Abstract
Objective: This study examines the reliability of antemortem diagnoses of schizophrenia using DSM-IV criteria.
Method: The case histories of 83 subjects with a provisional diagnosis of schizophrenia at autopsy were retrospectively reviewed using a semi-structured chart review and application of DSM-IV criteria. Agreement between antemortem and postmortem diagnoses of schizophrenia was examined, as well as the concordance between DSM-IV diagnoses and previously obtained diagnoses using DSM-Ill-R and ICD-10 criteria for schizophrenia.
Results: According to DSM-IV, 30.1% of cases did not have schizophrenia, compared to 36.1% using DSM-Ill-R criteria and 51.8% of cases using ICD-10 criteria. Concordance between DSM-IV and DSM-Ill-R diagnoses of schizophrenia was excellent (k = 0.81), but only fair between DSM-IV and ICD-10 (k = 0.57). Of the cases that did not meet the formal criteria for schizophrenia, the majority were reassigned diagnoses of schizoaffective disorder and affective disorder.
Conclusions: The use of human brain tissue in postmortem studies of schizophrenia must be linked to standardised diagnostic assessment procedures. Diagnoses can be upgraded with the development of new criteria, providing sufficient clinical data is available in case histories.