Abstract
In 1994 an initial cohort of 40 long-term psychiatric hospital patients transferred to community residences, in conjunction with a plan to close a psychiatric hospital. A further seven patients transferred to the community in the next year when vacancies arose due to hospital readmissions in the initial cohort. If not for the hospital closure these patients were unlikely to have been considered for discharge. This study provides a 6-year quantitative and qualitative evaluation of the progress of these 47 patients.
The deinstitutionalization plan was part of a NSWgovernment initiated rationalization of mental health services resulting in the amalgamation of two psychiatric hospitals. Upon hospital discharge the patients transferred to one of three community houses or a cluster apartment block located in four community mental health subareas of Sydney. The community services were well-established and provided continuity of care with 24-h supervision by a multidisciplinary team, intensive case management, individual rehabilitation programmes and crisis intervention. Finance of $40 000 per person was provided for the establishment of community residences. Far greater savings were made from the merger of the hospitals.
A detailed description of the residents' backgrounds, community services and implementation of the deinstitutionalization plan is presented in an earlier study of the first 2 years of community transition [1]. In summary, a primary diagnosis of schizophrenia was applicable for 98% of the 47 patients that included 25 men and 22 women. All had extensive periods of hospitalization with an 8-year mean predischarge length of stay (range 2–43 years). The mean age at final discharge was 41 years (range 25–64 years). During the first 2 years one male resident died of medical illness, seven residents (six females and one male) returned to hospital, and community tenure of 2 years was achieved by 35 of the 39 community residents with four residents having a shorter duration of community tenancy.
Community-based residents showed a significant improvement in their psychiatric symptoms over the first 2 years, without significant change in dose of neuroleptic medication. These changes could not be attributed to any increasing use of atypical antipsychotic medication at the time. While living skills, depressive symptoms and social behaviour problems were largely unchanged, residents reported a significant increase in life satisfaction. By the 2-year follow-up, three residents had moved from 24-h supervised accommodation to smaller group homes with supervision via a daily mobile team visit and 24-h phone access.
An intensive ethnographic study of the 2-year community transition [2] supported the clinical findings. The residents who continued to live outside hospital all reported this as their preference. Subjectively residents felt that community living had enhanced their quality of life. The community transfer was also found to be economically rational. The cost of care for a client in the community residence was calculated to be 43% of the cost of care in hospital [3].
In 2000, a 6-year evaluation was conducted to assess the progress of all the participants. This included both community and hospital-based residents. The objective of the 6-year follow-up was threefold: (a) to determine the accommodation style and level of care required by residents; (b) to evaluate clinical changes over time; and (c) to gain the residents' perspectives of their lives. To this end the 6-year evaluation included both quantitative and qualitative components.
Method
Qualitative semistructured interviews
Unlike the original study where ethnography comprised two and a half years of participant observation fieldwork, the 6-year follow up was a snapshot of hospital and community residents. The aim was to enable residents to express how they felt about their current situation. Essentially the semistructured interviews provided descriptive material to enhance the clinical data. Interview questions encompassed areas of living preference, freedom, restrictions, levels of support, future goals and expectations, social networks, notions of self, happiness and identity and what was of meaning or value in their present life.
Quantitative outcome measures
The clinical measures used in this study were the Brief Psychiatric Rating Scale (BPRS) [4], the Life Skills Profile (LSP) [5], the Social Behaviour Schedule (SBS) [6], the Montgomery Asberg Depression Rating Scale (MADRS) [7], the Quality Of Life index (QOL) [1] and medication levels. Apart from the QOL, the measures were all informant rated by psychiatrically experienced treating clinicians. All clinicians were trained in formal clinical evaluation and the BPRS and LSP were used for regular assessment in the service. These measures have established interrater reliability. Clinician ratings provided clinical validity of the data, due to their constant contact and knowledge of the clients, their symptoms and behaviour. The project researcher (CH) supervised completion of outcome measures and also administered the QOL to residents.
Evaluation for community-based residents was conducted at predischarge and then at 2- and 6-year intervals following community transfer. As research evaluation was discontinued following hospital readmission, hospital-based residents were only assessed at predischarge and 6-years.
