Giovanni de Girolamo; Mariano Bassi

Curr Opin Psychiatry 17(4):275-281, 2004. © 2004 Lippincott Williams & Wilkins

Posted 08/23/2004

 

In the last few decades, in parallel with the closing of many mental hospitals around the world, large numbers of severely mentally ill patients have been relocated to a variety of nonhospital residential settings, called residential facilities. These include different types of accommodation such as 'group homes', 'intermediate facilities', 'wards in the community', 'boarding-and-care homes', 'supervised hostels' or 'sheltered apartments'. To date, no internationally agreed and precise taxonomy has been developed to classify residential facilities and identify their distinct functions for specific groups of patients with severe mental disorders. Different facility names (as indicated above) often correspond to identical settings, while sometimes facilities having similar names exhibit very different characteristics and pursue nonconvergent objectives.

Despite the magnitude of this development, research in this area has so far been quite sparse. The aim of this contribution is to provide a nonsystematic review of the literature dealing with residential facilities for the mentally ill published since 2000. We aim to draw some general conclusions about residential, long-term care and to highlight needed areas of research.

 

Studies included in this review have been identified through Medline, using the key words 'residential facilities' or 'residences', and limiting the search to English language articles published from 2000 onwards. We also inspected the reference lists of published articles. We excluded all articles dealing with residential facilities for children and adolescents, substance abusers and the elderly. Although we are aware that many residential facilities for the elderly may host aged mentally ill, generally also showing cognitive impairment, this topic needs a separate, specific analysis.

We also excluded a few papers in which the term 'residential facility' (or 'intermediate facility') referred (arguably in an inappropriate fashion) to very large institutions, which should be considered variants of traditional asylums, with up to 100-400 beds.^[1-3] Studies which focused on residential facilities with more than 60 beds have consequently been excluded from this review.

The paper is organized under the traditional headings used for quality evaluation in the health field, that is structure (input) data (e.g. service provision, number of personnel, etc.), process data (e.g. how these facilities operate, rules, etc.) and outcome data (e.g. longitudinal and quality of life (QOL) studies, qualitative studies, etc.). Data on residents' characteristics have been grouped in the input section.

 

How many residential facilities are currently available in different countries, and how many residential facilities are needed? Unfortunately it is difficult to answer these questions; the only study which has directly evaluated the provision of residential facilities on a nationwide scale is the PROGRES (PROGetto strutture RESidenziali, Residential Project) study, recently carried out in Italy where all mental hospitals have been closed.^[4**] The aim of the first phase of this survey was to obtain data on regional provision of residential facilities, staffing arrangements, demographic and clinical characteristics of residents and discharge rates. In addition, the study aimed to clarify relationships between the availability of residential places and the provision of other mental health resources. The authors identified 1370 residential facilities in the whole country, with a total of 17 138 beds; three-quarters of these facilities had 24 h staff cover. Overall, in the whole country, there were 2.9 residential beds for every 10 000 inhabitants, with a large variability between regions.

The EPSILON study has recently gathered data about the provision of residential beds in five European catchment areas (Amsterdam, Copenhagen, London, Santander and Verona).^[5*] For service assessment, the study group used the European Service Mapping Schedule. The rate of nonhospital residential beds varied dramatically, with a 41-fold difference between the area with the highest provision (Copenhagen with 14.6 residential beds per 10 000 inhabitants) and the area with the lowest (Santander, Spain, with 0.35 beds), both with 24 h cover. London and Verona showed intermediate values (3.3 and 3.6, respectively). It is difficult to establish, however, to what extent these figures can be generalized to other areas of the same countries and therefore how they reflect the national average. Indeed, another report, focusing on the German state of Saxony-Anhalt and carried out with the same instrument (European Service Mapping Schedule), has found a much higher rate of residential beds, that is 2.4 places per 10 000 in medium stay residential facilities and an additional 24.0 places in indefinite stay residential facilities, with 24 h cover.^[6] In the past, in the UK 'Mental Health Residential Care Study' carried out in eight districts surrounding London, the authors found a rate of 9.46 beds per 10 000.^[7**] In another UK survey, however, of 35 districts, the rate was 4.29, closer to the 2.98 Italian rate.^[8]

In conclusion, although it may prove to be difficult to establish precise standards of residential facility provision, since 'housing needs assessments crucially depend on the range and quality of other local services and cannot be separated from the functioning and dynamics of the total service system ',^[9**] efforts are needed to identify acceptable ranges of residential facility places in all countries with a developed system of mental health care. Provision of residential facilities largely depends on two key variables: (1) the extent of informal family support, which can substitute and replace the formal support granted by residential facilities; and (2) the availability of comprehensive community resources, including assertive community treatment in each country. In the PROGRES study, the availability of residential places was indeed negatively related to the number of community mental health centres and day centres, and was positively related to the number of beds in private inpatient units in each region.

