Acute in-patient psychiatry: residential alternatives to hospital admission
Sonia Johnson, Helen Gilburt, Brynmor Lloyd-Evans, Mike Slade

There is a consensus that mental health in-patient services need improvement, and one potential approach is the development of residential alternatives in the community. We review published accounts of such alternatives and outline a new study that will describe and evaluate currently available services of this type.

Community residential crisis services

The lynchpin of current UK policy initiatives aimed at diverting patients from admission is the home treatment team or crisis resolution team model, in which patients' homes are the main setting where treatment is offered. However, home treatment is not always practical or desirable. Risk of harm to self or others might be too great for patients to be left for many hours without staff supervision, and crisis resolution teams might struggle to manage people who are severely neglecting themselves, for example by failing to eat and drink properly and to care for themselves and their environment in basic ways. Some service users have no home or live in very poor circumstances, and others may live in a home environment that exacerbates their difficulties (for example because of an abusive relationship). Another impediment to home treatment is that carers may feel unable to sustain their role in supporting someone at home.

Acute hospital admission is not necessarily required when one of these obstacles to home treatment is present, and community residential alternatives have the potential to address the needs of some people for whom home treatment is unsuitable. The availability of such alternatives also extends service user choice. In service systems in which reduction of acute bed use is a high priority, a danger is that unless they reach the high threshold for admission, service users have no option but to cope with crises at home, with brief contacts with a crisis resolution team just once or twice a day. This may be especially difficult for people who are socially isolated.

Several interesting types of residential alternative to hospital have been described over the past 40 years, although none has been well evaluated and no model seems to have been widely disseminated. Many of the model services described in the literature seem subsequently to have disappeared, and sustainability may be a significant issue. Three main categories of community residential alternative have been described

  • family sponsor homes

  • crisis houses

  • hybrid services (in which beds for management of crises are available within a community service such as a community mental health centre that has a variety of functions).

Family sponsor homes

Schemes involving placement of acutely unwell patients with carefully selected families have been described in several centres in the USA. One of the earliest published accounts relates to the family sponsor homes established by Paul Polak as part of an extensive network of community services that flourished briefly in Denver, Colorado in the 1970s. The families involved were carefully chosen and were supported by the home treatment service, which could be paged at any time. Families were paid a fee for receiving up to two patients in crisis as guests in their homes for a few weeks, and were encouraged to involve them in their households, for example through participating in chores and joining family meals (Polak et al, 1979).

A similar network of families was established by Leonard Stein and his colleagues in Madison, Wisconsin in the 1980s, and has continued to operate (Stein, 1991; Bennett, 2002). Stroul (1988) surveyed community-based residential crisis facilities in the USA and reported that such short-term housing and support at the homes of carefully selected families was the most widely available form of residential crisis care. No more recent reports have appeared regarding their availability, nor are evaluations available (apart from Polak's original investigation of the service system of which they formed part).

The only published description from the UK of a service of this type relates to the Accredited Accommodation Scheme in Powys, Wales, which aims to provide crisis care, although in practice it is often used for planned periods of respite and rehabilitative social care (Readhead et al, 2002).

Crisis houses

The residential alternative to admission which has been most prominent in the UK is the crisis house (Davies et al, 1994). This term tends to be applied to stand-alone community units that are usually based in converted residential premises. The care they offer seems to vary considerably in content and philosophy. A few descriptions and evaluations of such services have been published in the USA, although without very widespread adoption of this model. These studies have usually been small and have sometimes had substantial methodological difficulties, but results have suggested clinical and social outcomes at least as good as for standard in-patient care (Bedell & Ward, 1989; Bond et al, 1989; Fenton et al, 1998).

The US crisis residential services have varied considerably in the degree to which they adhere to conventional clinical practices and staffing patterns, or offer an alternative which is substantially different in philosophy and treatment approach from hospital services. Probably the best known more radical alternative is Loren Mosher's Soteria service, which operated from 1971 to 1983 and aimed to manage first-onset schizophrenia in a house in the community without recourse to antipsychotics (Mosher et al, 1975; Mosher, 1999). A randomised controlled trial, published more than two decades after its completion, suggested better or similar outcomes for Soteria compared with hospitalisation and lower subsequent use of antipsychotic medication, with 43% of Soteria residents identified as ‘drug-free responders’ who were well after 2 years without having received medication (Bola & Mosher, 2002). Despite these promising results, this model has been little replicated in the USA, although services influenced by Soteria have been established in Switzerland and Germany (Ciompi et al, 1992). There is some debate about whether all participants would meet diagnostic criteria for schizophrenia on rigorous assessment with a modern structured diagnostic instrument.

