The Team for the Assessment of Psychiatric Services (TAPS) was established in May 1985 with the explicit purpose of evaluating the national policy of replacing psychiatric hospitals with district based services. TAPS' remit was to mount the evaluation with respect to the closure of Friern and Claybury Hospitals in north London. Funding was provided by the Department of Health and North-East Thames Regional Health Authority (latterly North Thames Regional Health Authority) with supplements from the King's Fund and the Sainsbury Family Trusts. During 13 years of research TAPS has employed more than 50 members of staff and associated researchers. A full listing is given in Leff (1997). This summary of the research is organised by topics. However, there was very close integration of the various projects.
Long-stay, non-demented patients
This group was defined as patients staying more than one year in hospital who did not suffer from Alzheimer's disease or other forms of dementia. The number of patients meeting the selection criteria grew from 770 in 1985 to 1166 over the course of the study by the accumulation of ‘new long-stay’ patients from the admission wards. During the study 670 patients were discharged, of whom one-third were ‘new long-stay’ patients. The TAPS sample comprises the total long-stay population of Friern Hospital, but only a proportion of the Claybury patients.
TAPS developed a batch of nine schedules which together composed a comprehensive picture of the patients' quality of life. Here we summarise the follow-up data collected after the patients had been in the community for five years.
Nearly 80% of the patients were discharged to staffed houses, with an average of eight residents per home. At the end of five years, two-thirds of the patients were still living in their original residence.
Deaths and suicides
Reprovision did not increase the death rate or the suicide rate.
Lost to follow-up
We estimate that four patients became homeless, giving a vagrancy rate over five years of 0.6%. Three of the patients had a history of vagrancy prior to admission to hospital. No patient was lost from a staffed residential home. Patients with a history of vagrancy should not be discharged to unsupervised accommodation.
Over the five years there were 24 recorded criminal incidents committed by 18 patients. Of these, 15 were assaults committed by 13 patients, constituting 2% of the total sample. Three were assaults on staff or the police, three on fellow residents and nine on members of the public. Of the latter, one was attempted murder and three were sexual assaults. Three of the patients who assaulted members of the public were imprisoned, while four went to secure units. It would obviously be desirable, if not always feasible, to identify the small group of dangerous patients who are responsible for assaults, prior to discharge and to ensure that they were cared for in a secure facility.
The overall readmission rate for the five-year follow-up period was 38%. Of these, one-third remained in hospital for over one year, technically becoming long-stay patients once again. At the time of follow-up 10% of the sample were in hospital. This confirms our estimate that 9-10 beds need to be provided for every 100 long-stay patients discharged to the community, about half of these being in rehabilitation facilities. This need does not seem to have been taken into account in planning inpatient psychiatric facilities.
Clinical and social outcomes
Psychiatric symptoms and social behaviour problems remained unchanged overall. Patients gained skills in using community facilities such as public transport. They also improved their domestic skills in the first year but some of these gains were lost over the next four years. The community homes were much less restrictive than the hospital wards, with an average of 10 rules and regulations compared with 26 on hospital wards. Patients were very appreciative of their increased freedom, and 84% wished to remain in their present home. There was no marked deterioration in patients' physical health, but there was a steady increase in the number exhibiting restricted mobility and incontinence, which went with increasing age.
Patients' social networks did not enlarge, but they made more friends and increased the number of people they could confide in. There was also an increase in the proportion of their social contacts who were ordinary members of the public, including neighbours. However, these desirable changes benefited only a small proportion of the patients, the majority of whom continued to lead impoverished social lives. Overall, patients' quality of life was greatly improved by the move to the community, but deficiencies remained due to the nature of severe psychiatric illness.
