Test-retest reliability of the computerized CIDI (CIDI-Auto): Substance Abuse Modules

Maritza Rubio-Stipec*, Lorna Peters, Gavin Andrews

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    39 Citations (Scopus)


    This paper discusses the reliability of the Alcohol and Substance Abuse modules of the CIDI-Auto in two countries, Australia and Puerto Rico, and two languages, English and Spanish. CIDI-Auto is a computer-assisted version of the CIDI. Reliability estimates for DSM/ICD are presented at the diagnostic and symptom level. In total, 286 subjects, ages 17-60 years, who had at least 12 drinks of alcohol in their lifetime participated in the study. Adequate to good test-retest reliability estimates were obtained, with no major differences by nosology, site, substance, or time. Harmful use/abuse showed lower kappas than dependence. Reliability estimates for dependence ranged from 0.70 to 0.95. For harmful use, κ's ranged from 0.45 to 0.66. The findings are encouraging; CIDI-Auto produces reliable classification across two settings and in two languages with an instrument that has good coverage of different manifestations of the illness.

    Original languageEnglish
    Pages (from-to)263-272
    Number of pages10
    JournalSubstance Abuse
    Issue number4
    Publication statusPublished - 1999


    Dive into the research topics of 'Test-retest reliability of the computerized CIDI (CIDI-Auto): Substance Abuse Modules'. Together they form a unique fingerprint.

    Cite this