Background:
The validity of postmortem human brain research relies upon accurate clinical and psy-chopathological diagnosis. Current literature shows few instances where standardized diagnostic assessment tools such as the Diagnostic Instrument for Brain Studies (DIBS) have been used. The present study investigates the degree of concordance between predominant antemortem psychiatric diagnoses indicated in medical records and postmortem diagnoses derived through structured diagnostic instruments such as the DIBS and the Item Group Checklist (IGCL) of the Schedules for Clinical Assessment in Neuropsychiatry (SCAN).
Methods:
Fifty-eight subjects from the New South Wales Tissue Resource Centre with a recorded diagnosis of mental illness during their lifetime were included in the study. The predominant antemortem psychiatric diagnosis of each case was compared with its corresponding postmortem diagnosis obtained through structured case reviews to which either the DIBS or the IGCL of the SCAN were applied. Demographic variables such as age of illness onset, and alcohol or other drug use were also examined.
Results:
Comparison of diagnoses obtained from these two approaches produced an overall kappa coefficient of 0.66. Kappa coefficients for the schizophrenia cohort were 0.61, 0.35 for the schizoaffective cohort, 0.95 for the major depressive disorder cohort and 0.70 for the bipolar disorder cohort.
Conclusions:
These results indicate moderate to excellent interrater reliability for most cohorts in this sample. There is sufficient disagreement, however, particularly in the schizoaffective cohort, to suggest the value of applying standardized and structured assessment to enhance both the accuracy of diagnosis and the prospective validity of tissue-based research.
