Abstract
This report addresses some issues relating to the prevalence rate of depressive disorders in primary care settings. It is very likely that the majority of people suffering from depression do consult the non-psychiatric practitioner. Many patients consult doctors for unrelated physical illness or for nonspecific somatic manifestations of depressive disorders. However, relevant epidemiological studies of these depressed patients are not sufficiently reliable because of methodological problems. Since 1972,we have participated in the WHO collaborative study on public health implications of depressed disorders. In collaboration with internists from two general hospitals, we have been able to estimate the prevalence rate of depressive disorders in outpatient clinics of these hospitals on two different occasions. In the first study (1977-1980), 831 cases were systematically selected from new ambulatory medical patients and interviewed by research psychiatrists using a screen from and the SADD schedule. Subsequently, 50 patients were found to have a depressive disorder (ICD-9). The prevalence rate of depressive disorders in these settings was 6%. The second study (1991) was part of a more extensive, ongoing research project organized by WHO, Using the "Two phase sampling design" (W. E. Deming) and using both the GHQ-12 and CIDI, we were able to screen 1043 ambulatory patients in one of above mentioned hospitals. The estimated prevalence rate of depressive disorders (DSM-III-R) was 5.4%. In this paper we questioned whether screening for depressive disorders has any value in general medical settings. The performance of a screening test instrument is measured by the predictive value of a positive or negative result. The predictive values of a positive result falls as the prevalence declines. Understanding this point is of practical importance. The conclusion of this paper is that depression questionnaires should not be routinely administered to ambulatory medical patients. There may be merit, however, to employ such assessment practice in selected circumstances and with a well-defined population.
