The growth in the use of alternatives to psychiatric hospital care in Britain in recent years is an important trend which, one imagines, has been encouraged by the recovery movement. The recovery model has emphasised the importance of patient empowerment and interpersonal support, ^{Reference Warner1} and the articles in this supplement demonstrate that, in line with this focus, reduced paternalism and coercion and increased peer support are benefits of the hospital alternative models. These settings provide a different treatment atmosphere with more autonomy for staff and residents. Their greater emphasis on human interaction rather than medication, and improved user satisfaction are apparent from the papers in this supplement.

Advantages of alternative care

It is useful to remember that the themes common to these settings and to the recovery movement have a long-standing tradition in British psychiatry which can be traced back to early 19th-century moral treatment. The York Retreat was a transformative, domestic, non-coercive alternative to the institutional care of the time. ‘It is not at all the idea of a prison that it suggests, but rather that of a large farm,’ wrote a contemporary Swiss visitor, ‘No bars, no grilles on the windows.’ ^{Reference Warner2} Patients were treated with respect and were expected to exercise self-control and to participate in all usual social activities. The comparison is not an idle one, because some of our modern alternative settings embody many of the active ingredients of moral treatment. British crisis houses are in a tradition of small, domestic-style, normalising facilities that are open-door and genuinely part of the community, allowing residents to stay in touch with their friends, relatives, work and social life. ^{Reference Warner and Wollesen3} Such hospital alternatives are more flexible and non-coercive and often based more on peer relationships than on hierarchical power structures. They can offer opportunities for residents to be involved in the operation of the treatment environment, although the extent to which this happens in British crisis houses is in practice variable. If costs can be kept lower than those of hospital care, the pace of treatment in the alternative setting may not need to be so rapid and it becomes more possible to offer a quieter form of genuine asylum.

As in moral management, treating people with respect in a normalising and domestic setting leads them to exercise ‘moral restraint’ or self-control over their impulses. In someone's home, one feels obliged to treat other people and property with consideration, but in an institution, anything goes. This observation allows us to understand why it is possible to care for compulsorily detained patients in these open-door settings. If the alternative setting is more attractive to the patient than a hospital unit, then the patient is likely to call upon reserves of self-control in order to be allowed stay there rather than in hospital.

If the US experience is any value as a guide, a future stimulus to the diffusion of alternative settings in Britain may be the relative shortage of in-patient beds resulting from the closure of stand-alone British psychiatric hospitals. In the USA this shortage is more severe, having been exaggerated by the closure of psychiatric units in private general hospitals because of their poor profitability when compared with such sectors as cardiac surgery. There may come a point in the UK, however, when residential alternatives fill a need for patients who would traditionally have been admitted to hospital care but who cannot gain access to a bed. Another material factor that might make these alternative settings welcome in the future is their potential for cost-effectiveness. At present the cost-effectiveness of British alternative settings may be in doubt, but this factor is heavily influenced by facility size (larger facilities have lower per capita costs) and patient selection (less severely ill patients will show smaller clinical gains).

Alternative care in North America

Cedar House (recently renamed Warner House), a 15-bed hospital alternative that has been in operation for 30 years in the public mental health system in Boulder, Colorado, illustrates some of these points. The county mental health system has found that the facility can accommodate at least half of the catchment area patients in need of acute in-patient care at any point in time, including many patients requiring compulsory treatment. ^{Reference Warner and Wollesen3} Costing half as much as hospital care (which is purchased by the mental health system at the best price in the marketplace), there has never been any question as to its cost-effectiveness. Had it been much smaller than its current size, say, approaching the eight-bed average size of the British hospital alternatives, ^{Reference Johnson, Gilburt, Lloyd-Evans, Osborn, Boardman and Reese4} then the per capita cost would have been much greater and its survival in jeopardy. In other ways the clinical and specialist crisis house models identified in the Alternatives Study national survey are similar to Cedar House, in that these services provide a fairly extended period of care (mean 38 days), have staff awake at night and, in some cases, accept compulsory admissions. ^{Reference Johnson, Gilburt, Lloyd-Evans, Osborn, Boardman and Reese4} A substantial proportion of patients, in fact, are compulsorily admitted (16%) and most (60%) have symptoms of psychosis. It appears that these crisis houses serve a similar function to the public-system hospital alternative in Boulder, but would need to be larger to survive in a competitive health marketplace.

