Effect of treatment in an active rehabilitation hostel on the need for hospital treatment

LELLIOTT, P. & WING,J. (1994) A national audit of new longstay patients. II. Impact on services. British Journal of Psychiatry. 165. 170-179. MILMISPROJECT GROUP (1995) Monitoring inner London mental Illness services. Psychiatric Bulletin. 19. 276-280. RITCHIE.J.. DICK,D. & LINGHAM. R. (1994) The Report of the Enquiry into the Care and Treatment of Christopher Clunis. London: HMSO. SHEPHERD.G.. BEADSMOORE, A.. MOORE. C., et al (1997) Relation between bed use. social deprivation, and overall bed availability in acute adult psychiatric units, and alternative residential options: a crosssectional survey, one day census data, and staff interviews. British Medical Journal, 314. 262-266. TYRER.P.. EVANS.K.. GANDHI.N.. et al (1998) Randomised controlled trial of two models of care for discharged psychiatric patients. British Medical Journal. 316, 106-109. WORKINGPARTYOF THE SECTIONFOR SOCIALANDCOMMUNITY PSYCHIATRYOF THE ROYALCOLLEGE OF PSYCHIATRISTS (1988) Psychiatric Beds and Resources: Factors Influencing Bed Use and Service Planning. London: Gaskell.

The mental hospital closure programme has been associated with the development of a range of residential community facilities.The rehabili tation hostel, also known as the 'ward in a street', has been recognised in recent reviews (Shepherd, 1991;Young, 1991) and by the Department of Health (1996) as an effective alternative to hospital care for the 'new long-stay' patient.Although such units were generally set up to provide long-term care, studies have consistently shown their success in rehabilitating even the most difficult patients with challenging beha viour, into less dependent settings (see Wykes & Wing. 1982;Shepherd et al. 1994;Reid & Garrety, 1996).
In more recent years, descriptions have appeared in the literature of hostels providing more explicitly "fast stream rehabilitation" (Simpson & Middleton, 1994).or "targeting patients with the greatest chance of rehabilita tion" (Shepherd et al, 1994).Such units, adopting a more active rehabilitation approach of deliberately shorter duration, may intervene earlier in the patient's illness history, aiming to improve psychosocial function and clinical out come, and to reduce the requirement for hospital treatment following residential rehabilitation (Hawthorne et al 1994).Such findings require that residential facilities (and their associated treatment programmes) should be viewed as treatment modalities in their own right, and their effectiveness assessed accordingly.
The present study was an attempt to examine the effectiveness of such a unit over a relatively prolonged period of its operation.

Service context
The Vron is a seven-bed hostel which opened in 1982.It is a detached Victorian house in a residential area of Gloucester, close to the city and a range of leisure/day care facilities.It is operated by the Mayfield Homes Trust Ltd as a psychiatric nursing home, with 10 full-time equivalent staff provided through the rehabilita tion unit in the Severn National Health Service Trust.The unit is not registered to treat patients under the Mental Health Act 1983, and does not have the status of a hospital, but provides 24hour nursed care.A full multi-disciplinary team meets weekly, with senior psychiatric, occupa tional therapy, social work and psychological input.
The unit was set up to provide short-or medium-term active training in rehabilitation with emphasis on domestic and social skills training.Each patient participates in an indivi dualised programme of cooking and independent living skills, together with education about ill ness and training in medication management.He or she generally attends sheltered work, training or a day care placement daily through the week.The peer group of residents play an important role through daily team meetings, and the unit can be viewed as a modified therapeutic com munity.Length of stay has typically been less than 18 months, with discharge of patients into a well developed range of supported accommoda tion, patients are generally taken on by the associated community rehabilitation team to ensure continuity.It is anticipated that on discharge the skills developed through the residential treatment programme are maintained in the new placement and that the intensive support can gradually be withdrawn leading up to discharge.Four other Mayfield hostels provide longer term care for 28 patients (all having some, albeit slower, throughput), with backup from a 12-bed rehabilitation in-patient unit for readmission/crises.This ward also provides all the continuing in-patient care within the district served by the Trust (population 320 000).

