Improving clinical hematopathology quality using decision support methods

Clinical hematopathology has the potential to serve as the prototype of the post-genome-era pathology laboratory due to its varied and complex data management needs. Modern hematopathology laboratory testing is multifaceted and includes slide-based image analysis using detection techniques for lymphomas and leukemias, fluid-based sample analysis generating quantitative data, and flow cytometry processing for detection of cell-surface markers. The strategic integration of these analysis techniques for improving decision support to clinical hematopathologists is rarely seen due to a lack of system integration. Current laboratory information systems (LISs) are not effective in providing flexible reporting methods, particularly in integrating various types of data into a single report. Futhermore, LISs are unable to support data exploration (data mining) to further pathology research due to their lack of a cogent data model; much of the data stored within such systems is not readily accessible.; This research presents the design, implementation, and evaluation of a decision support system for improving outcomes in clinical hematopathology. Integrated reporting among various hematopathology laboratories was achieved through the development of a scalable data model as part of the decision support system (DSS). Using the DSS, laboratory personnel are able to create user-defined data types for clinical, quality assurance, and research endeavors.; Clinical laboratory quality indicators such as specimen turnaround time, technologist workload, errors and cost were used to measure changes in laboratory performance from the time prior to use of the DSS to the period after implementation. The use of the DSS led to a statistically significant decrease in specimen turnaround time, an increase in workload capacity, and an increase in the number of cases processed per technologist. The application also led to a decrease in cost through a reduction in clerical FTE's. The system's quality assurance algorithms discovered a relationship between specimen processing and flow cytometry data quality that had previously gone unnoticed. The integrated, longitudinal reporting of HIV-1 mRNA viral load and CD4+ T-cell values improved the level of decision support for antiretroviral therapy monitoring in HIV patients. The DSS serves as a foundation for further outcomes based research in clinical hematopathology as newer, higher throughput molecular techniques are used.