Factors associated with admission to hospital following emergency psychiatric assessment.

Psychiatric emergencies account for a large proportion of total referrals and admissions yet there has been little research in the United Kingdom into factors associated with admission after emergency psychiatric assessment. We conducted a one year prospective study of all emergency referrals from Borders Region in Scotland. Four hundred and eighty-seven emergencies were assessed and 153 (31%) of these were admitted. The main predictors of admission after stepwise logistic regression analysis were: risk to self (odds ratio 1.76 for each point increment on a five point scale, 95% confidence intervals 1.43 to 2.15), current contact with the psychiatric services (2.46; 1.52, 3.98) and psychotic diagnosis (2.38; 1.47, 3.87). Compared to subjects assessed at home, those assessed at the psychiatric hospital were more likely to be admitted (3.73; 1.99, 6.99) and self-referrals were less likely to be admitted than General Practitioner referrals (0.22; 0.08, 0.55). Our results suggest that future studies of emergency services should investigate and control for both clinical and service variables.