The validation of the Tamil version of the 12 item general health questionnaire

The 12 item General Health Questionnaire (GHQ-12), increasingly used to screen for common mental disorders (CMD) in primary care, has been validated in different languages and cultures. However, the validity of the Tamil version has not been established. Consecutive patients, attending a primary health care centre in Vellore, rural Tamil Nadu, India, were screened for CMD using the Tamil version of the GHQ-12. The subjects were also interviewed using the Revised Clinical Interview Schedule (CIS-R). The International Classification of Diseases-10: Primary care version (ICD-10 PHC) criteria were used to diagnose CMD. Various thresholds of the GHQ-12 were compared against the standards of the ICD-10 PHC. A receiver operator characteristic curve was drawn to obtain the best threshold value for screening. Principal Component Analysis was done to identify latent variables. The Cronbach's alpha and the split half reliability were also calculated. One hundred and eleven (33%) subjects of the 327 patients interviewed satisfied ICD-10 PHC criteria for CMD. The optimal threshold for the GHQ-12 was 2/3. This threshold had a sensitivity 87.4% and a specificity of 79.2%. Three factors were extracted with eigen values of 5.0 (depression-anxiety), 1.7 (social performance) and 1.1 (self-esteem) which explained 42.0%, 13.9% and 9.2% of the variance. The Cronbach's alpha was 0.86 while the split half-reliability was 0.83. The sensitivity and specificity of the Tamil version of the GHQ-12 is high. The factor structure is similar to that reported in other populations. The instrument can be employed as a screening instrument in this population.