The design of many of the alternative programmes in the public sector in the USA and Canada is driven more by financial considerations and service efficiency than by principles of recovery or social intervention. ^{Reference Warner5} One of the important variables determining style of working and individualisation of treatment has already been referred to – size. The drive for cost-efficiency leads to larger capacity, whereas a quieter and more personalised healing environment requires a smaller size. Progress Foundation in San Francisco, which accommodates public-sector patients, has been able to limit the size of each facility to 8–10 residents, ^{Reference Fields6} but this is unusual and is due to its selection of people who are less severely ill. Cedar House, in Boulder, struck the balance at 15 beds. Venture, in Vancouver, British Columbia, is substantially larger at 20 beds and the operators concede that this leads to a less home-like quality of the environment. ^{Reference Sladen-Dew, Young, Parfitt and Hamilton7} Private acute care facilities, such as Balsam House in Boulder, Colorado, and Crossing Place in Washington, DC, ^{Reference Bourgeois8} each with 8 beds, tend to be smaller, and the higher per capita cost is passed on to consumers and their families. Private-sector residential alternatives in the USA that are geared more towards rehabilitation than acute treatment can function well with a larger group size. At the 100-year-old Gould Farm in Massachusetts, for example, 60 people or more sit down for meals at the end of each day of farm, bakery or restaurant work. Many residential facilities, especially those in the private sector, adhere to a mission to use the sense of community among residents as a component of the healing process, but there are other private- and public-sector models that aim to provide acute care to more isolated individuals. The public-sector mental health centre in Madison, Wisconsin, places people with acute psychiatric distress in family foster-homes in the community for short periods, providing psychiatric services through a mobile team of professionals. This model offers one of the lowest-cost hospital alternatives available. ^{Reference Bennett9} The private-sector Windhorse Program in Colorado and Massachusetts provides support and treatment to people with acute and subacute disorders in their own homes with live-in aides and a ‘wrap-around’ team of professionals. ^{Reference Fortune10}

Another important treatment variable is coercion. When the Colorado mental health system sought to create more acute treatment units like Boulder's Cedar House around the State, the model changed dramatically, from a non-coercive, domestic-style household to locked facilities with the capacity to use restraints and seclusion. This change was designed to meet the need to manage acutely disturbed and agitated patients in rural parts of the mountain state that were several hours' drive away from a psychiatric hospital unit. The public-sector Northwest Evaluation and Treatment Center in Seattle, Washington, in operation throughout the 1990s, which was designed as a crisis reception facility as well as a hospital alternative, was locked and, with 32 residents, large. New proposed acute treatment facilities in Portland, Oregon, and Seattle, Washington, will be locked but, at 16 beds, smaller. Michele Tansella points out that hospital alternatives can help us listen to patients better, provide more personalised care and be less paternalistic. ^{Reference Tansella11} We must recognise that some elements of hospital alternative care which make this possible – elements such as small size, open doors and peer staffing – can disappear in negotiations with planners, if clinical concerns are not properly included in the discussions.

Future perspectives

Moral treatment, which was developed as a non-coercive, human-scale approach to managing people with serious mental illness in small residential facilities, became transformed into its opposite when it was used to market the growth of large insane asylums in the early 1800s. We should be on guard to preserve the domestic, non-coercive qualities of British acute-care residential facilities as they diffuse across the country, while acknowledging that there may need to be some give and take on the question of size and cost-effectiveness. In this way, we can continue to draw upon the knowledge accrued in 200 years of social psychiatry.