Of course, additional efforts will be necessary to clarify the need for residential facilities in those developing countries which are trying, with much difficulty, to overcome the traditional asylum system; quantitative standards will have to be adapted to the specific, local contexts and needs of such countries.

The Size of the Residential Facilities

In contrast to the Italian data, showing that residential facilities have on average small dimensions (mean size 12.5 beds, with only 5.7% of facilities having more than 20 beds), recent North American data show substantial variability in the size of long-term care settings. In a survey of 68 residential facilities providing acute care and managed by the US Veterans Administrations, Timko et al.^[10*] found that they had a mean size of 42.9 beds, with an average stay of 14.4 weeks. The proportion of such facilities on the total number of psychiatric residential facilities run by the Veterans Administrations is not specified, nor is the number or characteristics of residential facilities providing long-term care. There have been no other recent studies providing additional information on this very important issue.

Size represents a critical variable for any taxonomy of these settings, probably the single most important variable. Small residences help create a homely environment which is in huge contrast to the large institutional environments of the past, warehousing hundreds of patients. It is clearly economically unfeasible, however, in an era of reduced costs devoted to health care, to envisage a system based on a large number of very small (e.g. three to four beds) facilities, with intensive support.

Staffing in the Residential Facilities

In the Italian PROGRES national study, the 1370 residential facilities employed 18 666 professionals, of whom 60% (n=11 240) worked full time. Around 40% of staff had no specific professional qualification for working with psychiatric patients with severe mental disorders. The overall ratio of patients to full-time staff was 1.42 to 1 (range 0.82-22.3). In the Veterans Administrations study, the direct care staff-patient ratio was 0.24,^[10*] but these facilities were providing acute care and therefore this ratio may not apply to long-term care. Large percentages of poorly qualified staff were also found in the UK 'Mental Health Residential Care Study'^[7**] and in a large survey done in the US.^[11]

Staffing is another area needing clarification both in quantitative terms (e.g. how many staff are needed in different typologies of facilities) and in qualitative terms (e.g. staff qualification and training, tasks and roles of different professionals, turnover rate, etc.). Effectiveness trials have shown that a larger number of staff is not automatically associated with better outcomes (e.g. PRISM Psychosis Study).^[12**] Indeed, within community-oriented models of care, the quality of staff (and what they do) is more important than the quantity, provided that a 'minimum' quantity is ensured. Unfortunately, in residential care (as well as in other settings) we do not know what the minimum is; in other words, what is the threshold below which there will certainly be a deterioration in the quality of care and in selected outcome indicators?

The Physical Environment

In the years covered by this review (2000-2004) no studies have assessed the physical environment of residential facilities and its relationship with the satisfaction of residents (and staff). This is surprising, because one of the strongest (and well founded) criticisms of old-fashioned institutions referred to their dehumanizing, alienating physical environment. It would be important to identify the key variables which facilitate (or are strictly needed for) the creation of a home-like, pleasant physical environment for long-term residents. In particular, several studies have highlighted that most residents attribute great importance to privacy (virtually nonexistent in the former mental hospitals), and this has precise implications in terms of architectural features (e.g. availability of single rooms, private bathrooms, etc.).^[13,14]

Residents' Characteristics

In the years covered by this review, a few studies have investigated sociodemographic, clinical and treatment characteristics of residents. In the Italian PROGRES study,^[4**] the authors found 15 943 residents in the 1370 residential facilities, that is 11.6 residents per facility. The majority of residents (58.5%) had never been admitted to a mental hospital or a forensic mental hospital, although in this group all had been admitted to a general hospital psychiatric ward. Two-thirds of the residents had a diagnosis of schizophrenia, with mental retardation being the primary problem in around 10% of residents. Most residents (82.7%) had no current problems of alcohol or substance abuse. This low rate of dual diagnosis, however, may be due to intake screening that excluded residents with either alcohol or drug problems in more than half (59.1%) of the residences. While patients with substance abuse can be treated in a variety of residential facilities, Italian data seem to point to a shortage of settings treating patients with a dual diagnosis, which may also apply to other countries.