Statistical analysis
Assessment data were analysed by repeated measures analysis of variance. In this analysis the predischarge data of each resident were utilized as the control for the subsequent individual evaluations. Polynomial contrasts were used to examine any trends in the data over time. Non-parametric statistical methods were utilized for the analysis of categorical data. In some cases variables were transformed (collapsed into dichotomies) to analyse pertinent hypotheses.
Results
Resident movements
Seven residents were in hospital for long-term management at the 6-year follow-up. Six of the seven returned in the first 18 months although one of these residents relocated to the community after a 4-year hospitalization. Concurrently, another resident returned to hospital after 4 years living in the community due to deterioration in mental state. As both residents were female, the gender ratio of rehospitalized residents remained constant comprising six females and one male. Over the 6 years, three male residents died of medical causes unrelated to their mental illness.
Incidents, readmissions and respite care
There were no reports of any major community related incidents and no accounts of any community victimization or exploitation over the 6 years. Physical aggression predominantly resulted from resident altercations with three serious non-fatal incidents of assault recorded. Two incidents occurred in the first 2 years and involved one of the seven residents who returned to hospital after hitting a staff member and assaulting another resident with a knife. A second resident assaulted a flatmate with a knife just prior to the 6-year follow-up and was detained in a correctional services facility. Only two self-harm attempts occurred, both within the first 2 years of community transition. The two residents recovered and still resided in the community at the 6-year evaluation. During the study, one resident gave birth to a child who was adopted. Temporary readmission or respite care was required for 38% of community residents during the first 2 years and 28% in the successive 4 years. No major substance abuse problem arose during this study. Despite staff-initiated smoking reduction programmes, 75% of the residents smoked cigarettes with 58% designated as heavy smokers. There were no alcohol-related problems.
Hospital-based residents
Clinical measures
At the 6-year evaluation, the seven hospital residents were located in rehabilitation wards or a hospital hostel. Analysis of clinical measures in Table 1 reveals that residents' psychiatric symptoms significantly improved in the highly structured environment and on atypical antipsychotic medication. There was no significant change in living skills, social behaviour problems or depression. Although the average level of medication substantially decreased, the difference in means was nonsignificant due to the considerable range in medication levels.
Outcome measures for hospital-based residents (n = 7)
In marked contrast to community-based residents, there was no improvement in QOL scores of rehospitalized residents. Paradoxically whilst they were all dissatisfied with being in hospital, they resisted returning to the community in the short term. Three people preferred to stay in hospital. Four residents harboured a desire to return to the community in the future but currently viewed themselves as ‘too sick’, ‘not ready’ or described community living as ‘too stressful’.
BPRS scores for hospital residents were significantly higher than community residents at predischarge but were not significantly different at 6 years (f = 6.321; d.f. = 1,36; p = 0.017). There were no significant differences between hospital and community residents at 6 years on other measures of the LSP (f = 0.002; d.f. = 1,37; p = 0.964), SBS (f = 0.531; d.f. = 1,37; p = 0.471), MADRS (f = 0.327; d.f. = 1,37; p = 0.571) or medication (f = 1.193; d.f. = 1,37; p = 0.282). Hospital residents, however, would require intensive supervision during community transition to develop their skills and alleviate their anxieties about coping in the community.
Community-based residents
Community accommodation style
At the end of the first 2 years only three residents had moved to more independent accommodation. The other community-based residents lived in the three originally established 10-bed group homes or the cluster apartment block. In the ensuing 4-years however, there was a major change in accommodation style, as shown in Table 2. Only 13 of the 36 community-based residents remained in large group homes at the 6-year mark. The majority of residents (n = 23) moved to smaller three-bedroom units and houses or still resided in the cluster apartment block (with four three-bedroom apartments). The move to more independent accommodation was a function of both the capability of the resident to live independently and availability of suitable accommodation. This trend towards smaller scale accommodation was not just apparent for these residents but also in the wider community service.
Community accommodation style (n = 36)
The major clinical impetus for this change was ‘normalization’ of lifestyle for residents. In the large 10-bed houses, cooking, shopping and cleaning was not a normalizing experience. Similarly, staff recognized that rehabilitation was needed on an individual basis. In this respect the cluster apartments proved to be a more suitable model for accommodation as residents capable of shopping and cooking for themselves did so and individual training was provided for those needing assistance. In smaller households, residents made decisions for themselves regarding household responsibilities and it was easier to achieve conflict-free compatibility between residents.
