Decision Support through Data Integration: Strategies to Meet the Big Data Challenge

Abstract

Objectives: Presentation of an overview of the reasons why data integration initiatives should be seen as enablers for effective decision support in data-intensive healthcare settings. Methods: Typical challenges rising from the information requirements of clinical decision support systems are highlighted. We then propose a methodological solution where several heterogeneous data sources are integrated by the means of a common data model on top of which the DSS is built. Results: We report on two successful case studies based on the DSSs developed in the context of the MobiGuide and Mosaic projects, funded by the European Union in the Seventh Framework Program. The MobiGuide patient guidance system has been successfully validated during a recent pilot study involving 30 patients (10 with atrial fibrillation and 20 with gestational diabetes), while Mosaic is currently undergoing a validation phase involving 1000 type 2 Diabetes patients. Conclusions: In the era of big data, effective data integration strategies are an essential need for medical informatics solutions and even more for those intended to support decision processes. Building generic DSSs based on a stable (but easily extensible) data model, specifically designed to meet the information requirements of DSSs and analytics, has proven to be a successful solution in the two presented use cases.