Admission of the homeless mentally ill in the UK

The size of the homeless population in the UK has grown rapidly over the past three decades while the proportion with severe mental illness has remained high (Scott, 1993). During the same period many large psychiatric institutions have closed. In most instances former long-stay patients have been successfully resettled (Leffet al, 1996). However, there is concern that some people with severe mental illness, including many who are homeless, have neither been adequately provided for in the community nor given sufficient access to increasingly limited hospital resources (Ritchie et al, 1994). Whatever the relationship between the two, the chronological juxtaposition of rising homelessness and falling bed numbers is likely to have had an impact on the provision of psychiatric inpatient care to homeless people. We take advantage of published data from a 1960s study of 'no fixed abode' admissions to institutions in

The size of the homeless population in the UK has grown rapidly over the past three decades while the proportion with severe mental illness has remained high (Scott, 1993).During the same period many large psychiatric institutions have closed.In most instances former long-stay patients have been successfully resettled (Leffet al, 1996).However, there is concern that some people with severe mental illness, including many who are homeless, have neither been adequately provided for in the community nor given sufficient access to increasingly limited hospital resources (Ritchie et al, 1994).
Whatever the relationship between the two, the chronological juxtaposition of rising homelessness and falling bed numbers is likely to have had an impact on the provision of psychiatric inpatient care to homeless people.We take advantage of published data from a 1960s study of 'no fixed abode' admissions to institutions in Birmingham (Berry & Orwin, 1966;Orwin & Berry, 1968) to compare them with those from a comparable survey undertaken in the same city more than 30 years later.We examine changes in the number of homeless people admitted, their demographic and clinical characteristics, refer ral routes to hospital and discharge and after care arrangements.

The study
In Birmingham, the admitting hospital for home less people requiring in-patient care is deter mined by a no fixed abode rota.At the time of the first study (1961 to 1964) there was a monthly rota between the four large mental hospitals in the city.The study focused on admissions to one of these hospitals during a four-year period (givingthe equivalent of one year of admissions).By 1995-1996 separate rotas existed for the north (with four in-patient units) and south (with one in-patient unit) parts of Birmingham.Also, a city-wide specialist community mental health team (CMHT) for homeless people had been established (Commander et al, 1997a).The second study covered no fixed abode admissions between 1 February 1995 and 31 January 1996.Demographic data as well as admission and discharge details were collected from ward staff and the case notes in both studies.

Findings
During the year 1995-1996, there were 106 admissions under the no fixed abode rota (91 people) compared with 145 (135 people) in the earlier study.Homeless admissions in 1995-1996 also represented a smaller proportion of total admissions to Birmingham hospitals, ac counting for 2.8% compared with 4.4% in 1964.
The majority of admissions in both surveys were men (Table 1) but this was significantly more so in the later study.The proportion of people aged under 30 years was also significantly higher in the 1995-1996 study and consistent with this finding, patients in the second study were significantly less likely to be married than their 1961-1964 counterparts.There was no significant difference between the two samples according to ethnicity, the majority being White-British.However, there was a fall in the propor tion identified as Irish and a corresponding increase in people from other ethnic minority groups (see Table 1).
Previous admissions to psychiatric hospital were equally common in the two samples and schizophrenic disorders remained the most fre quent primary diagnoses (see Table 1).When subsidiary diagnoses were also considered, 16% (n=23) of the 1995-1996 sample were identified as having an alcohol use disorder while 29% (n=31) of the 1961-1964 sample were either labelled alcoholic or described as drinking excessively.In contrast, drug use disorders were twice as common in 1995-1996 compared with the earlier survey (20% (n=21) v. 10% (n=15)).
General hospitals and the police remained the main sources of admissions in both surveys (see Table 2).The proportion of prison referrals was considerably lower in 1995-1996, a substantial number of admissions coming instead from hostel staff (21% (n=22)).Although the proportion of involuntary admissions remained constant (at a level more than five times the national average.Wing, 1994), there was a significant reduction in length of stay for the 1995-1996 compared to the 1961-1964 sample (see Table 2).Also, a sub stantial minority in both surveys were discharged against medical advice.However, the proportion receiving no aftercare had fallen by the time of the second study.This was mainly attributable to the CMHT for homeless people who followed up over one-third of patients (37% (n=39)).

Comment
In keeping with the results from recent surveys of single homeless people in the UK (Anderson et al, 1993), younger men were found to predominate among homeless admissions.People from ethnic minority groups were also consistently overrepresented compared to the local population, although the number of Black and Asian people had increased while the proportion of Irish people had fallen over the intervening three decades.Schizophrenia remained the most common diagnosis but the greater frequency of substance use disorders in the 1995-1996 sample reinforces concerns about escalating drug misuse, especially among younger people who are homeless (Scott, 1993).
Many homeless people are not registered with a general practitioner and gain access to primary care through accident and emergency depart ments (Royal College of Physicians, 1994).This was reflected in the high proportion of referrals for admission coming from general hospitals and the low involvement of general practitioners in both eras.The improved liaison with psychiatric services stimulated by the CMHT for homeless people (Commander et al, 1997a) probably contributed to the increase in referrals from hostel staff.Although the fall in numbers coming from prison may be due to the success of forensic diversion schemes, there was no off-set in admissions from the police or courts and the decrease is as likely to derive from the difficulties in getting mentally ill prisoners transferred to appropriate in-patient facilities (Coid, 1991).
In stark contrast to the growth in size of the homeless population in Birmingham (Birming ham City Council, 1993), the number of home less people admitted to psychiatric hospital has fallen over the past few decades.While it is possible that this finding stems from methodo logical limitations, for example the eligibility criteria for the no fixed abode rota varying between the two time periods (Cowan & MacMillan, 1996), other explanations warrant consid eration.In the absence of any improvement in the mental health of Birmingham's homeless population (Commander et al, 1997c) one alter native is that innovations in community care have reduced the need for hospital provision.However, the fact that there was a fall in no fixed abode admissions not only in absolute terms (as might have been expected given reduced bed numbers) but as a proportion of total admissions argues against this.It is improbable that com munity care initiatives have been more effective in reducing the need for hospital admission in homeless compared with residentially stable populations.Research evaluating the impact of specialist homeless teams suggests quite the reverse, the introduction of dedicated services increasing access to in-patient care (CatÃ³n et al 1990;Commander et al 1997b).
Taken in conjunction with the substantial unmet needs identified in community surveys (Scott, 1993), the finding that no fixed abode admissions have fallen over the past 30 years confirms the view that psychiatric services are currently failing the homeless population.The changes in admission paths and the improve ments in aftercare highlight, yet again, the value of dedicated outreach teams targeting mentally ill homeless people.However, deficits in the provision of in-patient care must also be ad dressed if a comprehensive service is to be successfully delivered to this neglected section of the population.

Table 1 .
Demographic and clinical characteristics

Table 2 .
Admission and aftercare details