Level of care
The most dramatic change in clinical care over the 6 years was that none of the residents required intensive 24-h supervision. The majority of the residents (64%) however, still required daily assistance from case managers. While this was provided on a daily or nightly eighthour basis in the group homes, 28% of the residents required a daily visit from case managers for selective assistance. This assistance was for a specific area of need such as supervision with medication, meal preparation, shopping, cleaning or finances. Impressively 36% of the residents were living semi-independently and were responsible for their own self-care. This group needed only a weekly visit from case managers, and three residents were functioning well with only a monthly visit.
Community outcome measures
Data were collected on objective measures for all 36 community residents at the 6-year follow-up. However, as four residents had community tenure of 5 years and one person had only recently returned to the community after a 4-year hospitalization, their data were excluded from repeated measures analysis.
Psychiatric symptoms, living skills, social behaviour and depression:
Overall, community-based residents remained clinically stable over the 6-years without significant change in BPRS, LSP, SBS and MADRS total or factor scores. A mild level of depression (MADRS mean total score of 17) was detected with some residents expressing despondence due to having a mental illness. Anti-depressant medication was prescribed for 10% of residents at predischarge, 7% at 2-years and 16% at 6-years (Cochran's Q = 2.333; d.f. = 2; p = 0.311).
Psychopharmacotherapy
There was a prominent reduction in the overall level of antipsychotic medication over the 6 years. Daily mean doses decreased significantly in chlorpromazine equivalents and this occurred without any decrement to psychiatric symptomatology as was evident from the absence of change in BPRS scores.
A major impact on the level of medication that cannot be discounted is the use of atypical antipsychotic medication. The use of this medication significantly increased over the 6 years (Cochran's Q = 29.789; d.f. = 2; p = 0.000) especially between two- and 6-year assessments. Initially only 13% of the residents were prescribed this class of drugs at predischarge. This rose to 32% at 2 years and by the 6-year mark 74% were prescribed atypical antipsychotic medication. During the first 2 years, clozapine was the only atypical antipsychotic medication in use. Recent newer drugs were increasingly later prescribed.
Quality of life
Response rates: The response rate at predischarge was 84% but reduced to 61% at the 2-year follow up and improved to 77% at 6 years. All three QOL questionnaires were completed by 55% of residents over the 6-years with only two residents (6.5%) unable to complete any due to thought disorder or paranoia. In order to reduce the possibility of response bias, an independent groups t-test was performed using 6-year QOL assessment data. There was no significant difference in 6-year QOL scores (t = −0.830, p = 0.415) between consistent (mean QOL score 123) and inconsistent (mean QOL score 128) respondents.
Changes in quality of life: Consistent with results from the 2-year community transition period, residents reported a significant overall improvement in life satisfaction over the 6-years of living in the community (refer to Table 3). When interviewed at 6 years, the things that were most valued included freedom and the independence, along with the ‘peace and quiet’ of personal space in their own homes.
Outcome measures for community-based residents (n = 31)
Hobby and rehabilitation involvement: About half of the residents had a continuing independent active interest in hobbies over the 6 years (Cochran's Q = 1.286; d.f. = 2; p = 0.526). However, the remaining residents (61% at predischarge, 52% at 2 years and 61% at 6 years) had only fluctuating or passive interests (television or radio). These residents were inclined to stay at the community residence. Short trips to the local shops, superficial conversation, coffee and cigarettes were the mainstay daily interests.
This situation was compounded by a significant reducing involvement in any formal rehabilitation programme (especially day centre attendance) over the 6 years (Cochran's Q = 10.5; d.f. = 2; p = 0.005). Despite staff encouragement 22% of residents at predischarge, 26% at 2 years and 52% at 6 years, declined rehabilitation participation. The ageing of the residents (with a median age of 46 years at 6-year follow up) was a contributing factor as older residents favoured social activities in preference to formal rehabilitation programmes and would especially engage in social activities when arranged by staff or families. In addition, rehabilitation has undergone change over the 6 years with a greater emphasis on individual vocational training, education, and work less relevant to this age group. Three residents were participating in employment schemes at the 6-year evaluation.
