Stability of the diagnosis of first-episode psychosis made in an emergency setting

Objective: This study aimed at evaluating the stability of the first psychotic episode diagnosis in the emergency context. Methods: Fifty-nine patients were selected during a 15-month period and were followed for an average of 19.35 +/- 6.12 months. The admission and discharge emergency diagnosis were compared with the longitudinal diagnosis, obtained by the application of Structured Clinical Interview for DSM-IV Axis I Disorders - clinical version at the end of the follow-up. Severity rating scales (Brief Psychiatric Rating Scale, Young Mania Rating Scale and Hamilton Rating Scale for Depression) were applied in the emergency assessment. Results: Agreement between admission emergency diagnosis and longitudinal diagnosis was unsatisfactory (k=0.25), whereas that between emergency discharge and longitudinal diagnosis was satisfactory (k=0.57). Brief psychotic disorder diagnosis presented higher sensitivity rates but low specificity, comprising several false positives. Bipolar disorder had the highest rates of specificity. Conclusion: Brief psychotic disorder may not be a useful concept in the emergency assessment. A short period of observation can improve emergency psychiatric diagnosis. (c) 2005 Elsevier Inc. All rights reserved.