Is there a crisis about crisis houses?

The search for alternatives to psychiatric inpatient admission has driven several projects in recent years. Crisis houses across the UK are one such alternative to inpatient admission for some people with acute mental health problems. They may also serve to de-stigmatize the experience of inpatient psychiatric admissions. This article identifies the heterogeneity of crisis house models that exist and examines whether they offer greater value than inpatient care.

The publication of the ‘No health without mental health’ implementation framework in July 2012 advocated treating people with mental health problems in the least restrictive therapeutic setting and providing choice. The Schizophrenia Commission 2012 reviewed the delivery of services for the 220,000 people in England with psychotic experiences. ¹ This report recommended a radical overhaul of acute mental healthcare units, with greater use of alternatives to admission such as recovery houses to better manage the transition between hospital and community services.

There is currently a nationwide drive to reduce and consolidate inpatient psychiatric beds across the NHS. This has led to a corresponding growth in alternative levels of intensive care in community settings, including an increase in the number of crisis houses over the last decade.

A mental health crisis may be related to suicidal behaviour or intent, panic attacks or extreme anxiety, psychotic episodes or other behaviour that seems out of control or irrational and is likely to endanger the service user or others. ² Crisis houses provide a short-term alternative to hospital admission for those with mental illnesses who are experiencing a crisis and may be used to facilitate their early discharge from hospital. ³ They have also been seen as a response to dissatisfaction among service users and carers with the quality of care provided on acute psychiatric wards.

From our discussions with crisis house staff, we discovered that a variety of approaches for crisis houses have been adopted. Most admit a small number of patients of both sexes. Although there is close collaboration with acute wards, crisis houses are often located away from inpatient facilities within residential areas. Some crisis houses share buildings with and work alongside crisis teams, who may act as gatekeepers when admission is considered. In some models, service users can refer themselves directly to the crisis house. Exclusion criteria include patients who are detained under the Mental Health Act.

Patients in crisis houses have a range of mental health problems including schizophrenia, bipolar disorder and moderate depressive illnesses. Some medical, managerial and housing colleagues consulted voiced concerns about the potential risks of managing people with emotionally unstable personality disorders in such an environment. This is because most crisis houses are staffed by support workers rather than qualified nursing or medical staff. This has implications for risk assessment and management. In some – but not all – crisis houses, admissions and discharges are discussed with a Consultant Psychiatrist, who holds ultimate clinical responsibility for those admitted.

We found variability in the anticipated length of stay for users of crisis services. Some crisis houses encourage people to move on after several nights, while others accommodate them for up to four weeks.

Concerns were also voiced about the cost-effectiveness of crisis houses. For one crisis house we examined, it emerged that the cost exceeded that of inpatient admission and therefore was not economically sustainable. This may have been due to the longer length of stay caused by referrals of service users with housing problems.

One of the primary arguments for the provision of crisis houses is to facilitate early discharges and prevent unnecessary admissions. However, this may not be borne out in practice and a more detailed financial analysis is required. Similarly, a crisis house could be more dangerous were it to admit patients who require medical input as an inpatient or those at high risk of suicide (given the staffing of crisis houses with non-qualified staff). Here, we reference two studies that have evaluated patient satisfaction and cost-effectiveness of crisis house models.

The first compared the experiences of 314 patients in standard inpatient settings to those in alternative residential care, namely two crisis houses (one run non-clinically), crisis team beds and a therapeutic ward setting. ⁴ Using a variety of satisfaction questionnaires and controlling for detention under the Mental Health Act (an exclusion criterion for admission to most crisis houses), the following factors were significantly greater in non-hospital settings: patient satisfaction; patient autonomy and voice. There was no significant difference in the perceived levels of support provided, patient involvement or staff control.

Second, a study of 102 patients, with no significant baseline demographic differences, admitted to either a women’s crisis house or acute ward (combined randomized controlled and patient preference trial) found no differences in outcomes or costs between inpatient and crisis services as well as higher levels of satisfaction among those admitted to the crisis house. ⁵

The authors undertook a retrospective utilization review which aimed to identify patterns of how inpatient psychiatric care was actually provided for individual patients and consider whether alternative settings could have been appropriate. Fourteen NHS specialty registrars in psychiatry participated in the review by accessing 750 randomly selected patient records of people with mental illness who were admitted to a psychiatric ward between 2009 and 2011 within a defined geographical area of South East England. ⁶ To reflect the prevalence of conditions within this locality, the study was weighted to include patients from the following three broad categories: older adults and dementia; adults including mood, anxiety and personality; and psychosis. A bespoke tool was designed to consistently capture criteria for medical necessity on a weekly basis for the duration of each inpatient or home treatment team admission. This included demographic data, details on the current admission and treatment plan, history of inpatient treatment, mental state including risk assessment and the presence of social support.

The review showed that one in four inpatients were not documented to be seen by a Consultant Psychiatrist in the first seven days of admission. It also identified poor communication between inpatient, community and primary care services.

The reviewers determined that, on admission, over 50% of inpatients could have been treated effectively in a community setting such as a crisis house if appropriate services had been available.

Thus, there appears to be a role for crisis houses in the management of acute mental health problems, with potential advantages compared to other care models, such as improved user satisfaction. However, wide variety currently exists in the models of care offered and there is a need to develop and share a best practice framework. Furthermore, cost-effectiveness data remain limited and there are unresolved issues relating to risk management.

In summary, as the NHS seeks to do more with less, at first glance crisis houses appear to offer a value-based alternative to inpatient admissions for patients with mental health problems in acute crisis and a mechanism for speeding up inpatient discharges. They may also play a role in de-stigmatizing the experience of inpatient mental healthcare with higher levels of user satisfaction compared to inpatient care. However, these theoretical models are not yet supported by the available data. We propose that a more robust evaluation of the crisis house model is needed to better understand which patient groups to target, how to staff them appropriately and what outcomes to measure. This will inform future evidence-based models of mental healthcare delivery to optimally serve people with mental health problems.

Figure 1.

Frequency of inpatient review by a Consultant Psychiatrist. ⁶

Figure 2.

Number of cases per week with documented proactive discharge plan. ⁶