The provision of residential care homes for the severely mentally ill in the community is an important issue for psychiatry, and mental health teams regularly search for that will maximise a patient's skills and quality of life. Closure of institutions and current Government policies on community care have led to a rapid expansion of residential care homes in the private sector. Thirty-two per cent of residential places for the mentally ill are now managed by the private sector, and Lelliott et al (Reference Lelliott, Audini and Knapp1996) found an almost threefold variation between districts in the total number of residential places available per unit of populations. Generous provision of residential care can be advantageous, as it can provide much needed transitional or permanent placement for new long-stay patients. However, it can also be disadvantageous, especially if the facilities fail to provide adequate rehabilitative care, which may result in the development of ‘new institutions’ in the community akin to the asylums of the past. Furthermore, a localised excessive provision of such care homes can transfer morbidity into the catchment area in which they are located.

NHS reforms have also focused on quality management, providing the highest quality of service at the lowest possible price. The pace of private residential care expansion may suggest that such provision is cost-effective and benefits patients. However, there is little or no evidence to support or refute this. Thus, accurate information on costs, service provision and outcomes of placements is required so that informed, strategic development can take place.

The aims of this study were to describe the characteristics and referral agents of residents in seven midstaffed care homes (as defined by Reference Lelliott, Audini and KnappLelliott et al, 1996), to identify the mental health services currently used by the residents, including admissions to hospital, and to evaluated the cost and cost variations of their psychiatric health care.

The Study

The population of the study live in care homes in the borough of Lambeth, within a South London catchment area in what was then the Lambeth Health Care Trust. This now forms part of the South London and Maudsley NHS Trust. Of the boroughs in England and Wales, Lambeth has the sixth poorest Jarman index (Reference Jarman, Hirsch and WhiteJarman et al, 1992). The care homes were selected because they are owned by the private sector and provide for people with mental illness. Comprehensive demographic details for residents, including address prior to taking up residence at the care home, current mental health care provision, structured daily activity, care programme approach (CPA), supervision register registration, diagnosis, social services provider and source of referral were compiled via questionnaires completed by the home staff and augmented by hospital case notes, the CPA and supervision register. The medical records database was searched for admissions details for the entire duration of the individuals' residence in the trust catchment area. The time spent on client-centred activity and face to face contacts over a 3-month retrospective period was documented. The community service budget figures were used to calculate the facility costs and salaries paid to non-health care staff. The cost of care provided over a 1-year period was calculated using Netten and Beecham's (Reference Netten and Beecham1993) methodology. Costs of care provided by the psychiatric system, not total costs to society, were assessed.

Findings

In total, 83 residents managed by the trust were living in the seven homes, 61 (74%) of whom were male and 22 (26%) female. A further 18 residents were not in the care of the local catchment area mental health services: these services were provided by the health care trust that had placed them in the care home or they were managed by primary care services alone. These 18 were excluded from our study. The mean age of the group was 42.5 years (s.d.=12.8), with men being younger on average than women (40.9 and 46.9 years, respectively). The mean duration of residence was 4.33 years. There were three ethnic groups within the population of residents, with 48 (57.8%) White, 24 (28.6%) African-Caribbean and 11 (13.3%) Asian. Seventy-two (87%) of the residents were single and 9 (11%) divorced. None in the sample were in paid employment. The diagnoses recorded in the 83 patients' case records were as follows: 47 (56.6%) had schizophrenia; 5 (6%) had schizophrenia and a learning disability; 8 (9.6%) had schizophrenia and misused psychoactive substances; 3 (3.6%) had schizophrenia, a learning disability or cognitive impairment and misused psychoactive substances; 1 (1.2%) had schizophrenia and a personality disorder; 3 (3.6%) had cognitive impairment; 5 (6%) had a learning disability; 3 (3.6%) had a personality disorder; 3 (3.6%) had an affective disorder; 2 (2.4%) had an affective disorder and misused psychoactive substances; 2 (2.4%) had a learning disability or cognitive impairment and misused psychoactive substances; and 1 (1.2%) had no diagnosed psychiatric disorder. In addition, 6 (7%) had diabetes mellitus and 6 (7%) had epilepsy.