By contrast, in the Veterans Administrations survey of 68 residential facilities providing acute care (officially designed for a population of psychiatric patients), Timko et al.^[10*] found that only 20% of patients admitted every month had just a psychiatric diagnosis: the largest proportion (52%) had a dual diagnosis and the remaining (28%) only a substance abuse diagnosis.

In a German study, 244 patients with schizophrenia in residential care were studied. Eighty-nine of these patients were living in residential facilities of varying size and degree of cover. This study found, as predicted, that patients in residential facilities with intensive cover exhibited the highest numbers of areas of need.^[15] Another survey examined 251 patients with schizophrenia aged 40-97 years. Ninety-nine patients were living in residential facilities in the San Diego county and the rest were living independently. Auslander et al.^[16] showed that assisted living was associated with heightened characteristics of the disorder (e.g. earlier onset, longer duration, increased negative symptoms and cognitive impairment), lower probability of having ever been married and a poorer subjective well-being. Residential facilities host a selected population within the overall population cared for by mental health services, made up of patients with the most severe and disabling disorders (mostly psychotic): targets, programmes, and activities run in residential facilities should be clearly based on this assumption and be tied to the specific needs and demands of the residents.

 

An important area of investigation is the process of care, which can help identify the characteristics which residential facilities should have in order to maximize the likelihood of positive outcomes and improve the QOL of residents (as well as that of staff and relatives). Very few studies on the process of care have recently been conducted in residential settings.

Holmqvist^[17*] assessed staff feelings toward patients and treatment outcome after 5 years at 23 residential facilities in Sweden. These residential facilities housed 111 patients (mean age 29.3 years), of whom 52% had a diagnosis of schizophrenia. Outcome was measured with Kernberg's Structural Interview, while staff's feelings were evaluated with a specific feeling checklist administered twice a year. At the end of the survey, a total of 4568 checklists were assessed. In the whole group, there were no correlations between the general staff feelings and outcome. Interesting results were found, however, when data for psychotic and borderline patients were analysed separately. For psychotic patients, the absence of negative feelings, rather than the presence of positive feelings, was associated with a positive outcome. By contrast, for borderline patients, a positive outcome was associated with negative feelings at the beginning of treatment, followed by strong positive feelings later on.

In another study carried out in nine small residential facilities in Sweden, 33 residents (mean age 41 years) and 50 staff members filled in the Community Oriented Program Environmental scales,^[18,19] designed by Moos to assess community-based residential settings. According to the results of this study, the social environment of residential facilities should include high levels of social interaction, a supportive profile with organization, programme clarity, and a low level of staff control, anger and aggression. Both studies highlight the importance of a key process variable, that is staff characteristics. In this area, the expressed emotion construct and its application to the study of staff-patient relationships may prove particularly useful to chart the emotional climate between clients and professional caregivers in a reliable and analytical way.^[20*]

Development of Assessment Instruments

An important area of improvement for the quality of residential care is represented by the development of specific assessment instruments tailored to the specific features and needs of long-term residents. The Resident Assessment Instrument-Mental Health is a comprehensive, multidisciplinary mental health assessment system for use with adults in various types of facilities, including long-term residential facilities. It evaluates psychiatric, social, environmental and medical issues at entry in a residential programme, emphasizing patient functioning. The instrument includes 166 items covering 28 different areas.^[21] The inter-rater reliability and convergent validity of the instrument have been assessed in a sample of 261 patients, of whom 28% resided in residential facilities^[22]; almost all domain areas obtained average kappa values in excess of the 0.40 cut-off established for acceptable reliability.

Do We Know How to Effectively Run Residential Facilities?

In general, the few studies published in the years 2000-2004 and focusing on the process of care point to a very important issue: how to effectively run residential facilities. This includes the functional characteristics of such settings (also in terms of rules and procedures) associated with positive outcomes, and the staff's prevailing feelings and attitudes and how they should be managed in order to maximize residents' outcomes. If we want that these settings do not become simply custodial institutions, we need evidence-based guidelines on the best staff management strategies (e.g. the ratio of individual patient-centred work to group-based interventions, optimal use of staff's time in terms of meetings, educational activities, etc). So far, all these strategies have been almost entirely based on subjective views of facility managers or of 'local' charismatic figures, whose dicta were rarely disputed.

