Prediction of in-patient use in first-admitted patients with psychosis

Purpose: The aim of this study is to examine the utilization of psychiatric services for inpatients with psychosis over a period of five years after first admission, and to identify factors that contribute to intensive service use in terms of cumulative length of in-patient treatment and readmission rate. Methods: A cohort of 424 patients with psychotic disorders (out of a sample of 2565 first-admitted patients from a catchment area in Switzerland) was examined by means of register data. Results: Patients admitted for psychosis spent the longest time in hospital compared to other diagnoses, but there was considerable within-sample variation, and most patients (60.4%) had only one in-patient episode. Of the total time accumulated in this sample, 50.5% was 'consumed' by 10.7% of patients. Regarding the overlap between 'heavy use' and 'frequent use' (those 10% of the sample with the longest/most frequent hospitalizations), only 39.5% of the 'frequent users' were also 'heavy users' (vice versa: 38.6%). 'Clinical' measures were the main predictors of the overall time spent as an in-patient (schizophrenia diagnosis, therapeutic measures, clinical improvement). Concerning 'heavy use', homelessness was yet another important risk factor. Sociodemographic variables (age, level of education, living alone) particularly influenced the number of hospitalizations. A younger age at first admission was predictive of 'heavy' and 'frequent' use. Discussion: Clinical and sociodemographic factors explain differences in in-patient service consumption among patients with psychosis. Efforts to devise effective interventions have to take both into account, but different measures are needed to address 'heavy' and 'frequent' use. (c) 2006 Elsevier SAS. All rights reserved.