University of Minnesota
University Relations

Minnesota Supercomputing Institute

Log out of MyMSI

Research Abstracts Online
January - December 2011

Main TOC ...... Next Abstract

University of Minnesota Twin Cities
Medical School
Department of Laboratory Medicine and Pathology

PI: Donald P. Connelly

The Detection and Measurement of Bone Marrow Biopsy Diagnostic Discordance: A Foundation for Clinical Decision Support

Bone marrow diagnostic testing by the cytogenetic, flow cytometry, and hematopathology laboratories provide clinicians with a wealth of information utilized in clinical decision-making. It is expected that the diagnoses determined by these laboratories on aliquots of the same bone marrow would be from the same World Health Organization diagnostic category even though they utilize different testing methods. The first phase of research determined the diagnostic discordance rate of these laboratories on bone marrow specimens. The laboratory reports have been downloaded from Fairview Clinical Information System into an MSI database for easy searching, retrieval, and storage. Determination of statistically significant discordance rates using Cohen’s Kappa statistic utilized SAS statistical software on the MSI computers. The second phase of research categorized discordant cases from Phase I by the following methods. The first method determines the relationship between the discordant diagnoses, while the second method looks at issues contributing to diagnostic discordance by the laboratory or other means. The third means of categorization analyzes where issues contributing to discordance occur in the testing process, while the fourth methodology looks at where in the testing process issues occur and the etiologies that contribute to the discordances seen. Lastly, aspects of reporting are analyzed to determine if amended reports and typographical errors contribute to discordances. These categorizations all provide information about the information process and where issues occur contributing to diagnostic discordance so that the appropriate decision support tool can be designed and implemented.

Group Member

Andrea Pitkus, Graduate Student