Discussion
The major finding of this study was that residents had improved life satisfaction living in the community. Their clinical stability over 6 years of community life was maintained in the face of a significant decrease in both the overall level of psychopharmacology and in the level of case manager supervision. These results are comparable to the Team for the Assessment of Psychiatric Services (TAPS) 5-year follow up of 523 residents after community transfer [8]. TAPS found no decline in the residents' clinical state or social behaviour problems. While short-term improvement occurred in domestic skills, the only sustained significant improvement was in community skills due to greater use of public facilities. Significantly, 84% of TAPS residents expressed a preference for community living and although social networks did not increase, residents reported having more friends and confidantes.
Community transition in both studies, however, was not without incidents. As suggested in the TAPS study [9] patients with a long-term serious mental illness will experience exacerbation of symptoms requiring periodic readmission, and allocation of nine to 10 beds per 100 patients discharged to the community was recommended. While not minimizing incidents occurring during the 6 years of this study, suicide attempts were minimal and substance abuse problems were negligible especially as residents had limited personal finance. Factors contributing to the low incident rates in both studies were the older average age of the residents and continuity of care that in particular prevented major homelessness and medication non-compliance. In this study residents received intensive 24-h supervision in the early stage of community transition, which may account for the absence of criminal activity and major incidents involving community members.
The main change over the 6 years of this study was in accommodation style and intensity of supervision. A move to smaller two to three person accommodation was favoured by both residents and staff. By the 6-year mark 24-h supervision had been reduced to 8-h supervision and 36% of residents were living semi-independently with only minimal supervision; however even after 6 years of community living, 64% of residents still required some form of staff assistance to attend to their daily needs.
Although the majority of residents engaged in visits to the local shops during the week, an increasing level of inactivity was apparent over the 6 years. While active interest in hobbies remained constant, there was a marked reduction in rehabilitation programme involvement. This could partially be attributed to the decreasing relevance of rehabilitation for this group who were largely middle-aged. Very few reported new goals for the future and most saw their lives staying the same. Also, residents reported a desire for involvement in social rather than rehabilitation activities. The apparent dilemma is that residents value the freedom of greater independence but also express the need for staff who can spend more time doing activities with them, as they find it difficult to enhance their own social networks. Success with systems akin to the buddy system have been reported [10] whereby more able residents assist less able residents with chores, coping skills, activities, socialization and aspects of case management. Indeed a consumer-driven social facility has recently been established in the service area in this study with a growing network of consumer and community volunteers to assist others with transport and support.
Although a few residents were linked into social organizations through hobbies and interests, for the majority social contact was limited to locals in the community, such as shopkeepers, neighbours and service providers, but these acquaintances were of a superficial nature. The residents' social networks were still largely connected to the residential facilities, families and mental health services. In part, successful community adaptation depends upon acceptance of the mentally ill by their community. A public education intervention in the UK [11] has not only resulted in a reduction of stereotypic images of the mentally ill as aggressive but also increased the likelihood of social contact and interaction with residents which perpetuates a further reduction of fear, rejecting attitudes and exclusion of mentally ill people.
Two pertinent accommodation issues emerged in this study. A 24-h staff supervised setting is necessary for residents to cope with community transition. Any further process of deinstitutionalization with similarly disabled residents would require this level of care, at least in the initial stages. Indeed, the residents who had returned to hospital expressed a need for continuing intensive staff supervision, were they to return to community living.
Assuming this need for 24-h supervision, a purpose built clustered housing model would have been the most appropriate. Such funds were unavailable to our service at the time. This allows a combination of autonomy, companionship to reduce loneliness, and access to staff on a more regular and needs-based basis.
Another important issue that needs addressing is the ageing of this population. Some residents returned to hospital following this study, due to physical infirmity from ageing or medical illness that reduced their ability to cope with daily living tasks in the community. Yet these residents were not of an age or level of physical frailty suitable for a nursing home environment. Clearly the provision of aged care hostels will become a pressing need as this population ages.
Conclusion
In conclusion, people with long-term serious mental illness can achieve improved life satisfaction, remain clinically stable with less medication and maintain community tenure when supported by a mental health system with adequate community resources and continuity of care. Social integration can be further improved by consumer networking and public education.
Footnotes
Acknowledgements
This research was funded by the Australian Commonwealth Department of Health and Family Services Research and Development Grants Advisory Committee (RADGAC), the Northern Sydney Area Health Services (NSAHS) and the Centre for Mental Health, New South Wales Department of Health. The research team would like to thank community and hospital residents, staff and management for their continued assistance with this project.