Investigation of the ontology and information model of morbidity reporting in the electronic health record environment

The emergence of electronic health records in the Australian health environment is anticipated to improve the quality and availability of secondary data for health care service planning, public health and health research. Achievement of this goal requires in-depth knowledge about the business requirements and data collection format to enable secondary data collections to be automated whilst ensuring data access, accuracy and data integrity. The research objective was to understand the clinical admitted episode morbidity entityrelationship model in the context of ontological structures used to represent clinical data in EHRs. The applicability of the current national morbidity data collection format used to collect information on admitted patient episodes of care from Australian hospitals to collection automation and analysis is dependent upon the capacity of that data to be represented in formats that accurately transfer knowledge and retain meaning. The hypothesis that an ontological categorical structure exists within the data instructions and component definitions of the national morbidity data collection was tested by the adoption of ontology engineering. This methodology included 1) a review of existing data collection contents and formats, and 2) detailing the inclusion and exclusion instructions provided in ICD-10-AM, the classification system used to describe clinical information in the collection. This resulted in the identification of categories used for classification purposes and the relationships between these categories in the collection of these data from which an ontology of Australian morbidity data was developed. A significant outcome of this study is a clearer understanding of what types of knowledge are represented in these data collections and the relationships between the different knowledge components. It is recommended that this knowledge be used to inform the (EHR) system requirements suited to maximizing the efficiency of a future automated data collection process and optimise the usability of the data collected in this manner.