Retest stability of DSM-III-R diagnoses among adolescents using the Diagnostic Interview Schedule for Children (DISC-2.1C)

Abstract

Retest stability of DSM-III-R diagnoses was assessed using the DISC-2.1C with a sample of Anglo, African, and Hispanic American adolescent patients 12 to 17 years of age. Based on the kappa statistic, retest stability was fair for any disorder (kappa = .50), for any anxiety disorder (kappa = .44), for any affective disorder (kappa = .53), for any disruptive behavior disorder (kappa = .58), and for substance use disorders (kappa = .46). Although there was a trend for reliability to be somewhat higher for African Americans (kappa = .58) than for Anglos (kappa = .42) or Hispanics (kappa = .49), these differences were not statistically significant. In general, 15- to 17-year-olds had somewhat better reliability (kappa = .58) than did 12- to 14-year-olds (kappa = .44). Males had somewhat higher reliability (kappa = .63) than females (kappa = .43). These findings are congruent with those reported recently using the DISC-R (Schwab-Stone et al., 1993) and suggest that the DISC appears to be at least as reliable as other available child diagnostic instruments. In view of the fair-to-moderate levels of reliability of these instruments in general, future research should focus on the joint effects of instrument, subject, interviewer, and nomenclature on operating characteristics of diagnostic interview schedules, focusing in particular on factors affecting accurate recall and reporting of symptoms and episodes.