Introduction: The aim of this research was to validate the General Health Questionnaire- 28 (GHQ-28) as a screening instrument of mental disorders in the population of aged 15 yr and above in Tehran City, capital of Iran.

Methods: Overall, 879 people (455 females and 424 males) were chosen by the randomized cluster sampling from families under the coverage of health centers of the universities of Tehran, Iran and Shahid Beheshty. The study was carried out in two stage procedures. In the first step the GPs completed the GHQ-28, and in the second step all the samples were clinically interviewed by psychiatric residents based on DSM-IV diagnostic criteria.

Results: By using conventional scoring method of GHQ, the cut off point for the sample was 6. The sensitivity and specificity of this questionnaire were 84.7% and 93.8% respectively, and the overall misclassification rate was 8.2%. Likert scoring method of GHQ showed a cut off point of 24 and the sensitivity, specificity and the overall misclassification rate of 70.5%, 92.3% and 12.3% respectively. The concurrent validity of this questionnaire carried out by the Symptom Checklist- 90- Revised indicated that there was a meaningful correlation between scores of samples within the scales of these instruments. The Pearson correlation between the scales and total scores of GHQ-28 suggested that fluctuation of anxiety symptoms had an important character in predicting abnormality than the other scales.

Conclusion: The evaluation of reliability of GHQ-28 carried out by the test- retest procedure a week after the first stage, and validation of this instrument indicated that the instrument could be used successfully in the epidemiological study of mental disorders as a screening instrument.

Hakim Research Journal 2009 11(4): 47- 53.