Original articles Severe persistent mental illness

Many patients with persistent mental illnesses enjoy a better life in a community setting than would be possible in a long stay mental hospital. Furthermore, the available evidence indicates that most such patients get better while living in the community. Unfortunately, community care has not served all patients well. Much of the difficulty can be attributed to lack of resources. However, there is also a tend ency by planners to under-estimate the severity of patients' disabilities. A realistic appraisal demands a detailed examination of the problems of patients whose needs have not been met by community care. One important issue is that of patients who fall through the net of community care and another is that of patients who have not but nonetheless have not survived in the community. This paper addresses the question of the needs of this latter group.

Many patients with persistent mental illnesses enjoy a better life in a community setting than would be possible in a long stay mental hospital. Furthermore, the available evidence indicates that most such patients get better while living in the community. Unfortunately, community care has not served all patients well. Much of the difficulty can be attributed to lack of resources. However, there is also a tend ency by planners to under-estimate the severity of patients' disabilities. A realistic appraisal demands a detailed examination of the problems of patients whose needs have not been met by community care. One important issue is that of patients who fall through the net of community care and another is that of patients who have not but nonetheless have not survived in the community. This paper addresses the question of the needs of this latter group.

The setting
From 1982 to 1987, St Bernard's Wing of Baling hospital provided in-patient services for patients with intractable mental illness from a catchment area of population 0.9 million in west London, when there was an active policy of striving to prevent long stay in hospital. When patients admitted to the acute ward could not be discharged to their former address, placement in supported accommodation such as group homes or social service hostels was sought. An extensive system of community resources, including group homes run by the Health Service, Social Services and voluntary bodies, had been established. Nonetheless, in 1988 there were 42 in-patients aged less than 65 at admission, admitted from 1982 to 1987,who had remained as in-patients for more than one year. This is a group of patients representative of those remaining in contact with services whose needs could not be met in the community.
The sample does not include patients detained in the Regional Secure Unit serving North-West Thames Region, and consequently forensic cases are under-represented. Nonetheless, five of the 42 patients were detained under Section 37 of the Mental Health Act as a result of dangerous criminal behaviour, and in many other instances, dangerous behaviour was a factor contributing to the index admission to hospital.

Assessment of the problems
These 42 patients were assessed as a part of a survey of all the long-stay patients (excluding those on psychogeriatric wards) in 1988. Social function was assessed using the Social Functioning Questionnaire (Clifford, 1987a). The SFQ rates skill in 41 of the activities of daily life on a scale from 1 (major difficulty/high level of dependency for that activity) to 4 (satisfactory performance without assistance). A social function index, a measure of readiness for dis charge from hospital, can be derived by determining the average on ten cardinal items (including per sonal appearance, simple domestic skills, shopping, getting-up in the morning). A score greater than 3.0 suggests social function adequate for discharge. Problem behaviour was assessed using the Problems Questionnaire (PQ; Clifford, I987b) which rates the presence of 37 types of problem behaviour in the preceding month. Each item was rated 0 or 1 indi cating absence or definite presence of the relevant problem. The total number of problems per patient, and the occurrence of violence in the preceding month, were examined. The occurrence of violence was defined as definite occurrence of at least one of the following items of behaviour from the aggressive or dangerous behaviour section of the PQ: item 20, tendency to violence; item 21, self-harm (overdoses, self-mutilation); item 22, sexual assault; and item 23, arson. Behaviours such as verbal abusiveness (item 8) and threatening behaviour (item 9) were not counted as violence.
A comparison between the 42 new long-stay patients and the 114 old long-stay patients is pre sented in Table I. In both groups, the majority suffered an illness that began early in adult life, and had received psychiatric treatment for many years before the current admission. The new long-stay patients tended to have had a longer period of prior treatment, often entailing multiple admissions with intervening periods of unsettled life in the com munity. The eight patients in the new long-stay group with diagnoses other than schizophrenia comprised three with organic psychosis, three with personality disorder and two with manic-depressive psychosis.
While both groups of patients exhibited many problems, the mean number per patient in the new long-stay group was significantly greater than that in the old long-stay group. In particular, violence was significantly commoner in the new long-stay group. More than half the group had been violent. On the other hand, the new long-stay group exhibited a significantly higher level of social skills, though the group mean for the 10 index items was below the criterion of 3.0, indicating substantial impairment.
Three new long-stay cases satisfied criteria for discharge from hospital according to both the SFQ (mean index item score >3.0) and the PQ (number of problems <5). One was a former patient of Broadmoor Special Hospital restricted under section 41 of the Mental Health Act; one had attempted arson in the preceding month; the third was making good progress in a rehabilitation programme and was subsequently discharged to a residential care home.

