Objective: To assess the reliability of the Substance Abuse Module of a Persian translation of Composite International Diagnostic Interview (CIDI), in a multi-center study in Iran.
Method: the sample consisted of 101 inpatients and outpatients from five psychiatric hospitals in Tehran. Trained psychologists conducted the interviews. The completed CIDI was administered to 65 subjects and CIDI-substance abuse module (CIDI-SAM) was administered to 36 others. To study the test-retest reliability, a second interview by a different person was administered to all subjects within 3 to 7 days of the first interview. Kappa and Yule’s statistics were used to evaluated reliability.
Results: Overall, the reliability of substance disorder diagnosis was higher for complete CIDI than for CIDI-SAM. In complete CIDI in both diagnostic systems of ICD-10 and DSM-IV, Kappa statistics was 0.5 or higher with a generally higher Yule’s statistics; and opioid dependency had the highest kappa (0.64) among different disorders. In CIDI-SAM, the highest reliability regarding Kappa and Yule results was related to cannabis dependency (Kappa=0.41; Yule=0.6).
Conclusion: Overall, the reliability of complete CIDI for the diagnosis of substance abuse disorders was better than CIDI-SAM. Weak reliability, especially in the Substance Abuse Module, may be due to the method of sampling or the high prevalence of the disorder.
Type of Study:
Research |
Subject:
Special Received: 2005/02/16 | Accepted: 2005/04/22 | Published: 2005/06/22