Collected and neglected: are Oxford hostels for the homeless filling up with disabled psychiatric patients?

OBJECTIVE To assess the severity of psychiatric symptoms among residents of hostels for homeless people. DESIGN Survey of residents in two hostels in Oxford, comprising three weeks of background fieldwork, a demographic questionnaire, and rating behaviour over two weeks with a behavioural rating scale (REHAB) and mental state with the brief psychiatric rating scale. SETTING Two hostels for homeless people in Oxford. SUBJECTS 146 Medium to long term residents, of whom 48 were selected by hostel workers by the following criteria: continuous residence for at least two months, signs of persistent severe mental disability, and difficulty in coping independently in the community. Two subjects died during the study; three (previously long term psychiatric inpatients) declined to be assessed on the psychiatric scale. MAIN OUTCOME MEASURE Behavioural disturbance and mental state. RESULTS Only a third of the total sample had been born in Oxfordshire. Subjects had been accepted into the hostel either by arrangement with the local psychiatric service (22) or straight off the streets (26); 43 had had a previous (non-drug related) psychiatric admission. Subjects were significantly more likely than other residents to have spent longer (greater than 80 weeks) in a hostel in the past three years (p less than 0.02). With reference to norms for deviant behaviour, the 46 subjects assessed showed considerable deviant behaviour (average weekly scores: 0 (11 subjects), 1 (14), 2-3 (16), and greater than or equal to 4 (5] not significantly different from that expected in moderately to severely handicapped psychiatric inpatients (chi 2 = 1.3, df = 3, p greater than 0.7); 22 had scores equivalent to those in most severely handicapped inpatients. Of the 43 subjects assessed with the psychiatric rating scale, 16 had symptoms of neurosis, 29 of florid psychosis, and 32 of a deficit state. Symptoms of deficit state were positively correlated with ratings of low social activity on the behavioural scale (Spearman's rank correlation coefficient 0.30, p = 0.03). CONCLUSIONS Hostels are having to care for long term severely affected psychiatric patients discharged into the community. The suitability of the services offered to such subjects should be assessed.