Homeless people and psychiatric care.

SIR,-The last paragraph of Dr S L George and colleagues' paper on their census ofsingle homeless people in Sheffield gives the impression that large numbers of long stay patients in psychiatric hospitals have been discharged to live in "reception centres, in substandard rented accommodation, or on the street."' By contrast, we did not find that any of the long stay patients in hospital in the same city in 1982 were homeless at follow up eight years later.2 Evidence is accumulating that the outcome of resettling long stay psychiatric patients is perhaps not as bad as some campaigning organisations feared. For example, in New South Wales not one of 208 long stay patients discharged into supported accommodation had drifted to a refuge for the homeless at follow up at 21 months.' In a companion study, similar in design to the census in Sheffield, hospital records for a cohort of homeless men in a refuge showed that only three of the 22 residents with schizophrenia had had a prolonged admission to a psychiatric hospital.4 Admissions for the remaining 19 had been brief and frequent, but the duration of stay in hospital had not altered appreciably over the years. It would be interesting if psychiatric records for the people studied in the Sheffield census were available for comparison. They may show that the population in the census did not generally come from long stay wards. In evaluating evidence about homelessness and the fate of discharged psychiatric patients it is important to distinguish between the needs of long stay patients who have been discharged and the needs of people who may no longer be able to get psychiatric care. My concern is that the results of surveys of homeless people are being interpreted as showing that traditional psychiatric hospitals should be preserved.5 Government policy recognises the need for "asylum," but this can be arranged in smaller, well staffed units.