The concept of major depression. III: Concurrent validity of six competing operational definitions for the clinical ICD-9 diagnosis

The comparative validity of six operational diagnoses of major depression was evaluated in 600 psychiatric inpatients using the independently assessed clinical ICD-9 diagnoses as a yardstick. Agreement with, and positive predictive value for the ICD-9 categories of pure (endogenous and psychogenic) depression served as validation criteria; sensitivity of major depression diagnoses for detecting ICD-9 bipolar depressions was additionally used for examining the adequacy of width, time and exclusion criteria of the competing operational definitions. Three essential results were found. First, the "old" diagnostic definitions of RDC and FDC are superior to all newer definitions because they define the time criteria and the schizophrenic exclusion criteria more adequately than, for example, both DSM-III and DSM-III-R definition. Secondly, the current ICD-10 definition of 1989 ("mild", "moderate" or "severe" depression) comes closer to the concurrent validity of RDC and FDC than DSM-III, DSM-III-R and the previous ICD-10 definition of 1987. Thirdly, using the criterion of identifying a high proportion of ICD-9 bipolar depressions, all six competing diagnostic systems are too restrictive. Evaluations of predictive and criterion-related validity will be needed to substantiate these findings.