Interestingly, while in the past there have been a number of studies exploring the relationships between residential facilities and the local neighbourhood,^[23,24] in recent years no studies have investigated this area, which is important in order to foster residents' integration and avoid the creation of self-segregating institutions.

The only recent study which has assessed certain aspects of residents' pharmacological treatment has analysed the association between medication adherence and a number of environmental and clinical variables. This took place in a small, convenient sample of 74 patients with schizophrenia living in four residential facilities in New York City.^[25] The authors found that lack of direct medication supervision, negative medication attitude and lower GAF score were associated with increased medication nonadherence in the recent past. They recommend adequate supervision in these settings in order to ensure medication compliance.

 

Under this heading we include not only a few longitudinal studies, but also different investigations exploring discharge rates and QOL of residents.

Longitudinal Studies

The most important longitudinal study providing data about patients resettled in residential facilities is the Team for the Assessment of Psychiatric Services (TAPS) project, carried out in the greater London area and assessing the 5-year outcome of a representative cohort of 523 patients discharged from two mental hospitals to a variety of community residences (specific information about number of residents living in various types of residential facilities is not given).^[26**] Although there was no change in the patients' clinical state or in their problems of social behaviour, they gained domestic, community living skills and more freedom than in the previous hospital setting, they acquired friends, and they wanted to remain in their current state. In the framework of the same project, a specific analysis has recently focused on a subgroup of 61 long-stay psychiatric inpatients, initially regarded as unsuitable for community placement and followed up for 5 years.^[27*] Substantial improvements occurred in clinical and social functioning which enabled 29 patients (40% of the study group) to be resettled in various residential facilities, gaining better access to community amenities and living more independently.

Other follow-up studies have consistently shown good community tenure in most patients resettled in residential facilities, although clinical improvements were often of limited magnitude.^[28-33]

Violence risks of severely mentally ill patients may represent a significant problem for community resettlement. A prospective study has analysed assaultive behaviour, in the course of 6 years, directed against staff by 554 male and 472 female residents of several residential facilities, originally discharged to these facilities from Massachusetts state hospitals in the early 1990s.^[34] Observed rates of assault were found to decline by 61%; men and women had similar rates of aggression. Such a result is consistent with figures obtained in the recent TAPS study,^[26**] and seem to show that violence and aggressiveness may settle down in most residents as they become more and more acquainted with their place of residence and feel comfortable in their living environment.

Qualitative Studies

While there has been much qualitative research carried out in traditional mental hospitals (e.g. Goffman's pioneering work), qualitative studies conducted in residential facilities have been very sparse.^[35,36] Such work, however, may represent (if carried out with appropriate, rigorous methodology) an important tool to thoroughly assess residents' daily life, their needs, expectations, concerns and habits.

Residents' Turnover: Rehabilitation Centres or 'Homes for Life'?

Current data on residents' turnover show that discharge to independent living represents an infrequent event. In the Italian PROGRES study,^[4**] during the course of an entire year (1999), more than a third of residential facilities (37.7%) did not discharge any patients and 31.5% discharged only one or two patients.^[4**] Also in the 5-year TAPS follow-up study, very low rates of turnover were shown for patients discharged to community residences.^[24] In general, even when optimal treatment is provided to all long-term patients, a substantial proportion of nonresponders will remain and for many patients residential facilities will probably be 'homes for life'.^[37*]

The issue of discharge from residential facilities to independent living is closely related to the model of residential care adopted. In the past, a 'continuum' dimensional model has been proposed as the most appropriate for planning and managing residential facilities in the context of a wider system of mental health care.^[38,39] This model posits that there are different residential settings with various levels of support and restrictiveness: the most intensive treatment is offered in the most restrictive setting. The patient can move along the continuum, from more restrictive to more open environments, and patients' needs have to be matched to the most appropriate setting based on their level of functioning. Carling^[40**] has summarized the main criticisms to this model (referred to by him as the 'sheltered housing model'): (1) it often confuses housing and treatment needs, and this may lead to unnecessary dislocations through successive moves, because improvements in functioning often require a move to another setting; (2) alternative housing is often not available at precisely the time when the person's needs change; (3) acquiring skills for independent living in an artificial environment such as a residential facility may not ensure that those skills will be generalized and saved in another setting.

