OBJECTIVES: To assess the reliability of a Portuguese version of the Structured Clinical Interview for DSM-IV - Clinical Version (SCID-CV). METHODS: Forty-five psychiatric patients at the university hospital of Ribeirão Preto, Brazil, were assessed twice in two independent interviews (test-retest). Inter-rater agreement analysis was done using the kappa coefficient (K). RESULTS: The weighted Kappa was excellent (Kw=0.83). Reliability was statistically significant for affective disorders (K=0.87); psychotic disorders (K=0.90); substance-related disorders (K=0.76); anxiety disorder (K=0.61); and for all specific diagnostic categories analyzed, except for agoraphobia without history of panic attacks (K=-0.04). CONCLUSIONS: The present Portuguese version of the SCID-CV showed good inter-rater reliability, but the lack of specific questions in the interview and specific diagnostic criteria for some disorders, such as agoraphobia without history of panic attacks, resulted in poor reliability.
Mental status schedule; Reproducibility of results; Diagnosis