness and four neuropsychiatry services. Single instances of clinics for affective disorders, mentally disordered offenders, young adults, new mothers, neuroses, social disorders and gender dysphoria were reported. All these would be likely to have wide catchment areas, but evidently in most areas the problems they cover would be treated, if at all, in general clinics.

While this picture appears detailed, it is almost certainly incomplete. Out-patient clinics have historically been a core element in mental healthcare provision and it is thus surprising that so many areas report none. Brief enquiries with some of those reporting none indicate that while in most cases this is missing data, in a few, newer arrangements have led to the abandonment of out-patient clinics.

Around 286 (82%) out patient clinics reported staffing profiles. Around 79% of these were staffed by doctors alone, 9% included psychologists, 17% included nurses. General clinics reported a median of 42 attendances per week (interquartile range 18–81) and 88 referrals (interquartile range 39–209) in the preceding 6 months. Annualised attendance and referral rates per 100,000 population aged 18–64 (in the areas with clinics which reported figures) were 4,790 and 480 respectively. The ratio of attendances to referrals for individual areas had a median value of 10.8 (interquartile range 5.8–24.1). These activity figures are broadly similar to the national figures reported above. This again indicates that a substantial proportion of the areas failing to report this type of provision in service mapping must nevertheless have had it.

Community mental health teams

In the service mapping for 2003, 808 appropriately coded community mental health teams were reported with all but one LIT (173/174) reporting at least one. Detailed comments were available about half (406). These made it clear that by no means all served the broad generic function described above. Around 112 (14%) were described as focusing specifically on severely mentally ill patients. This may reflect a functional separation of roles within areas, local policy to focus limited available resources on severely mentally ill patients, or the narrower historical role of community psychiatric nursing teams. For 693 teams, available evidence suggested a broad generic function. Around 156 LITs (90%) described at least one of these.

Teams reported a median complement of 14.5 whole-time-equivalent (WTE) staff (interquartile range 10.7–19.4), comprising 43% nurses, 13% doctors, and 23% social workers. Around 757 (94%) reported their current caseload. For these teams this represented 25.6 patients per staff member for generic teams and 22 for teams focusing on severely mentally ill patients. In relation to the population covered, the current caseload of generic teams was almost exactly 1% of the population aged 18–64.

Crisis resolution/acute home treatment teams

In the service mapping for 2003, 119 crisis resolution teams were reported, with 90 LITs (52%) reporting at least one. This figure needs to be taken in the context of the rapid implementation of these teams following the publication of the Policy Implementation Guidelines. When the first service mapping was undertaken in 2000, a best estimate is that it included 25 properly classified teams. Corresponding figures for the two subsequent years were 45 and 59, reflecting strong central government direction backed with substantial new funding. For the purposes of monitoring it raises the question of the point in its development at which a team should begin to be counted. (Usually staff are appointed over a period of six months or so, and spend a few more months developing working arrangements and liaising with future referrers before starting operations, often on an initially restricted basis.) Local areas were understandably keen to report the presence of teams, since important national performance indicators were attached to these.

The size of teams probably reflects this. Given their emergency role, it is necessary that they are available at least most of the 24 h of the day, 7 days per week, and that sufficient staff be available to visit patients in pairs. A staff complement below 10 makes this unlikely. Of the 119 teams, 48 had fewer than 10 staff, 29 fewer than 8 and 13 fewer than 5. The median staff size was 11 (interquartile range 8–15.5), 73% of this being nurses, 10% social workers and 4% doctors.

Around 90 of the teams reported caseload and referral figures. The median caseload per staff member was 1.6 (interquartile range 1.1–3.1), appropriate to their sort of work. In relation to the size of population served by LIT areas reporting these teams, the total figures represent 1.9 persons currently on caseload and 31.6 referrals annually per 10,000 population. (These rates are cited per 10,000 as the role these teams undertake should be compared with in-patient beds rather than wider ambulatory services).

Psychosis early intervention teams

The new mental health policy envisaged the setting up of a network of 50 early intervention teams across England to provide early diagnosis, and care for individuals in the prodrome and the first three years of psychotic illness. Service mapping in 2003 showed 35 in place, of which two were still at too early a stage to be operational. The remaining 33 reported a median staff complement of three (interquartile range 1.9–6.6) and a median caseload per staff member of 5.8 (interquartile range 2.0–10.1). This staffing level is too small to undertake the roles adequately and suggests either incomplete reporting or a high proportion of teams at an early stage of development. As implementation of these teams is currently the subject of a prominent government initiative, the latter seems more likely.

