Mining and retrieving medical processes to assess the quality of care

Abstract

In a competitive healthcare market, hospitals have to focus on ways to deliver high quality care while at the same time reducing costs. To accomplish this goal, hospital managers need a thorough understanding of the actual processes. Process mining can be used to extract process related information (e.g., process models) from data. This process information can be exploited to understand and redesign processes to become efficient high quality processes. Process analysis and redesign can take advantage of Case Based Reasoning techniques. In this paper, we present a framework that applies process mining and case retrieval techniques, relying on a novel distance measure, to stroke management processes. Specifically, the goal of the framework is the one of analyzing the quality of stroke management processes, in order to verify: (i) whether different patient categories are differently treated (as expected), and (ii) whether hospitals of different levels (defined by the absence/presence of specific resources) actually implement different processes (as they auto-declare). Some first experimental results are presented and discussed. © 2013 Springer-Verlag.