Threatened eviction and hospital readmission from community homes

large psychiatric hospital to staffed group homes managed by a voluntary organisation were compared with 11 patients discharged to a hostel ward managed by an NHS Trust.The former were found to have a higher rate of police involvement, readmission and threatened eviction. The study highlights the need for an integrated range of different residential facilities and the necessity for close monitoring of long stay patients discharged to more independent community settings.

The number of hostels and group homes has increased enormously to meet the needs of patients discharged through the closure of the large psychiatric hospitals. Most services employ a range of different rehabilitation placements. It has long been recognised that those with the most severe disabilities are likely to require accommodation with high staffing levels (Mann & Cree, 1976). Wykes & Wing (1982) evaluated a hostel ward for patients with severe disability. This was managed by the Health Authority and staffed with trained nurses but had a domestic atmosphere. The model was found to be success ful and has subsequently been adopted in other areas (Kingdon, 1991).
With the development of community care, homes run by non-statutory organisations and social services are now admitting severely disabled patients, many of whom suffer not only from a primary psychiatric disorder but also from the secondary effects of long-term institutional care. Not all such homes are run by trained staff. Some projects may always have qualified staff on duty while in others, trained staff take on managerial roles. Wagner (1988) recognised the inherent shortcomings of un trained staffing. TAPS (Team for the Assessment of Psychiatric Services) has found that patients are less likely to be readmitted from staffed group homes than more traditional settings such as hostels, council flats or board and lodgings which may not have a specific mental health remit (Thomicroft et cd, 1992;Dayson, 1993).
Our study explores this theme further by comparing simple outcome measures for longstay patients discharged from a large psychiatric hospital to a traditional NHS Trust 'hostel ward' with those resettled in group homes managed by a voluntary sector organisation.

The study
The study population comprised 30 long-stay patients allocated to one part of the Friern Hospital reprovision scheme, and discharged between 1987 and 1990. Although the patients were drawn from similar long-stay wards, they were not randomly allocated to the hostel ward or voluntary sector homes, but on the basis of ward of origin.

Health Trust hostel ward
Eleven patients were resident in a home run and staffed by the Health Trust where a Registered Mental Nurse (RMN) qualified nurse was always on duty. A staff member remained awake at night. Generally, three staff members would be on duty during the day and two at night.

Voluntary organisation homes
Nineteen patients were resident in three homes run by a consortium of agencies, including the Health Authority, Social Services and a Housing Association. Staff in the homes were drawn from a variety of backgrounds including health and social services although not all held specific qualifications. Sleeping in night cover was pro vided with one staff member on duty at night and two during the day.
Patients in the homes were all under the care of a consultant psychiatrist and were subject to regular community reviews. A multidisciplinary team provided day to day input. Additionally, all homes operated a 'key-therapist' system.
Age, diagnosis and total length of hospital stay prior to community placement were ob tained from medical case-notes. Evaluation of level of disability was undertaken by keytherapists using the 'Rehab' schedules (Hall & Baker, 1983). Details of readmission between 1st January 1990 and 31 December 1992 were obtained from the medical records department of the psychiatric in-patient unit.
In each setting the living environment was assessed using the Environmental Index which is an adaptation by TAPS of the Hospital and Hostel Practices Profile (Wing & Brown, 1970). This is a measure of the autonomy of individuals within their residential setting.
The multidisciplinary team and key-workers provided details of contact with friends and relatives, attendance at sheltered workshops and day centres. Patients' records were also examined for details of problems with drugs or alcohol, police involvement and whether patients had ever faced eviction from their residential setting, for example, by having a warning letter as part of an eviction process.

Findings
The baseline details are summarised in Table 1. The patients were comparable prior to place ment in terms of diagnosis, age and 'Rehab' scores. The total 'general behaviour' scores place the study group in the moderate to severe handicap range and on the 25th to 74th percentile for long-stay patient norms. A score of 40 or below is taken to signify 'discharge potential'. Patients resident in the NHS Trust hostel ward had spent significantly longer in hospital prior to community placement (t=2.616; d.f.=28; P<0.01). Table 2 summarises social and work activities; there are no significant differences between the groups on these.
Hospital readmission was required by ten (53%) of the patients in the consortium homes compared with two (19%) in the hostel ward. P=0.049). Alcohol or drug problems were noted in seven (23%) patients with no significant difference between the Trust hostel ward and consortium homes. A sub-analysis of the three consortium homes yielded no significant differ ences between each of them on the above measures.
The 'Rehab' scores for patients presenting these difficulties were no higher than average for the group.
The Environmental Index yielded a score of 3/ 89 for the consortium homes and 9/89 for the Trust hostel ward. Low scores reflect a more liberal environment.

Comments
The findings demonstrate a large discrepancy between the Trust hostel ward and consortium homes with respect to readmissions, police involvement and patients facing eviction. Of the five patients in our study facing eviction, two eventually had to leave their homes and were subsequently readmitted to hospital. None were evicted onto the street.
Our original expectation was that patients who had difficulties coping in community settings would have higher 'Rehab' scores whereas, in fact, no such differences were found. The overall 'Rehab' scores were high, reflecting our clinical impression that the patients were a severely disabled group.
There is little difference between the overall staffing levels of the projects, however the hostel ward had RMN trained staff and waking night cover. It is possible that the RMN staff had a greater tolerance of abnormal behaviour, con struing as illness rather than antisocial behaviour, but they were also able to carry out daily monitoring of mental health state and adminis tration of medication. From an administrative point of view it may have been easier for the voluntary sector homes to evict patients com pared with the hospital managed hostel ward.
The Environmental Index showed that the Trust hostel ward had a slightly more restrictive environmental than the consortium homes, although both are liberal. In a recent study (Anderson et al 1993), long-stay wards scored between 25 and 28. A score of 9 for the hostel ward is the same as that recorded by TAPS for long-stay patients discharged into the commu nity. It is possible that a number of the dependent patients had difficulty coping with the liberal setting of the consortium homes. TAPS empha sises the role of social networks in aiding community placement (Dayson et aL 1992). It was thus disappointing to find less than half the patients in this study in contact with friends or relatives, though the homes did not differ from one another in this regard.
This study highlights the need for close mon itoring and follow-up of patients in community schemes. There needs to be a spectrum of provision including some NHS Trust staffed placements of the hostel ward type to enable patients who are not coping with greater inde pendence to be transferred into more structured settings and vice versa, depending on need. It seems important that community schemes are integrated to allow patients to move between projects according to their needs, providing a support to voluntary sector homes and alterna tives to threatened eviction.