Using Protégé for Modeling Clinical Practice Guidelines: The Experiences of the EON Project

The EON project at Stanford was a 6-year project, funded by the National Library of Medicine, to develop a component-based architecture for creating decision-support systems for guidelinebased care. As part of the architecture, the project created a suite of models and software components, including a set of middleware servers that perform the computation necessary to support specific tasks in guideline-based patient care. One such server, the guideline execution server [1], takes as inputs formalized clinical guidelines and relevant patient data to generate situation-specific recommendations. A second server, the Tzolkin temporal data mediator [2], extends the traditional relational database server to include capabilities to resolve queries involving complex temporal relationships and to create temporal intervals representing abstractions derived from primitive time-stamped data. A third component, known as WOZ [3], provides explanation services for other components. The technology developed in the EON project forms the basis for the ATHENA decision-support system for the management of hypertension, which had been deployed and validated at a number Department of Veteran Affairs sites [4].