Hospital to hostel: what benefit?

Resettlement of 15 long-stay psychiatric in-patients from a hospital rehabilitation unit to a new hostel was assessed by measuring symptoms, social behaviour and client satisfaction beforehand, then six and 12 weeks afterwards. Three residents experienced relapses with varying degrees of recovery. Overall, symptoms and social behaviour improved significantly. Change in satisfaction with the services offered proved difficult to measure accurately. A well resourced resettlement project can enhance long-stay patients9 well-being.


The study
The purpose of this study was to assess the effect of resettlement of 13 males and two females from Tyrone and Fermanagh Hospital, Omagh, to Hillside, a new hostel. Fourteen of these patients had been in a rehabilitation ward and one patient, who had previously resisted attempts at resettlement, was in an adjacent continuing care ward. When the group were discharged to Hillside in May 1992, they were accompanied by 11 of the rehabilitation ward staff in an attempt to provide continuity of care. The new purposebuilt hostel adjoined a private nursing home. and was situated in a pleasant suburban area, three miles from the Tyrone and Fermanagh Hospital and one mile from the town centre of Omagh.
Ages of the patients ranged from 26 to 67 years, with a mean of 48.3 years. Twenty had a primary diagnosis of schizophrenia, two suffered from an organic psychosis post head injury and one suffered from depression. The total length of admission per patient, ranged from one year to 38 years, with a mean of 16.5 years. The number of admissions ranged from one (of 38 years' duration) to 18, with a mean of 6.8 admissions. The length of the last admission ranged from two months to 38 years, with a mean of 12.5 years. Only five patients were in hospital less than three years and eight had been in-patients for over nine years continuously. The time spent in the rehabilitation unit ranged from 0 to 34 months, with a mean of 17 months.
The hostel was staffed 24 hours a day and there was regular, frequent input from the rehabilitation multidisciplinary team. Residents were registered as out-patients with the local community health care team, and with the local health centre. A programme was implemented which involved all residents in the daily running of the hostel. This included personal laundry, cleaning their own single rooms, and setting tables and washing dishes on a rota.
Residents were encouraged to cook their own evening meals and to organise their own finances where appropriate. They were also encouraged to self-medicate. Residents were strongly encouraged to maintain contact with their relatives and also to use local leisure facilities to the full.
The Brief Psychiatric Rating Scale was used to measure changes in mental state. Hedlund & Vieweg (1980), in their review of studies using the BPRS, found inter-rater reliability to score 0.80 or higher. Changes in social behaviours were assessed using the Social Behaviour Schedule. Wykes & Sturt (1986) reported Kappa coefficients of 94% for interrater reliability, 86% for inter-informant reliability and 83% for test-retest reliability. With the aid of the relevant key worker, 21 behaviour areas were covered.
Client satisfaction with the service was assessed with the self-recording Client Satisfaction Questionnaire (Larsen et oÃ-, 1979). They reported a coefficient alpha of 0.93 for internal consistency and it correlated with therapists' estimÃ¢tes of client satisfaction.
Assessments were made before the move was announced, then six weeks and 12 weeks after the move.

Findings
The Brief Psychiatric Rating Scale at baseline showed moderate and moderately severe ratings in "emotional withdrawal" and "blunted affect". Of the two patients who experienced an exacerbation of symptoms at the six week assessment, one recovered well and showed a marked improvement in her score when seen again at 12 weeks. The other patient required a brief hospital admission to the acute ward. His rating did not return to base line. It is of note that he was planning his return to Australia. A third patient, although stable at six weeks, experienced an exacerbation of his symptoms at 12 weeks. He later moved on to a smaller home where he was much more content as he was not expected to participate fully in domestic chores. At 12 weeks seven patients had improved in their ratings, six showed no or minimal improvement and two had become worse. The total score for the group reduced from 159 at baseline, to 150 at six weeks, and then there was a further decrease to 117 at 12 weeks. This second decrease was statistically significant (P<0.05). Multivariate analysis of variance revealed a significant change in scores over the total time period (P<0.05). Individual items which dropped by at least 50% in their rating score were hallucinatory behaviour, hostility and tension.
Baseline assessment on the Social Behaviour Schedule underlined the high dependence of this group, with 11 scoring highly on deficits in personal appearance and hygiene. The deterioration in mental state of the two previously named patients was also noted in the worsening of their scores on the Social Behaviour Schedule. At 12 weeks nine patients had improved in their ratings, six showed no or minimal improvement and none had deteriorated. The total score for the group reduced from 182 at baseline to 175 at six weeks, and there was a statistically significant reduction to 128 at twelve weeks (P<0.05). Multivariate analysis of variance again showed a significant improvement over the twelve weeks (P<0.05). In particular, there were reductions in "bizarre topics of conversation", "talking to self', "unacceptable habits" and "personal hygiene." Taking both scales together, four patients improved by more than two points, whereas four patients showed minimal or no change on both scales. Three patients improved on the Social Behaviour Schedule but were static or minimally improved on the Brief Psychiatric Scale.
The Client Satisfaction Questionnaire was given to the 20 patients in the rehabilitation unit to complete anonymously. Of these, 15 were later confirmed to be moved to the new hostel. The mean score of 71% at baseline showed a moderate degree of satisfaction (the scores ranged from 53% to 91%). At the second assessment there was little change in the mean of 76% but the range had widened from 50% (fair satisfaction) to 100%. It is possible that the two clients who rated the service as only fair, and whose scores were markedly below the rest of the group, were the two individuals who resented the imposition of increased responsibility for domestic chores in their new environment. The lady who had been so reluctant to leave the hospital rated the new service as being 100% satisfactory.

Comments
The move appeared to trigger relapses of psychotic symptoms in two patients soon afterwards, and in a third patient at a later stage. However, one of the former recovered and even improved further than expected compared to her previous mental state. There were considerable stresses on the second patient and these were thought to contribute to the fact that he had only a patria! recovery. The third individual, despite having the lowest score on the Brief Psychiatric Rating Scale, and the second lowest score on the Social Behaviour Schedule had already clearly expressed his wish "to be looked after". Overall, there was a significant improvement in mental state and social behaviours.

Hospital to hostel what benefit?
Although there was no significant change found in user satisfaction, it is possible that the Client Satisfaction Questionnaire was not sufficiently sensitive, especially in a group such as this with chronic communication difficulties. The staff noted that residents were more content and were more actively involved in the daily running of the hostel. Many appeared to derive enjoyment from this, and from the fact that they had the security of a permanent home.
There was a great advantage in the staff being familiar with the patients as the supportive therapeutic relationships, which still encouraged independence of the clients, were allowed to continue. The fact that the hostel regime was probably rather similar may have had a protective effect in lessening the impact of the major change of discharge from hospital. Ideally the assessments would have been performed 12 and six weeks prior to discharge as well as immediately before the move was announced.
This would have allowed one to see if the improvements were continuations of trends already established in the rehabilitation unit. The impression from staff had been that they had improved, in varying amounts, after entering the rehabilitation programme, and had then mainly plateau ed. These patients continued to improve after their discharge despite having completed a rehabilitation programme and it could be speculated that at least some may continue to progress further.
Although acknowledging that this sample size was small, it is reassuring to see that, when resources are adequate, there is a definite benefit for the consumer in moving from hospital to hostel.