Psychometric Properties and Factor Structure of General Health Questionnaire 28 (GHQ-28) in Iranian Psychiatric Patients

The aim of this survey was to assess psychometric properties of GHQ-28 in Iranian psychiatric population and appropriating it as a screening instrument for assessing general health in this population.

Method and Materials: 

This was a psychometric study executed on 80 psychiatric patients and 80 matched normal participants. The assessments consisted of a psychiatrist’s interview based on DSM.IV criteria, and administration of GHQ-28 and a demographic questionnaire. Data was analyzed using factor, discrimination and regression analysis methods, as well as Pearson’s correlation co-efficient, Spearman- Brown’s and Guttmann’s split tests. 

Four factors were extracted from the GHQ-28, of which the first three explained the most percenttage of variance. Clinical cut-off point, sensitivity, specificity and classification error were found to be 24, 0.80, 0.99, 0.10 respectively. Criterion validity Chronbach's alpha and split reliability co-efficient were found to be 0.78, 0.97 and 0.90 respectively. 

Factor analysis revealed that the four factors which are the basics in this questionnaire have a high internal consistency. The calculated optimal clinical cut-off point for screening general health in Iranian population was 24 which ensure optimal psychometric indices. Coefficients of criterion validity, structural validity and reliability showed that GHQ-28 is one of the most valid instruments for screening general health.