Hidden psychiatric illness: use of the general health questionnaire in general practice.

A 10% random sample comprising 234 adults attending a general practitioner was studied to obtain an estimate of conspicuous and hidden psychiatric morbidity and to determine the value of the general health questionnaire in improving case recognition in general practice. Patients completed the 28-item general health questionnaire before seeing the general practitioner, who completed a rating sheet without seeing the general health questionnaire score. The doctor identified a psychiatric component in 38% of men and 53% of women and diagnosed psychiatric disorder in 22% of men and 31% of women. Using a cut off point of nine or above, high general health questionnaire scores were found in 25% of men and 29% of women. Agreement between the general health questionnaire and the doctor's assessment was better for males (misclassification rate 16%) than for females (20%). A subsample of patients scoring over the recommended threshold (five or above) on the general health questionnaire were interviewed by the psychiatrist to compare the case detection of the general practitioner, an independent psychiatric assessment and the 28-item general health questionnaire at two different cut-off scores. The general health questionnaire may be a useful tool for improving recognition of psychiatric morbidity in general practice if the cut-off point is raised above that recommended for epidemiological research.