Patients who are ‘difficult to place’
Shortly before Friern Hospital closed, a group of 72 patients was identified who were considered too disturbed or disturbing to the public to manage in standard staffed homes. These patients who were ‘difficult to place’ were transferred to four high-staffed facilities (up to 1.7 staff:1 patient), three of which were in hospital grounds. About half the patients had had recent problems with aggression, while some were sexually disinhibited. After one year there was little change in the patients' problematic behaviour and only a handful had moved to standard homes. However, in the next four years everyday living skills improved significantly and there was more than a 50% reduction in problem behaviours, particularly aggression. As a result, 40% of the group had been discharged to standard community homes by the end of five years. Thus, a high staff input can reduce severely problematic behaviours over several years, allowing difficult-to-place patients to be settled in ordinary community homes.
Comprehensive costing of services used by former long-stay patients
The economic evaluation was carried out by the Centre for the Economics of Mental Health, Institute of Psychiatry, this being a continuation of the work previously undertaken by the same team at the Personal Social Services Research Unit, University of Kent. The economic evaluation, closely linked to the TAPS research, examined service use patterns, costs and the links between them and individual needs and outcomes. It focused primarily on long-stay, non-demented patients.
A wide variety of community care packages were being received. The NHS, local authorities, voluntary organisations and private sector bodies were all involved in providing services. The costs of these and other services were calculated using appropriate combinations of local and national figures. Because the most able patients were selected for discharge first, community costs were generally higher for each successive annual cohort (or group). The average total cost of care for the first cohort, discharged in 1985-1986, was £347 per week during the first year in the community (all costs expressed at 1997-1998 price levels). Mean costs rose to £819 for Cohort 8 and £1158 for the ‘difficult to place’ group. Over the whole study sample, the mean cost per week was £664 (median £638).
Cost changes over time
Community care costs had usually increased slightly by the time the study group was followed up five years after leaving Friern or Claybury (overall, the mean weekly cost rose to £693 and the median to £714). This may be partly due to service responses to the greater age and frailty of the study group, and to increasingly complex management arrangements by the facilities providing accommodation. By the five-year follow-up many of the ‘difficult to place’ group, however, had been able to move from the intensively-staffed units which had been set up for them upon discharge. As a result of these moves to less highly supported accommodation facilities, the mean cost of care for this group fell by 14% to £994 per resident week. Providing intensive rehabilitation for people who present a severe challenge to community services is crucial to contain the cost of community care within the resources available from the closure of the psychiatric hospital.
Over the years of the study there was a shift in funding sources, notably a fall in the NHS share of funding from 61% to 54%. Service users' contributions (almost all from social security payments) rose from 25 to 35% of the total. These figures raise questions about the destination of resources saved when hospitals close, particularly the ‘dowry’ money (Section 28 payments) attached to individual hospital-leavers. It may be that some of the funds saved by the NHS as a result of closing the hospitals have been diverted away from mental health services.
Overall, it appeared that there was little difference between hospital and community costs. Coupled with the outcome findings, the economic evaluation suggests that community-based care is more cost effective than long-stay hospital care.
Psychogeriatric wards compared with community facilities
The psychogeriatric wards at Friern Hospital were compared with community facilities set up to replace them by comparing groups of patients in the two locations. As the majority of the patients suffered from dementia it was not possible to seek their opinions. Instead an observational study was conducted in both settings, and relatives were asked their opinions about the quality of care. Relatives cited a number of aspects in which they considered the community facilities to be an improvement on the hospital wards. Some of these were confirmed by the observational study, which also revealed that staff spent more time interacting socially with the patients in the community facilities.
Patients using admission services
Patients preferred the admission wards at Friern Hospital because of the space and freedom afforded by the extensive grounds. The general hospitals were on busy central London streets and had no space around them. Patients also disliked the noise level in the general hospital units. Nurses considered that after the closure of Friern Hospital patients were staying too short a time to gain the maximum benefit from treatment, and the nurses' morale dropped significantly. A survey of out-of-hours referrals to the general hospitals found that following the closure of Friern Hospital there was a significant increase in the number of patients considered by the duty psychiatrist to need admission, but for whom no bed could be found. There was a concomitant rise in the number of patients admitted to private hospitals.