All the residents in the study were in contact with psychiatric services. The variety of services provided are included in Table 1. Twenty-four (29%) residents were receiving care from two community care professionals (e.g. during medical out-patient clinic appointments and community psychiatric nurse visits), and the mean number of contacts with community care services of any type was 4.13 per patient per month. Data were available on the daily activity of 36 patients, 33% of whom were attending a structured activity with a rehabilitative or therapeutic function. One resident saw an occupational therapist regularly.

Table 1. Cost residential of community care over 12 months

Mental health worker	Mean time (hour/resident)	Average number of residents per worker	Total cost (£)	Mean cost per contact (£)
Case manager
Grade F	126	3	1494.74	15.57
Grade G	66.42	1.7	5210.54	20.73
Grade H	56	1	885.89	31.64
Outreach nurse1
Grade F	50.34	4	1194.37	13.24
Grade G	96	3	1284.75	13.38
Consultant's clinic	2.25	2	67.01	8.3
Senior house officer's clinic	7.33	4	97.04	4.41
Staff grade psychiatrist	22.64	14	789.36	8.85
Senior registrar's domiciliary visit	264	34	4385.17	8.30
CAPS group	16	4	448.37	9.34
Total cost to LHC Trust			15979.16

Thirty-one (37%) patients were residents in the trust catchment area before moving into the care homes, and 39 (47%) had moved into the area in order to take up the residential placement. The proportions of residents from outside the trust area placed in the seven care homes were 100%, 91%, 83%, 60%, 43%, 39% and 25%, respectively. A previous address could not be ascertained for 13 (15.7%) of the patients. Nine social service boroughs were involved in providing for the residents, with the local borough responsible for 49 (59%). Seven (8.4%) were funded from social security benefits alone. (Data were missing on 3 (3.6%).)

The average duration of residence was 4.33 years (minimum 1 year, maximum 15 years), and an average of 12% of that time was spent in hospital. However, only 27 (32.5%) of the 83 subjects had been admitted to hospital since taking up residence, and the mean number of beddays in hospital was 183 (maximum 895, minimum 2). The patients admitted differed from those not admitted in that the mean duration of stay in the care home was 3.16 years (s.d. 2.09) v. 4.86 years (s.d. 3.62). They were also younger, with a mean age of 38.92 years (s.d. 11.39) v. 44.16 years (s.d. 13.16). There was also ethnic variation between the two groups, with African-Caribbean patients forming a larger proportion of those admitted v. those not admitted (42% v. 17%).

To cost hospital in-patient services, bed usage over a period of 1 year (1996) was examined. The monthly rehospitalisation rate during that year varied between 1% and 4%. A total of 840 bed-days were used by 12 of the residents, two of whom were admitted twice. The average number of bed-days used per hospitalised resident was 70 (minimum 5, maximum 218). A week in hospital has been costed at £650 per person (Reference Wilkinson, Piccinelli and FalloonWilkinson et al, 1995), giving a cost of £78 000 for in-patient care of the residents that year.

The total cost of community care provided by the trust to the residents during one year (1996) was £15979, giving an average cost for each resident of £192.52. This does not include the cost of social care, funding of the placement, voluntary sector rehabilitation, medication and primary care. The variation between disciplines in total cost of community care provision, as well as cost per patient contact, is shown in Table 1.

Discussion

This study is limited in that it accounts for only a proportion of the privately managed care homes in a catchment area. Furthermore, there are no data on drug costs and there is no outcome measure against which to explore cost effectiveness. Disease measurement would also be desirable, to ensure comparability. However, the method is easily repeatable and some interesting findings have emerged.

The shift in provider agency for the long-term mentally ill has seen a rapid expansion of private sector residential provision, particularly in this catchment area. Two reasons for this may be that property has been relatively cheap in the area, and that Streatham is surrounded by mental health trusts, all of which require residential care for a regular number of patients who have benefited as much as possible from acute care, but remain too disabled to live independently. These patients form the growing body of ‘new long-stay’ patients and the team for the Assessment of Psychiatric Services Study suggests that the number of new long-stay patients increases each year (Reference Knapp, Beecham and AndersonKnapp et al, 1990). This would suggest that care provision by the private sector is likely to grow as direct provision by the NHS and social services reduces.

