Guidelines and protocols have gained support as vehicles for promulgating the best practices in clinical medicine, and many researchers have proposed frameworks for modeling them in a computer-interpretable format. However, for guidelines to be most useful for decision support, they need to be integrated into the patient care process and to be patient-specific. OBJECTIVE: To achieve this, as a part of the development of interoperable and sharable clinical decision support system in the context of electronic health records, we identified the system requirements on the CDS for integrating hypertension management into workflow of a health center. We also introduced our approach to the acquisition of medical knowledge.

A method combining the top-down and bottom-up approaches was used. For the top-down, the service scope and target group of patients were defined, and the knowledge sources, including JNC 7 and other guidelines from which decision logic was extracted, were identified. For the bottom-up approach, scenario-use cases were applied and an analysis of care data and a survey of the care providers were conducted.

This bidirectional method revealed the opportunities of our approach and the means by which improvements can be suggested in practice. The content and functions expected by users were also identified. These results will be used for further encoding knowledge at the computable level and application development at the implemental level.