This section of the replacement services for the psychiatric hospital was the one area found by TAPS to be beset by serious problems. Admission services throughout the country are experiencing similar difficulties, many of them operating at well above 100% occupancy. It is difficult to determine how much of the problem is due to an increased demand for admission and how much is a consequence of the closure of psychiatric hospitals. The reprovision plans for Friern Hospital included a reduction in the number of admission beds in the general hospitals. This policy had to be reversed when the service rapidly ran into crisis. This was partly due to the unrecognised need of discharged long-stay patients for admission beds (see Readmission section) and partly to the failure to reprovide the rehabilitation function of the psychiatric hospital, leading to some patients remaining for many months on the admission wards. The problem could be tackled by providing more admission beds. A less costly option is to provide a rehabilitation facility in each health district and to invest in alternatives to admission such as acute day hospitals, which have proved their value.
A survey of care staff found that a half of the community staff and one-third of the hospital staff felt they needed more training to help them understand the patients' behaviour. All 54 staffed group homes that were part of the Friern Hospital reprovision programme were approached for information on staff training, and 48 participated. Just over a half of the 502 staff surveyed had no formal qualification. This was most marked in the private sector, where 68% of the staff were unqualified. Of the staff who completed a questionnaire, 68.7% stated a need for further training.
In response to this obvious training need, we developed a training programme of nine sessions. Evaluation of the course showed that staff members learned to use more strategies involving an attempt to effect change and utilising more resource facilities. There was also a shift away from communicating the carer's perspective and towards confirming the patient's view of himself or herself. These beneficial changes lead us to recommend that brief focused training courses should be established for all care workers in staffed group homes.
We mounted two surveys of public attitudes in north and south London. The findings were very similar: the public stereotype of a person with mental illness was of someone who was difficult to communicate with, dishevelled and unpredictable. In view of the media highlighting of homicides by the mentally ill, it was surprising that aggression was mentioned by only 20% of the respondents. There was also a large amount of goodwill towards people coming out of long-stay institutions. An overwhelming majority of local residents in the first survey showed positive attitudes towards the opening of mental health facilities in their neighbourhood (over 80%), and many showed interest in offering help. However, two-thirds believed that preparation of the local residents prior to opening a mental health facility in their area was important. Such preparation should include information about mental illness as well as the opportunity to develop practical skills in dealing with people with mental illness.
In response to these encouraging findings we mounted an experiment in south London around reprovision for Tooting Bec Hospital. We studied two staffed homes for discharged long-stay patients, situated in adjacent districts. In one of the streets we ran an education campaign for the neighbours and surveyed their attitudes before and after the intervention. In the other street we conducted the two surveys at the same time interval, but without any educational input. Comparison of the experimental with the control street showed that the campaign was successful in increasing neighbours' understanding and reducing their fear of people with mental illness. These changes in attitude were reflected in behaviour, since some of the neighbours in the experimental street visited the patients and invited them to their homes, whereas there was no social activity of this kind in the control street. Moreover, the social networks of the experimental patients enlarged while those of the control patients remained static. We conclude that localised education campaigns are effective in improving the social integration of patients into their neighbourhoods.
For long-stay, non-demented patients, and for those with dementia, the transition from care in the psychiatric hospital to community care has been a qualified success. Although psychiatric symptoms and social behaviour problems remained virtually unchanged, community living skills were acquired, patients appreciated the increased freedom in their new homes, and some enriched their social networks. Community-based care has also been shown to be affordable, although it is important that funds saved by the NHS as a result of hospital closure are not diverted from mental health services.
Hospital acute admission services are currently failing patients and staff. Although this may not be entirely attributable to the change in the locus of care, the need of discharged long-stay patients for admission beds was not recognised, and the rehabilitation function of the psychiatric hospital was not preprovided. Provision of more beds is only a part of the solution. A rehabilitation facility in each health district and investment in alternatives to admission, such as acute day hospitals should be seriously considered.
- © 2000 Royal College of Psychiatrists