The residents are predominantly young, single, unemployed men. Schizophrenia and its spectrum make up 78% of the total diagnoses. There is an increased incidence of diabetes, epilepsy, learning disability and multiple diagnoses among the residents, suggesting that this population has high care needs and that they are appropriately placed in 24-hour supported accommodation. In the past, the majority of these residents would probably have been housed in the Victorian long-stay institutions, and privately managed care homes are filling some of the void left by their closure.

Two-thirds of the residents had not been admitted to hospital since taking up residence, and all received care from the community psychiatric services. This would seem to suggest that many residents were coping well and costing the mental health service relatively little. However, admissions when they did occur tended to be lengthy, as demonstrated by the mean duration of hospital stay.

The calculation of the cost of community care does not take into account the total cost to society, which would include benefit payments, social service contribution to placement cost, primary care provision, voluntary sector rehabilitation and so on. The cost of placements in the care homes varied for each resident, depending on their level of need and disability: the average cost was between £300 and £412 per week.

The cost variations between disciplines are interesting and mainly accounted for by the different styles of working. Case managers tend to work with patients who are clinically mentally unstable and on average tend to have a higher frequency and duration of contact. The out-patient clinic model and in-house reviews by the visiting specialist registrar were by far the cheapest options. However, without measures of effectiveness it is difficult to attach further significance to these findings.

Service provision was tailored to maintain patients in the community, with the expectation that their rehabilitation needs were to be met by the care homes. Of the patients on whom daily activities data were available, only one-third were engaged in regular structured activity outside the home. The care homes differed in the activities they provided and organised internally, with participation in gardening and cooking available in two of the homes. Activities outside the home included attendance at voluntary sector projects, aimed at rehabilitation. No symptom measure was used in this study, but the management of negative symptoms was a clinically significant problem. Negative symptoms make the encouragement of activity difficult and often contrary to what the patient wants. In the 1991 Hampstead schizophrenia survey, Jeffreys et al (Reference Jeffreys, Harvey and McNaught1997) found that lack of daytime activity was an increasing problem in community patients. Therefore, it is worrying that many residents did not participate in structured rehabilitation, given that clinical deterioration in people with chronic schizophrenia is known to be significantly associated with lack of occupation (Reference Wing and BrownWing & Brown, 1970).

Lambeth borough met the costs of the residential care for 59% of the residents. Lambeth was responsible for many of these clients because they had been resident in the borough before the Community Care Act came into force. Social Security benefits alone funded places for 8.4%, suggesting that these residents did not have the benefit of an individual social services review of their progress and the suitability of the placement. The Community Care Act made the social services of individual boroughs responsible for life-long provision of care to vulnerable clients. In our resident group, 30% are provided for by boroughs other than Lambeth. This makes liaison between health and social services departments more difficult, as they are geographically separated and regular contact is hard to coordinate. It also raises issues for patients who are uprooted and may lose their connections with a place, friends and family; they may also lose continuity of care. The homes showed marked variation in where they sourced their residents, with one home recruiting 100% of residents from outside the trust catchment area and another taking 75% from the local district. This might relate to economic variables, such as the relative ability of social services providers to meet varying costs of the homes.

The care homes explored in this study are a private sector initiative and their development in particular locations seems somewhat arbitrary. Certain locations need more local residential care homes: for example, the King's Fund Report (Reference Johnson, Ramsay and ThornicroftJohnson et al, 1997) on London's mental health services indicates a serious lack of placement in Southwark. The Lambeth catchment area is rich in placement provision and so has to provide, from existing services, for a large number of high-need patients from neighbouring boroughs. Given the ever growing need for supported and supervised accommodation, it is time that health and social services and the private sector looked at the development of future care homes in partnership, with the aim of providing residential care to clients local to their existing health and social service teams. This would bring care provision closer to home and could also lead to rationalisation of the workload and resource provision associated with these high-need clients.