Description
Clinical-guideline based decision support for multimorbidity patients encounters interactions among recommendations originating in different guidelines. We developed an algorithm that reasons on computer-interpretable guidelines (CIG) and on the FDA’s repository of side effects to detect adverse events and rerun the CIG decision that recommended the drug that contributes to that effect, to find alternatives. We evaluated the algorithm using a case study of a patient with diabetes, taking Gliflozin that caused a fungal infection.
Learning Objective: To reason about multimorbidity, knowledge from several sources can be combined: computer-interpretable guidelines (CIGs) for each morbidity and a knowledge base of reported side effects (AEOLUS). This can enable detection of an adverse event (which can be a second disease) due to a side effect of a medication given for a first disease. The audience will learn how CIGs can be specified in a goal-oriented way in order to support such reasoning.
Authors:
Alexandra Kogan, University of Haifa
Mor Peleg (Presenter)
University of Haifa
Samson Tu, Stanford University
Presentation Materials: