Child global rating scales: selecting a measure of client functioning in a large mental health system.

The Global Assessment Scale for Children (GAS-Children) and the Children's Impairment Scale (CIS) were examined for inter-rater reliability, discrimination of outpatients from inpatients, and clinician acceptance. Forty-four clinicians used the two scales to rate 146 recently admitted children and adolescents in eight mental health programs. An additional study assessed the inter-rater reliability of the two scales compared with the Connors Parent-Teacher Questionnaire and the correlation of the GAS-Children with both the Connors Parents Questionnaire and Achenbach's Child Behavior Checklist. The GAS-Children showed better interrater agreement than any of the four subscales of the CIS, although the sum of the CIS subscales also showed adequate inter-rater reliability. Among adolescents, the GAS-Children correlated highly with the adult GAS, but the scales showed different mean values. Clinicians slightly preferred the GAS-Children over the CIS. To assess comparability of scale usage across sites, clinicians rated six case vignettes. Inpatient clinicians rated the vignettes as more dysfunctional than did their outpatient counterparts. Bias-adjusted scores still discriminated outpatient from inpatient children. Nevertheless, these rater biases should make evaluators cautious about comparing functioning scores across programs, even when the rating scale is ostensibly the same.