So far, no countries have shown that it is feasible to set up, on a nationwide scale, a system of residential care based on a continuum model. An alternative to the dimensional sheltered housing model may be represented by a categorical 'supported housing' model. Within this model there are, on the one hand, high-cover residential facilities and, on the other hand, residential facilities with limited cover, and a flexible system of on-site support can allow a temporary increase in the quantity of required help in less intensive settings when patients face a crisis. While in the continuum model the system provides fixed levels of supervision and patients are expected to move around as they get better (or worse), in the supported housing model, aside from residential facilities with 24 h support, the remaining facilities are organized according to flexible levels of supervision around 'ordinary' housing options, and staff are expected to move around according to the fluctuating needs of patients. For both models, a problem faced by the high-cover residential facilities may be the difficulty in providing individually centred, individually sensitive care. If mismatches occur between levels of dependency and levels of supervision (more often in the direction of too much supervision), there is the risk of decreasing functioning by overproviding, especially if the staff are undertrained.

An important implication of the supported housing model is that discharge to independent living should not necessarily be considered a (feasible) target. A substantial number of patients may remain indefinitely within the residential network, while other objectives (e.g. clinical stabilization, ensured compliance, QOL improvements, etc.) become the objective of the intervention. Longitudinal studies, using large, representative samples of residents, are needed in order to clarify the feasibility of this model on a large scale.

Quality of Life of Residents

The issue of residents' QOL is important for two reasons: first, as just noted, for many residents these settings represent 'homes for life'; second, several critics of deinstitutionalization have argued that the institutional change has sometimes been less substantial than advocated by many, or even only 'cosmetic', and that residential facilities might represent 'small asylums'. In the years 2000-2004 five studies have assessed residents' QOL, and they have consistently shown that 'subjective' QOL among patients living in residential facilities was generally better than subjective QOL of patients living in traditional hospital settings, and residents are generally satisfied with their residential status.^[41-45] These studies, however, have also shown that subjective and 'objective' QOL (that is, evaluated by external raters) are weakly correlated, and substantial improvements in living conditions may not be rapidly reflected in parallel improvements in subjective satisfaction.

 

Papers of particular interest, published within the annual period of review, have been highlighted as:

