Factor solution of the BPRS-expanded version in schizophrenic outpatients living in five European countries

<bold>Rationale: </bold>The expanded version of the Brief Psychiatric Rating Scale (BPRS-E) has improved the instrument's coverage and interrater reliability, but there is little knowledge on its subsyndromes.<bold>Objectives: </bold>To assess: (1) whether there are common underlying BPRS-E subscales in patients living in different countries and (2) if this is the case, whether these subscales behave the same in all populations and, if not, what are the differences over these populations.<bold>Methods: </bold>Data are part of the EPSILON study, a collaborative project carried out in Denmark, England, Holland, Italy and Spain. A random representative sample of 404 adult patients with a ICD-10 diagnosis of schizophrenia who have been in contact with mental health services of a defined catchment area in each site were assessed. Simultaneous component analysis (SCA) was used to find component weights that optimally explain the variance of the variables in different populations simultaneously.<bold>Results: </bold>Symptom severity differed significantly among the five EPSILON sites in 12 out of 24 BPRS-E items, but a common component solution could be found. It explained 48.8% of the variance and gave four well-interpretable components: manic excitement/disorganization, depression/anxiety, negative and positive symptoms. Each component's internal consistency and intercomponent correlation matrix differed significantly among sites. The four components mean score differed significantly among sites for negative symptoms and depression/anxiety.<bold>Conclusions: </bold>In spite of the heterogeneity of symptom's severity in the various countries, the way symptoms cluster in schizophrenia is rather stable cross-culturally. Data demonstrate that to explore schizophrenia a third component, including mania/disorganization items, is necessary beside the positive-negative symptom dimensions. The subscales derived from these analyses can be readily used in clinical trials and epidemiological studies. [ABSTRACT FROM AUTHOR]