Teams caring for long term mentally ill people

The Mental Health Policy Implementation Guidance [1] discussed assertive outreach teams in detail. In addition to these, the ordinary community mental health teams described above looks after many long-term clients. However in some areas other specialist teams are also provided; in mapping work these appear as ‘rehabilitation teams’ or the community mental health teams, described above, which report specializing in this client group. The three are described together here.

It might be logical to expect that these teams would substitute for each other. However the evidence does not suggest such a simple pattern. Out of 174 LITs, 152 (87%) had an assertive outreach team, 56 (32%) a rehabilitation team and 41 (24%) a specialised community mental health team. Around 70 (40%) had an assertive outreach team alone, 45 (26%) assertive outreach and rehabilitation teams, 29 (17%) assertive outreach and specialised community mental health teams and 8 (5%) all three. Around 16 areas (9%) reported none of these.

Table 3 shows a comparison between these three types of teams. It shows that the supposedly specialized community mental health teams are quite different from the other two team types in terms of staff profile and caseload, similar in fact to ordinary community mental health teams. Rehabilitation and assertive outreach teams, on the other hand, appear quite similar to each other. Calculating the prevalence of use of these types of care is complicated by the fact that their local roles are presumably dependent on the local configuration. If it is assumed that all the areas reporting any teams have reported comprehensively, then overall, in those areas 30.4 people per 100k total population (49.2 per 100k aged 18–64) currently use assertive outreach or rehabilitation teams while a further 78.2 (126.7) use specialized community mental health teams.

Table 3 Teams caring for long-term mentally ill adults of working age in the community, staff numbers and profiles and caseload statistics
Full size table

Day hospitals and day centres

Daytime provision is one of the hardest areas to describe nationally. Traditionally in England it has been divided into day hospitals, provided by the NHS and undertaking medical functions, and day centres, providing broadly social care functions, funded, and in some cases provided by local government. In practice the situation is far less clear. Briscoe et al. [18] undertook a national survey of day hospitals, which demonstrated a wide range of roles from those providing supportive facilities for long-term mentally ill to those providing for the treatment of acute illness as an alternative to hospitalisation.

Mapping of services for working age adults in 2003 identified 242 appropriately classified day hospitals, with 102 out of the 174 LITs (59%) reporting at least one. Descriptive comments were available for 43% and these indicated a wide range of functions similar to that described by Briscoe. Of these units, 24% reported focusing on acute illness, 16% on eating disorders or psychotherapy and a further 20% on wider, but still distinctively treatment programs. 40% reported a focus comprising some type of support for long-term mentally ill patients.

Around 707 day centres were described, with 94% of LITs reporting at least one. The majority were reported as being provided by the independent sector, mostly non-profit organisations. Roughly half of the staff of independent sector day centres were volunteers. Day centres generally undertake a range of functions. Around 92% indicated that they provided advice and information, 86% leisure activities and 70% support groups and ‘drop-ins’. Less common were counselling (36%), befriending services (34%), advocacy (34%) and carer support (33%). While, in the majority of cases, these will have been at least partly funded by local government or the NHS, 59% of the day centres were reported as being provided by independent, not-for-profit organizations. Around 16% were reported to be run by the NHS, 17% by local government and the remainder by combinations of agencies. Centres run by independent organisations were more likely to offer drop-in services but less likely to provide counselling and service-user groups. Among centres run by the public agencies, those run by local government were more likely to provide organised education and leisure and service user groups.

No similarly detailed survey is available of day services for older people.

Mental health care staff

The most detailed source of information about staff providing mental health care for working age adults is the mental health service mapping [17]. In addition to the total number in each professional group this also provides an indication of how staff are deployed in each area. Of 248 paid care staff per hundred thousand working age adults, 127 are nurses, 18 doctors, 18 social workers, 11, occupational therapists, and 7 clinical psychologists.