*of special interest
**of outstanding interest

1. Anderson RL, Lyons JS, West C. The prediction of mental health service use in residential care. Community Ment Health J 2001; 37:313-322.
2. Anderson RL, Lewis DA. Quality of life of persons with severe mental illness living in an intermediate care facility. J Clin Psychol 2000; 56:575-581.
3. Anderson RL, Lyons JS. Needs-based planning for persons with serious mental illness residing in intermediate care facilities. J Behav Health Serv Res 2001; 28:104-110.
4. de Girolamo G, Picardi A, Micciolo R, et al. Residential care in Italy: a national survey of non-hospital facilities. Br J Psychiatry 2002; 181:220-225.
   ** The first survey of all RFs active in an entire country (Italy), giving details about staffing levels, characteristics of facilities and rates of residential places in different areas.
5. Becker T, Hulsmann S, Knudsen HC, et al. European psychiatric services: inputs linked to outcome domains and needs. Provision of services for people with schizophrenia in five European regions. Soc Psychiatry Psychiatr Epidemiol 2002; 37:465-474.
   * An analysis of all in- and outpatient services available for people with schizophrenia in five European Cities, carried out with the European Service Mapping Schedule.
6. Brieger P, Wetzig F, Bocker FM. Institutions and services of psychiatric care in Saxony-Anhalt: assessment with the European Services Mapping Schedule. Eur Psychiatry 2003; 18:145-147.
7. Lelliott P, Audini B, Knapp M, et al. The Mental Health Residential Care Study: classification of facilities and description of residents. Br J Psychiatry 1996; 169:139-147.
   ** A detailed survey of all residential (acute and long-term) places in 8 districts surrounding London,with details about staffing,facility characteristics and patients' populations.
8. Faulkner A, Field V, Lindesay J. Residential care provision in mental health: the current picture and future uncertainties. J Ment Health 1993; 2:57-64.
9. Shepherd G, Murray A. Residential care. In: G Thornicroft, G Szmukler, editors. Textbook of community psychiatry. Oxford: Oxford University Press; 2001. pp. 309-320.
   ** A nice overview of the main problems and prospects of residential care in the context of a community-based mental health system of care.
10. Timko C, Lesar M, Calvi NJ, et al. Trends in acute mental health care: comparing psychiatric and substance abuse treatment programs. J Behav Health Serv Res 2003; 30:145-160.
    * A large survey of facilities for acute psychiatric and substance abuse patients within the US Veterans Administrations system of mental health care.
11. Randolph FL, Ridgway P, Carling PJ. Residential programs for persons with severe mental illness: a nationwide survey of state-affiliated agencies. Hosp Community Psychiatry 1991; 42:1111-1115.
12. Thornicroft G, Wykes T, Holloway F, et al. From efficacy to effectiveness in community mental health services. PRISM Psychosis Study 10. Br J Psychiat 1998; 173:423-427.
    ** A large experimental trial comparing intensive and standard community treatment for patients with psychosis and addressing important service issues.
13. Johnson LC. The community/privacy trade-off in supportive housing: consumer/survivor preferences. Can J Commun Ment Health 2001; 20:123-133.
14. Timko C. Physical characteristics of residential psychiatric and substance abuse programs: organizational determinants and patients outcomes. Am J Community Psychol 1996; 24:173-192.
15. Leisse M, Kallert TW. Normative needs for community-based psychiatric care of patients with schizophrenia in different residential settings. Psychiatry Res 2003; 118:223-233.
16. Auslander LA, Lindamer LL, Delapena J, et al. A comparison of community-dwelling older schizophrenia patients by residential status. Acta Psychiatr Scand 2001; 103:380-386.
17. Holmqvist R. Associations between staff feelings toward patients and treatment outcome at psychiatric treatment homes. J Nerv Ment Dis 2000; 188:366-371.
    * An interesting analysis of staff feelings and attitudes towards RF patients, and their correlations with patients' outcome.
18. Brunt D, Hansson L. A comparison of the psychosocial environment of two types of residences for persons with severe mental illness: small congregate community residences and psychiatric inpatient settings. Int J Soc Psychiatry 2002; 48:243-252.
19. Brunt D, Hansson L. Characteristics of the social environment of small group homes for individuals with severe mental illness. Nord Psykiatr Tidsskr 2000; 56:39-46.
20. Van Audenhove C, Van Humbeeck G. Expressed emotion in professional relationships. Curr Opin Psychiatry 2003; 16:431-435.
    * A brief review of the main topics related to the use of the expressed emotion model for the study of staff-patient relationships.
21. Hirdes JP, Marhaba M, Smith TF, et al. Development of the Resident Assessment Instrument-Mental Health (RAI-MH). Hosp Q 2000-2001; 4:44-51.
22. Hirdes JP, Smith TF, Rabinowitz T. The Resident Assessment Instrument-Mental Health (RAI-MH): inter-rater reliability and convergent validity. J Behav Health Serv Res 2002; 29:419-432.
