The Diagnosis is Critical for Psychiatric Brain Banks: an Interrater Reliability Study

Therese Garrick, Xanthe Glaw, Clive G. Harper, University of Sydney Pathology, and Amanda North, Neuroscience Institute of Schizophrenia and Allied Disorders, Sydney, New South Wales, Australia:

The New South Wales Tissue Resource Centre (TRC) at the University of Sydney Australia has been collecting human brain tissue for use in psychiatric research since 1994. Sixty per cent of the next-of-kin (NOK), when approached within 60 hours of their relative's death, agreed to donate their relative's brain tissue for research purposes [1]. Cases are obtained from the New South Wales Coroners' Office in Sydney. The TRC (brain bank) obtains consent for the donation in accordance with our ethics approval from the Sydney South-west Area Health Service, Australia.

All processes involved in the collection, storage, case characterization and subsequent distribution of tissues to researchers are critical. However, the confirmation of the lifetime psychiatric diagnosis of the deceased is the most challenging aspect [2]. It is achieved by reviewing all available medical records and/or interviewing the NOK.

The instrument used for the diagnostic process is the Diagnostic Instrument for Brain Studies (DIBS) [3], which is compliant with the Diagnostic and Statistical Manual IV (DSM-IV) [4]. As part of our ongoing ‘gold standard’ approach to this process, we perform inter-rater reliability on select cases to ensure that all diagnoses are accurate and reliable. Therefore, this study aims to determine the accuracy of the use and outcome of the DIBS and to examine the level of competency of the clinicians involved in the process.

All cases in this study (n = 20) were subject to medical record review and/or NOK interview. The information gathered was formatted to produce a clinical summary. To ensure that clinical summaries are auditable, they are structured in a standardized manner including page and line numbers. Two independent clinicians (two psychiatric nurses) examined the clinical summaries and applied the DIBS blind to the primary Axis 1 diagnosis, subtype and longitudinal course components of the DSM-IV. Cohen's kappa coefficient was computed using SPSS version 11 (SPSS Inc., Chicago, IL, US).

Inter-reliability for the three components was very high. For the Axis 1 diagnosis component, the kappa coefficient was 1.0, for the subtype component: k = 0.94 and for the longitudinal course component the kappa coefficient was 0.84.

We have demonstrated that it is possible to produce reliable lifetime psychiatric diagnosis in cases retained in brain banks. However, this can only be achieved with a commitment to the training and ongoing review of the clinicians using the diagnostic instruments. The outcome of the research using the brain tissue is dependent on the accuracy of the processes used by brain banks. Inaccuracies in the lifetime psychiatric diagnosis will jeopardise the results of research using post-mortem brain tissue and the subsequent interpretation of the results.