The study
All patients admitted to the Vron between April 1986 and February 1996 were included in the study.The unit was actually opened in April 1984, but health records were not available for the first two years of the unit's operation.There were 103 admissions over this time.Demo graphic data are shown in Table 1.
Health records were used to determine admis sion details.In most cases the patients remain under the care of the Gloucester rehabilitation service, and their progress can easily be followed through to the present day.ICD-9 diagnoses (WorldHealth Organization, 1978) were as at the time of Vron admission.In five cases, patients were admitted twice to the Vron.Under such circumstances, admissions were treated as separate episodes and analysed accordingly.However, where Vron readmission took place within the study follow-up time-frame, the readmission was counted as a hospital admission (on the principle that the unit acts as a Vard' in the Community).In practice, this occurred in only one case.

Statistical analysis
Data were entered in to the SPSS software system for analysis.A non-parametric test, the Wilcoxon matched-pairs signed-ranks test, was used to analyse change in in-patient stay one year and two years before and after treatment in the Vron.

Results
Demographic, illness-related data and place ments pre-and post-Vron admission are presented in Table 1.Mean duration of stay in the Vron was 29.2 weeks (range 0-107 weeks, s.d.=22.5).Length of time in the Vron was significantly negatively correlated with age (r=-0.21;P=0.035); patients over 35 years of age spent mean 32.0 weeks, versus mean 25.0 weeks for the under 35-year-olds.The age of patients had decreased since the time of the Vron opening, mean age 36.6 years (range 19-69, s.d.= 15.2) over the first six years, versus mean 33.5 (range 19-58, s.d.= 11.0) over the last six years of the study period.

Discussion
Our impression was that the unit adapted over the 12-year period studied, initially treating more elderly, institutionalised 'old' long-stay patients who had spent many years in institutions, but shifting its focus (with the closure of the longstay wards) to a younger, more heterogeneous patient group, admitted at an earlier stage of illness.The data showed that in the most recent seven years, six patients with personality dis order and five patients with 'other' diagnoses (two with learning disability, one substance misuse, one anxiety neurosis and one with an organic disorder) were admitted, as opposed to two personality disordered and one 'other' diag nosed patient in previous years.Most patients were admitted to the Vron from acute wards, with a significant proportion coming directly from community placements such as supported lodgings, which were breaking down.As a 24hour nursed facility, it has in some cases acted as an alternative to acute hospital treatment.
Over the study period pressure on the acute wards has increased, the Vron acting to reduce this pressure and to provide a more therapeutic setting for patients needing intensive support over the medium-term, and an active rehabilita tion programme to rebuild skills.
The results of this study show a marked, highly significant change in duration of hospital treatment between the year and two years before and after admission to the Vron.Although this was consistent with our impression that the unit was effectivelypreventing new long-stay patients and blocking 'revolving door' admissions, some caution is needed in interpretation of these findings: This was not a controlled study, and the effectiveness of 'standard' hospital treatment cannot be directly compared.There was no recording of clinical change through the period of Vron treatment (although rating of this is now in place), and actual health/social gains were not measured.Due to the lack of a control group, it is not possible to rule out the possibility that the reduction in admissions may be due to the natural progression of illness.Similarly, It is possible (as most patients were discharged to supported accommodation) that the longer term support after discharge is what prevented further hospitalisation, rather than the impact of the Vron treatment programme.
However, the change in requirement for hospi tal care was considerable, and was in keeping with a finding in the only similar previous study found in the literature: Hawthorne et al (1994) found that 104 patients treated in two hostels in San Diego had significantly fewer admissions and shorter hospital stay in the year after (versus the year before) residential care.In the American study, the researchers also demonstrated sig nificantly less psychopathology in patients, following hostel rehabilitation, and concluded that psychosocial residential treatment could offer "cost-effective and clinically efficacious care to persistently mentally ill patients".These authors concluded that projected savings in hospital costs alone exceeded the annual costs of hostel facilities.Lelliott (1996) showed that in the UKthere are now only one-third of residential places for patients with severe mental illness as compared with 40 years ago.The NHS Executive has recommended that health authorities should commission a further 5000 24-hour nursed beds nationally (NHS Executive, 1994), and our findings add further support to the potential value of such units.
The experience of staff working in the Vron suggests that a number of factors may be important to the success of such units: Engage ment in a therapeutic programme is crucial and needs to be established well before admission, usually by a programme of gradually increasing visits.Family work is generally necessary, in cluding at times the employment of contracts to manage boundary issues in over-involved fa milies.Regular input from outside agencies for staff and patients, through the course of admis sion, is encouraged.Training and staff support are important.Strong links with the supported accommodation officer and the community re habilitation team have proved helpful in finding appropriate placement and ongoing support after residential rehabilitation.

Table 1 .
Demographic/illness-related data and placement pre-and post-Vron treatment