23. Aubry TD, Tefft B, Currie RF. Public attitudes and intentions regarding tenants of community mental health residences who are neighbours. Community Ment Health J 1995; 31:39-52.
24. Cook JR. Interactions between group homes and neighbours: neighbour preferences. J Behav Health Serv Res 1998; 25:425-436.
25. Grunebaum MF, Weiden PJ, Olfson M. Medication supervision and adherence of persons with psychotic disorders in residential treatment settings: a pilot study. J Clin Psychiatry 2001; 62:394-399.
26. Leff J, Trieman N. Long-stay patients discharged from psychiatric hospitals. Br J Psychiat 2000; 176:217-223.
    ** A landmark follow-up study of a large cohort of patients discharged from two mental hospitals in the London area and followed-up in the community, with an indepth evaluation of their clinical and psychosocial outcomes.
27. Trieman N, Leff J. Long-term outcome of long-stay psychiatric in-patients considered unsuitable to live in the community. TAPS Project 44. Br J Psychiat 2002; 181:428-432.
    * In the framework of the TAPS follow-up study, a detailed analysis of the specific problems raised by long-stay patients unable to live in the community.
28. Duurkoop P, van Dyck R. From a 'state mental hospital' to new homes in the city: longitudinal research into the use of intramural facilities by long-stay care-dependent psychiatric clients in Amsterdam. Community Ment Health J 2003; 39:77-92.
29. Hobbs C, Newton L, Tennant C, et al. Deinstitutionalization for long-term mental illness: a 6-year evaluation. Aust N Z J Psychiatry 2002; 36:60-66.
30. Hobbs C, Tennant C, Rosen A, et al. Deinstitutionalisation for long-term mental illness: a 2-year clinical evaluation. Aust N Z J Psychiatry 2000; 34:476-483.
31. Barbato A, D'Avanzo B, Rocca G, et al. A study of long-stay patients resettled in the community after closure of a psychiatric hospital in Italy. Psychiatr Serv 2004; 55:67-70.
32. Guazzelli M, Palagini L, Giuntoli L, et al. Rehab rounds: outcomes of patients with schizophrenia in a family-style, residential, community-based program in Italy. Psychiatr Serv 2000; 51:1113-1135.
33. King C, Singh K, Shepherd G. An analysis of process and outcomes for new long-stay patients in a 'ward-in-a-house'. J Ment Health 2000; 9:179-191.
34. Flannery RB, Fisher W, Walker A, et al. Assaults on staff by psychiatric patients in community residences. Psychiatr Serv 2000; 51:111-113.
35. Humberstone V. The experiences of people with schizophrenia living in supported accommodation: a qualitative study using grounded theory methodology. Aust N Z J Psychiatry 2002; 36:367-372.
36. Wilton RD. Poverty and mental health: a qualitative study of residential care facility tenants. Community Ment Health J 2003; 39:139-156.
37. Trieman N, Smith HE, Kendal R, et al. The TAPS project 41: homes for life? Residential stability five years after hospital discharge. Community Ment Health J 1998; 34:407-417.
    * A study of residential stability and discharge rates in a sub-sample of the TAPS cohort, showing low discharge rates to independent living.
38. Drachman D. A residential continuum for the chronically mentally ill: a Markov probability model. Eval Health Prof 1981; 4:93-104.
39. Wing JK, Furlong R. A haven for the severely disabled within the context of a comprehensive psychiatric community service. Br J Psychiat 1986; 149:499-457.
40. Carling PJ. Housing and supports for persons with mental illness: emerging approaches to research and practice. Hosp Community Psychiatry 1993; 44:439-449.
    ** A stimulating critique to different models of residential care.
41. Hanrahan P, Luchins DJ, Savage C, et al. Housing satisfaction and service use by mentally ill persons in community integrated living arrangements. Psychiatr Serv 2001; 52:1206-1209.
42. Mares AS, Young AS, Mcguire JF. Residential environment and quality of life among seriously mentally ill residents of board and care homes. Community Ment Health J 2002; 38:447-458.
43. Horan ME, Muller JJ, Winocur S. Quality of life in boarding houses and hostels: a residents' perspective. Community Ment Health J 2001; 37:323-334.
44. Leisse M, Kallert TW. Social integration and the quality of life of schizophrenic patients in different types of complementary care. Eur Psychiatry 2000; 15:450-460.
45. Chan GW, Ungvari GS, Shek DT, et al. Hospital and community-based care for patients with chronic schizophrenia in Hong Kong: quality of life and its correlates. Soc Psychiatry Psychiatr Epidemiol 2003; 38:196-203.

Acknowledgements
 

Special thanks are due to the following colleagues who provided invaluable help in the preparation of Services Research and Outcome section: Giorgio Bignami, Len Bowers, Danilo Di Diodoro, Graham Dunn, Sue Estroff, Ian Falloon, Angelo Fioritti, Francesco Grossi, Matt Muijen, Angelo Picardi, Diana Rose, Giovanni Santone, Geoff Shepherd, Graham Thornicroft, Richard Warner.

Reprint Address
 

Correspondence to Giovanni de Girolamo M.D., Department of Mental Health Viale Pepoli 5, 40123 Bologna, Italy Tel: +39-0516584204; fax: +39-0516584244; e-mail: gdg@iss.it

Abbreviation Notes
 

QOL = quality of life; TAPS = Team for the Assessment of Psychiatric Services

Giovanni de Girolamo and Mariano Bassi, Department of Mental Health, Viale Pepoli 5, 40123 Bologna, Italy