Other US crisis residential models adhere to more conventional clinical models, and a randomised controlled trial conducted by Fenton et al (1998) suggested that around two-thirds of patients destined for acute general hospital wards could be managed in a community residential alternative, with outcomes similar to hospital admission.

In the UK, investigation of the Drayton Park women's crisis house in Islington, a 24-hour staffed crisis house for women, suggested that most women managed at the house had a previous history of hospital admission, and that it was highly valued by service users (Killaspy et al, 2000; Johnson et al, 2004). They reported that their recovery was promoted by a home-like environment, absence of disturbed male patients, ready availability of staff for talking through current and past difficulties, and good support from other residents. Admission to the house was often experienced as less stigmatising than hospital. Some crisis houses have adopted a user-led model of care, as in the residential crisis facilities described in a report from the Mental Health Foundation (2002), which again reported that service users valued highly the availability of such alternatives.

Hybrid services

Another strategy for providing residential alternatives to in-patient care is to establish hybrid facilities that offer crisis admission alongside other types of community care. Probably the best known example is the community centres in Trieste which combine crisis beds with a comprehensive range of other services (Mezzina & Vidoni, 1995). Similar services have also been described elsewhere in Italy and in France (Katschnig et al, 1993).

In the UK, Boardman and colleagues (Boardman et al, 1999; Haycox et al, 1999) have investigated beds integrated into community mental health centres in North Staffordshire. Results suggested greater client satisfaction and better outcomes on some measures for the group managed in the community mental health centres than for hospital in-patients. Wesson & Walmsley (2001) have described a community-based unit in Southport that combines day care and crisis admission beds.

Alternatives Study

Although over the past 40 years sporadic descriptions and, occasionally, evaluations have appeared of a variety of types of community residential alternative to hospital, little is known about the clinical effectiveness of such services and whether they manage crises of similar severity to those currently resulting in hospital admission. In particular, the reviewed approaches were in use sometimes several decades ago, when service systems and in-patient populations are likely to have been different in many ways from now. There is little evidence about the extent to which they might act as genuine alternatives to admissions in modern mental health services.

Clear descriptions and operational definitions of the different types of residential service that offer an alternative to admission are also lacking, and the extent to which such services are available in the UK and elsewhere is unknown. The reasons why they have not been widely adopted or sustained also remain uncertain, especially as the very limited evidence available, much of it now outdated, has tended to suggest they are useful models, at least in the contexts in which they were originally implemented. Other potential reasons might include local funding considerations (if the traditional acute ward is viewed as a necessity and the alternative is a luxury), scepticism among clinicians about the feasibility of managing severe crises in settings other than a hospital, or the end of involvement from a ‘product champion’ who initiated and led the service.

The Alternatives Study is a national research programme that has been commissioned by the NHS Service Delivery and Organisation Programme in order to begin addressing the many unanswered questions relating to these services. The research is being conducted by a study team from the Institute of Psychiatry and University College London, led by the authors, and is due to be completed by November 2008. As well as community residential services, it encompasses innovative hospital services intended to provide acute care that is distinctly different from standard generic acute wards.

Box 1 summarises the main components of the study. Methods used will include investigation of the characteristics and outcomes of a consecutive cohort of users of the alternative services and of standard services in the same areas, and exploration of service users' views both through qualitative interviews at the alternative services and through a quantitative comparison between residents of the alternative and standard services.

The overall aim is to inform the development of meaningful and sustainable alternatives to acute in-patient services, a task for which a secure evidence base is currently lacking. Randomised trials in this area, although methodologically challenging to conduct, have potential to be very relevant to future service development.

Box 1. Alternatives Study

Stage 1

This is a national survey of in-patient alternatives to traditional acute psychiatric in-patient care for adults of working age in England. The aims are:

  • to identify all in-patient alternatives to traditional acute psychiatric in-patient care for adults of working age in England

  • to describe their organisational structures and operational methods and use these data to develop a typology of alternatives.

Stage 2

This involves identifying six alternative services, typifying each category in the typology and comparing them with matched standard acute in-patient psychiatric wards in order to:

  • examine the social and clinical features of people admitted to alternative services and compare them with those of people admitted to standard services

  • assess using a very simple set of indicators, the short- and medium-term clinical outcomes for people admitted to alternative and standard services

  • assess the cost and cost-effectiveness of alternative services

  • understand the experiences of people who use alternative services

  • understand the experiences of carers of people admitted to alternative services

  • assess and compare patients' satisfaction with alternative and standard services

  • investigate and compare the content of care in alternative and standard services.


The Alternatives Study is funded by the NHS Service Delivery and Organisation Programme.