Patients' limited knowledge about medications is a key factor in post-discharge adverse drug events. This can be improved by providing optimal advice to patients in discharge summaries for managing their post-discharge care. This paper presents a prototype system, Medication Information Decision Support System (MI-DSS), which brings together natural language processing and ontology based semantic annotations to develop a clinical decision support system to assist Electronic Discharge Summary (EDS) authors in providing medication advice to patients in EDSs. Our approach involved the selection and modeling of high risk discharge medications and their consumer related advice. We present our medication information ontology which models the medication knowledge necessary for consumers to manage their post-discharge self care. This ontology serves as the knowledge source to semantically annotate the EDS text and determine the medication-specific recommendations in the MI-DSS.
展开▼