These data should be looked at with some caution as a result of two aspects of the way the mapping work is undertaken. First it is not considered feasible to collect staffing levels for long-term residential placements outside the NHS. Some of these will be nurses, but the majority residential care workers with lower qualifications. Second, there are clearly some problems with the accuracy of staffing data. Teams are asked to report whole time equivalent staff numbers. In practice, it is evident that in some cases they report ‘headcount’ figures. This is most obvious for voluntary sector facilities and applies particularly to volunteers and counsellors. A similar problem of overestimation seems to have occurred in a few NHS services where staff, such as senior doctors, work in several different locations.

It is hard to check these data as relatively few corroborative sources are available. One exception is acute in-patient beds where a survey was undertaken in 2004, which reported staff in relation to bed numbers [19]. Findings were available for qualified nurses, nursing assistants, occupational therapists and assistant occupational therapists. All were closely similar.

Staffing rates in relation to population numbers vary considerably around the country. London has 1.3 times the number of nurses and social nurses, 1.5 times the number of doctors and 1.6 times the number of clinical psychologists seen in the rest of the country.

The figures also provide a correction to the common perspective that mental health care in England has become largely non-institutional. It is striking to see that 65% of the nursing staff reported (who make up just over half the paid care staff) is still working in inpatient or residential settings. Exploration of the pattern of staff deployment around the country suggests considerable variation. These figures should again be treated with considerable caution because of questions about their accuracy. Probably they are most accurate for nurses who are usually employed to work in only one setting. Here the proportion assigned to acute inpatient care has a median of 44% and an interquartile range from 35% to 52%. Corresponding figures for general ambulatory settings are 20% (15%–25%) and for long-term community-based care 4% (2%–6%).

Process data

A number of examples of ‘process’ data have been quoted above as, in some cases, these give the only really quantitative view of the facilities available. However there are three other, major sources of mental health care process data, which provide very detailed evidence of patterns of care.

Hospital episode statistics (HES)

The oldest and most reliable process dataset available for England describes admissions to NHS hospitals. This system has been in place since 1949, although with two major structural reorganisations each including roughly three years loss of data (in 1961 and 1987). Readily available data are categorised by broad age group, consultant specialty and diagnosis [20]. For the present purpose the first two types of analysis are the most useful.

As a rough check of completeness it is appropriate to compare the total occupied bed days reported with the bed availability figures described above. Data for 2003 shows average numbers for daily occupied beds at 15,645 in the three working age adult specialties (mental illness, forensic and psychotherapy) and 7,061 for psychiatry of old age, a total of 22,706. Set alongside the figures in Table 1 these suggest occupancy rates of 73% for working age adults and 56% for older people. These figures are much lower than contemporary data reported from a more detailed survey by Garcia et al. [19]. These authors found a national figure for working age adult in-patient services of 100% bed occupancy with variations between regions from 109% to 91%. Hence there must be a difference in the scope of the two data sources. Probably this is most marked in the figure for older people where 2/5 of the beds are described as providing for long-term patients where percentage occupancy rates would normally be expected to be in the high 90s.

The principal units reported in these data are ‘finished consultant episodes’; that is to say periods during which an individual is looked after by a consultant during a single hospital stay. This figure is slightly higher than the admission rate since some individuals will be transferred between consultants during a single hospital stay. Overall, the national rate for finished consultant episodes is 371 per hundred thousand population per year. (The admission rate, 340 is about 10% lower). The rate for men (380) is just under 5% higher than that for women (363). It shows a biphasic relation to age (age 15–59: 412, 60–74:381, 75 and over: 895). Durations of consultant episodes vary considerably between the subspecialties. The median stay for adult mental illness is 18 days, psychotherapy 55, forensic psychiatry 160 and old age psychiatry 30 days.

In estimating the extent to which admission rates vary around the country, it is necessary to make allowance for evident imperfections in data from individual hospitals in single years. These are common. A simple approach is to exclude from calculation areas where the admission rate differs from the previous year by more than some chosen threshold of plausibility—15% was chosen for the present study, thus excluding 37.5% of local areas. Among the areas where data looked plausible, the admission rate for working age adults showed an interquartile range from 16.9% above to 18.2% below the median.

A separate statistical collection indicates that 88 people per 100,000 population were detained in NHS hospitals under legal compulsion in 2003. The frequency of this varies considerably around the country with the figure for London being 140, while only three other regions exceed 80 [21].

