יובל שחר

We report on new projection engine which was developed in order to implement a distributed guideline-based decision support system (DSS) within the European project MobiGuide.In this model, small portions of the guideline knowledge are projected, i.e. ‘downloaded’, from a central DSS server to a local DSS in the patient’s mobile device, which then applies that knowledge using the mobile device’s local resources. Furthermore, the projection engine generates guideline projections which are adapted to the patient’s previously defined preferences and, implicitly, to the patient’s current context, which is embodied in the projected knowledge. We evaluated this distributed guideline application model for two complex guidelines: one for Gestational Diabetes Mellitus, and one for Atrial Fibrillation. We found that the initial specification of what we refer to as the customized guideline should be in the terms of the distributed DSS, i.e., include two levels: one for the central DSS, and one for the local DSS. In addition, we found significant differences between the customized, distributed versions of the two guidelines, indicating further research directions and possibly additional ways to analyze and characterize guidelines.