The validity of two malay versions of the general health questionnaire (GHQ) in detecting distressed medical students

ABSTRACT

Objective: To compare the sensitivity, specificity and reliability between the Malay GHQ-30 and the Malay GHQ-12 in detecting distressed medical students. This study determined which version is more sensitive, specific and reliable in detecting distressed medical students. Methods: Three validated instruments, the Malay version GHQ-30, the Malay version GHQ-12 and the Malay version Beck Depression Inventory-II (BDI-II), were administered to a total of 187 medical students. Distress diagnoses were made based on the Malay version BDI-II. ROC curve analysis was applied to determine the sensitivity and specificity of the both GHQs by testing against the BDI-II. Reliability analysis (Cronbach’s alpha) was applied to test internal consistency of the GHQ. The analysis was done using SPSS version 12. Results: 141 (75.4%) medical students participated in this study. The GHQ-30 sensitivity and specificity at cut-off point of 6 was 87.5% and 80.6% respectively with positive predictive value (PPV) of 70% as well as area under ROC curve was 0.93. The GHQ-12 sensitivity and specificity at cut-off point of 4 was 81.3% and 75.3% respectively with PPV of 62.9% as well as area under ROC curve was 0.82 The Cronbach’s alpha value of the GHQ-30 and GHQ-12 was 0.93 and 0.85 respectively. The Kappa value for the GHQ- 12 and GHQ-30 was 0.65 (p < 0.001). Conclusion: This study showed the Malay GHQ-30 had better ability to discriminate between distressed and non-distressed medical students compared to the Malay GHQ-12. The shorter version was found to be as reliable as the long version GHQ-30 in detecting distressed medical students. The optimum threshold score for the GHQ-30 and GHQ-12 to detect distressed medical students were 6 and 4 respectively. The GHQ-12 and GHQ-30 showed a good level of agreement in detecting distressed medical students.