Social services activity statistics

Legislation in the early 1990s changed the role of local government social services departments from one of providing services to one primarily of assessing the service’s individuals need and assisting in their procurement and funding. Some local government-run residential care homes, day centres and other support facilities still exist, mostly predating this change, but the expectation is that a substantial proportion of these types of care are provided by independent sector organisations.

To monitor this new role, a new set of statistics was introduced in April 2000 [10]. Care packages are broadly divisible into residential or community-based. Nationally, 480 people per hundred thousand population received some type of care. Around 399 received community-based services, 81 residential care in independent sector homes, 13 in local authority residential homes and 42 in nursing homes, (some individuals are counted in more than one category as a result of having used more than one type of care during the year). Community-based care is roughly twice as common for people over 65 as for working age adults, residential care nearly 14 times.

The various types of community-based care are documented in some detail. Table 4 sets out an overview, showing separately individuals for whom the principal mental illness is dementia. Much the commonest type of intervention is ‘professional support’. This is defined as the continuing involvement of a professional, usually a social worker, beyond the process of needs assessment and care procurement. The next commonest is day care (which would usually take place at a day centre), and home care which would involve a non-professional social care worker visiting the client at their home to assist with basic domestic tasks.

Table 4 People receiving social care, excluding long-term residential care, as a result of mental illness: numbers (rounded to nearest 100), rates per 100,000 population and proportions (by column) receiving various types of support
Full size table

The rates of care provision vary considerably between local authorities. For any type of care, the median figure is 486 per 100,000, interquartile range 362–620, for community-based services 374 (273–484). The range for all types of residential care combined cannot be exactly calculated because of double counting, however it is roughly from 35% below to 50% above the median.

The mental health minimum data set

The most recent addition to the range of statistics for mental health care in England is the mental health minimum data set (MHMDS) [15]. This was developed in the late 1990s and finally introduced fully in April 2003 [22]. It records periods of care for adults of all ages with specialist mental health services, showing how the care contacts they receive are related in time. Data from this source are only just beginning to become available and, inevitably, the early concerns are with their completeness and accuracy. At the time of writing only data for the 2003/2004 data were available in detailed form. No published analyses were available. I undertook the analyses presented here, using data made available by the national quality inspectorate, the Healthcare Commission.

Examination of the profiles of figures for the 303 health care commissioning areas suggest that between a quarter and a third produced seriously incomplete figures in this, the first year of data returns. Table 5 shows appropriately selected points on the percentile distributions for health care commissioning areas. Overall, these show median figures of 2.3% of the population in touch with specialist mental health care in that period. Of these 60% attended outpatients, 37% saw a community psychiatric nurse and 14% spend at least 1 day in hospital.

Table 5 Population-based rates of receiving any, and specifically in-patient mental health care (rounded to nearest 10), and percentages of patients receiving specific types of care (by columns, rounded to integers)
Full size table

The table shows separate figures for working age and older adults. The conventional pattern of higher usage rates for older people is seen, for overall care rates by about 60% and for inpatient care by about 80%. CPN and day hospital care are commoner components of care packages for older people, while clinical psychology and outpatient attendances are less common. These figures should be read with caution since it is likely that data about inpatient care are the most completely reported. However they seem reasonably compatible with the inpatient admission rate figures reported above, since in most cases individuals having two or more admissions within a year will not be multiply counted in this table.

Discussion

This paper has attempted to survey the overall pattern of mental health care for working age and older adults in England. It is a difficult undertaking for two reasons. First, data sources are almost all incomplete in their scope or coverage or variable in quality. Second, the actual design of the system varies considerably between places. Funding levels in relation to need are substantially greater in some areas than others. An inevitable consequence of this is that some services are able to address a wider spectrum of mental health problems.

The paper has considered the issue of data quality in some detail in many places. The caveat of an anonymous reviewer was reported in the introductory section and readers should be aware of these concerns in reading this paper. It is the authors view that routinely collected statistical data, which cannot be examined in this type of detail, is more worrying than those in which the scale and nature of the defects can be inspected readily.

For working age adults, data from service mapping in earlier years indicate that the pattern of services is not static. Mental health care in England is currently undergoing significant change as a result of a major government initiative. Some new types of service appear to be being introduced quickly. Others are being scaled down, or possibly redesignated. The task of identifying whether new style services will prove more effective will require other types of analysis. However the types of data presented at least allow us some evidence about the nature and scale of the changes.