A study of outcome in patients treated at a psychiatric emergency unit

Abstract

BACKGROUND: Although being an important part of the psychiatric treatment chain, there are few outcome studies of treatment at psychiatric emergency units (PEU). AIMS: The aim was to measure changes in psychopathology and humiliation during admission at a PEU. METHODS: The sample consisted of 147 patients examined at admission and discharge. The instruments used were the Brief Psychiatric Rating Scale (BPRS), the Narcissistic Personality Inventory-29 (NPI-29), the Hospital Anxiety and Depression Scale (HADS), a combination of questions measuring negative experiences and Cantril's ladder measuring experienced humiliation. Outcome measures were clinically significant improvement [>10% reduction of the BPRS converted (0-100) score] and changes on the other instruments. RESULTS: Median hospitalization time was 13 days (mean 20.4 days). Fifty-six per cent of the patients showed clinical significant improvement (95% CI 48-64%), 42% showed some degree of improvement and 2% were unchanged. The more improved patients had higher scores at admission than those with less improvement on all scales, indicating a floor effect. Small changes were observed for narcissism and experienced humiliation and negative admission events. In multivariate analyses high admission scores on BPRS subscales, thinking disorder and activation and HADS total score were significantly associated with clinically significant improvement. Type and length of admission did not significantly affect the outcome. The BPRS, HADS and NPI-29 scores at discharge were mainly explained by corresponding admission scores. CONCLUSIONS: More than half the patients admitted to PEU have clinically significant reduction of psychopathology during their stay. Higher levels of psychopathology at admission were significantly associated with improvement. Negative admission experiences and involuntary admission did not influence outcome.