BACKGROUND: This paper describes the rationale, development, reliability and validity of a new screening psychiatric instrument. METHOD: The instrument comprises 26 items that tap the cardinal features of main psychiatric categories as defined by ICD-10 and DSM-IV. These items were adapted from various structured and semi-structured diagnostic interviews that yield ICD-10 and DSM-IV psychiatric diagnoses. After a training course, 12 trainees and the trainer rated blindly the 26 items on 45 subjects (22 with psychopathology and 23 without). Inter-rater reliability coefficient (Kappa) was estimated between trainees and the trainer on each item of the instrument. The total score on the new instrument was then correlated with the total score on the Arabic Self Reporting Questionnaire (SRQ-20) and the Arabic version of the General Health Questionnaire (GHQ) in a random sample from the general population (n = 365). Logistic regression was utilised to estimate the power of the total score on the new instrument in discriminating between cases and non-cases as classified by the SRQ-20. RESULTS: Excellent levels of agreement (Kappa > 0.80) were found for all items except for obsession (Kappa = 0.65) and for depressed mood (Kappa = 0.70). Moderate correlations were found between the total score on the new instrument and total score on SRQ-20 (r = 0.69) and the total score on the Arabic GHQ (r = 0.7). The new instrument correctly classified 89% of subjects into cases and non-cases. CONCLUSIONS: The results of this study indicate that the new instrument is a highly reliable and valid screening instrument. The authors are now investigating its test-retest reliability and its procedural validity.
BACKGROUND: This paper describes the rationale, development, reliability and validity of a new screening psychiatric instrument. METHOD: The instrument comprises 26 items that tap the cardinal features of main psychiatric categories as defined by ICD-10 and DSM-IV. These items were adapted from various structured and semi-structured diagnostic interviews that yield ICD-10 and DSM-IV psychiatric diagnoses. After a training course, 12 trainees and the trainer rated blindly the 26 items on 45 subjects (22 with psychopathology and 23 without). Inter-rater reliability coefficient (Kappa) was estimated between trainees and the trainer on each item of the instrument. The total score on the new instrument was then correlated with the total score on the Arabic Self Reporting Questionnaire (SRQ-20) and the Arabic version of the General Health Questionnaire (GHQ) in a random sample from the general population (n = 365). Logistic regression was utilised to estimate the power of the total score on the new instrument in discriminating between cases and non-cases as classified by the SRQ-20. RESULTS: Excellent levels of agreement (Kappa > 0.80) were found for all items except for obsession (Kappa = 0.65) and for depressed mood (Kappa = 0.70). Moderate correlations were found between the total score on the new instrument and total score on SRQ-20 (r = 0.69) and the total score on the Arabic GHQ (r = 0.7). The new instrument correctly classified 89% of subjects into cases and non-cases. CONCLUSIONS: The results of this study indicate that the new instrument is a highly reliable and valid screening instrument. The authors are now investigating its test-retest reliability and its procedural validity.