Outcome
In the following 3.5 years, 10 of the 42 patients were successfully discharged from hospital care, all to settings providing 24-hour staff supervision. Often the after-care plan demanded extensive support by professional staff. For example, a young woman who had twice resorted to arson when she was lonely and distressed prior to admission was discharged to a staffed hostel and subsequently to a supervised flat, with regular contact with community psychiatric nurse, social worker, psychologist, and psychiatrist. The social worker acted as key-worker responsible for coordinating the care.
Successful discharge was associated with absence of violence in the assessment month. Two of the ten patients discharged had exhibited violence, com pared with 20 of the 32 remaining cases (/><0.02). Three of the 42 died. The number remaining under hospital care (including those transferred to an offsite hospital hostel opened in 1989) decreased at the rate of 10.6% per year. The rate of discharge in this group of severely disturbed patients is similar to that reported by Garety et al (1988) for clients served by a hospital hostel.

Predicted bed requirements
On the basis of the observed accumulation of cases over six years, and the rate of discharge or death, it is possible to estimate the annual rate of admission of new long-stay cases, and the number of beds required for such cases in a steady state in which admission is balanced by discharge and death. The estimated rate of admission is 1.05cases per 100,000 population per year, and the number of beds required in the steady state is 9.9 beds per 100,000. The rate of accumu lation is slow. If a district established a new service without having to provide for previously accumu lated long-stay cases, the number of beds required after ten years would be 6.3 per 100,000. These esti mates do not include requirements of those aged over 65 at admission, nor those of forensic cases cared for in a Regional Secure Unit. And because the sample did not include cases who had fallen through the net of mental health service care, the calculations of bed requirements are approximate.

Type of accommodation required
The number and type of problems exhibited by the new long-stay patients require a range of care extending from a hospital hostel, staffed by trained psychiatric nursing staff (Goldberg et al, 1985;Garety et al, 1988) up to a level of supervision and security approaching that provided by a Regional Secure Unit. Prior to the opening of a six-bedded offsite hospital hostel in 1989,each patient was assessed to establish potential suitability for transfer. Fifteen of the 42 cases exhibited problems which could not have been managed safely in an off-site hospital hostel. These included violence with minimal ability to collaborate in domestic activities; tendency to abscond and behave dangerously; and repeated arson.
This small group of severely disruptive patients require accommodation in a unit that offers sus tained care with adequate supervision. Such a Sus tained Care Unit needs to be located on a site which can contain violence. It needs trained staff; access to locked facilities for use during episodes of disturbed behaviour; and a setting sufficiently spacious to allow the residents freedom of movement. To avoid under mining self-responsibility, it is necessary to design the accommodation in a way that fosters autonomy, and to provide occupational training facilities. It is unlikely that the required level of supervision and facilities could be provided economically in a unit of less than 40 beds, though these might best be dis tributed in hostel-type buildings on a single site. That 15 cases in our sample required this level of care indicates that approximately 3.6 such beds are required per 100,000 population. Hence it would not be feasible to develop a Sustained Care Unit to serve a population of less than 1million.
Despite attempts to prevent the accumulation of long-stay patients, a small number present problems demanding a level of care hitherto only available in a hospital setting. These new long-stay patients differ from the residual old long-stay patients in that 745 they exhibit more violent behaviour. However, even among this group, a substantial proportion can be discharged to less intensely supervised accommo dation over several years. Thus, accumulation of new cases will eventually be balanced by discharge. Although the estimates are approximate, the findings suggest that, once a steady state has been achieved, a health district serving an urban population of 300,000 would require approximately 20 beds located in hospital hostels, and access to 10 beds in a regional Sustained Care Unit. Because of the slow rate at which cases of severe, persistent mental illness accumulate, a District Health Authority can over look the needs of these patients in the short term. However, continued failure to take account of the high level of care demanded by this small minority of patients will lead to continued failure to plan for one of the most disadvantaged groups in our society.