The validity of the 28-item General Health Questionnaire in a Nigerian antenatal clinic.

The 28-item General Health Questionnaire (GHQ-28) was validated against the Psychiatric Assessment Schedule (PAS) using data collected during a prospective study of psychiatric disorders associated with childbirth among 277 Nigerian women attending an antenatal clinic. Using the results of correlational analyses of the total scores on the GHQ and on the PAS and the traditional validity coefficients of sensitivity, specificity, and misclassification rate, the instrument was shown to be a valid tool for the detection of psychiatric morbidity in this population. With a sensitivity of 82% and a specificity of 85%, the revised scoring method devised by Goodchild and Duncan-Jones performed better in discriminating 'cases' from 'non-cases' than the conventional scoring method, which values respectively, of 75% and 83%. In a discriminant function analysis, only three of the four subscales of the instrument contributed to its discriminating power.