The Structured Clinical Interview for DSM-III-R (SCID)

TLDR

Diagnostic disagreement stems from inadequate interviewer training, information variance, and low base rates of many disorders. The SCID’s test‑retest reliability was evaluated in 592 subjects across multiple patient and nonpatient sites, including one German site. In patient samples, kappa values exceeded .60 for current and .68 for lifetime diagnoses, whereas nonpatient agreement was lower (.37 current, .51 lifetime), comparable to other structured instruments.

Abstract

A test-retest reliability study of the Structured Clinical Interview for DSM-III-R was conducted on 592 subjects in four patient and two nonpatient sites in this country as well as one patient site in Germany. For most of the major categories, kappa s for current and lifetime diagnoses in the patient samples were above .60, with an overall weighted kappa of .61 for current and .68 for lifetime diagnoses. For the nonpatients, however, agreement was considerably lower, with a mean kappa of .37 for current and .51 for lifetime diagnoses. These values for the patient and nonpatient samples are roughly comparable to those obtained with other structured diagnostic instruments. Sources of diagnostic disagreement, such as inadequate training of interviewers, information variance, and low base rates for many disorders, are discussed.