Automated clinical decision model construction from knowledge-based GLIF guideline models

Abstract

Many guideline-based decision models use rule-based criteria to set qualitative preferences. In our method, we envision incorporating expected values computed from a decision-theoretic model to the hierarchical representation framework. We take the knowledge-based Clinical Practice Guideline (CPG) model in Guideline Interchange Format (GLIF) as the input knowledge model. In addition, the domain medical ontologies provide data models and controlled vocabularies for referencing patient conditions and therapies that are relevant to managing diseases. These two parts build up the knowledge base for the clinical decision making. We develop an algorithm to automatically build a rough decision model (RDM) represented in influence diagram, from the knowledge base described above. Although the RDM is not complete in the structure, or parameters, it greatly reduces the efforts needed for constructing decision model manually. Decision makers could complete the decision model by modifying the RDM and filling in additional information (probabilities and utilities).