Acceptability and Difficulties of (Fuzzy) Decision Support Systems in Clinical Practice

Medical knowledge in modern health care is vast and constantly changing, as well as expanding. The envisioned role of computer programs in health care is perhaps the most important. The provisioning of clinical decision support systems (CDSS) would enable the discovery of patterns in health data which might be important for the fight against incorrect diagnosis. Medicine uses empirical knowledge about superficial associations between symptoms and diseases, and uncertainty is without any doubt a central, critical fact about medical reasoning. Many of intelligent CDSS are based on the fuzzy set theory, which describes medical complex systems mathematical model in terms of linguistic rules. Considering the fuzzy nature of the data in a medical environment, it becomes obvious that the ability of managing uncertainty turns to be a crucial issue for clinical fuzzy decision support systems (CFDSS). Since the potential of medical decision making was first realized, hundreds of articles introducing decision support systems (DSS) have been published in the last three decades. But even today, only few systems are in clinical use, and their full potential for optimizing the healthcare system is far from realized. Clinician's acceptance and utilization of CDSS depends on its workflow-oriented context sensitive accessibility and availability at the point of care, and on the integration into a hospital information system (HIS). This paper describes advantages and disadvantages of several approaches, and experiences, gained by clinical use of two introduced CFDSS, integrated in the HIS of the Vienna General Hospital, to analyze the little use of CDSS in today's clinical practice.