The authors describe the previous studies of the stability of
psychiatric diagnoses over time that have found such diagnoses to be
unreliable, stating that the assumptions underlying these studies have been
tested and disproved. On the other hand, the use of standardized interview
techniques and categorization by computer yield reliable symptom ratings
and precise diagnoses. The authors discuss the importance of these findings
to the selection of the most appropriate treatment modality for individual
patients.Abstract Teaser