Global Assessment of Relational Functioning Scale (GARF): a validity study in patients with recurrent major depression in Brazil.

The Global Assessment of Relational Functioning Scale (GARF) has been proposed as an additional domain on Axis V, yet few validation studies have been conducted. This study examines the validity of the Brazilian-Portuguese version of the GARF. An experienced family therapist interviewed 34 families in which one member had a diagnosis of DSM-IV recurrent major depression. A semi-structured family interview was used to classify families into functional or dysfunctional. A second interviewer, blind to the earlier evaluation, used the GARF to assess the families' functioning. Validity coefficients with the best balance between sensitivity and specificity were obtained with a cut-off point of 70 on the GARF: sensitivity, 78%; specificity, 86%; positive predictive value, 95%; and negative predictive value, 50%. Regression analyses identified demographic and clinical variables that influenced GARF scores. The following variables accounted for 41% of the GARF's variance: patients' educational level, severity of the current depressive episode, and the role of the family member participating in the study. The Brazilian-Portuguese version of the GARF has acceptable validity coefficients, and may be a useful tool in cross-cultural settings.