LiSA - Live System Analysis: A robust approach for the real time monitoring of clinical environments

Abstract

The use of data produced in the University Hospitals of Geneva for steering and governance is becoming increasingly important. In this regard, many tools have been developed, both for aggregating and consolidating storage and also in terms of analysing Big Data. However, these tools are not generally designed to provide a real-time analysis. Indeed, there is an increasing amount of data available and it is easy to get lost. The potential uses of real-time monitors are numerous and can greatly increase the responsiveness of the institution in various situations by providing relevant data needed in that situation. The LiSA project aims to develop the conceptual and technical architecture for building real-time monitors, objects potentially useful in critical situations. A robust architecture is necessary to build a nearly real-time system. Management of alerts and events in a healthcare environment is so complex and scattered that it has become necessary to centralise. In software engineering, one of the most reliable architectures for such systems is the Publish-Subscribe Design Pattern. This architecture has been implemented in the project in order to give access to any kind of event in the Clinical Information System (CIS). Once data are processed and organised, a data stream that can be read by humans is sent to a web based monitoring interface. This new versatile monitoring model could lead to better and clearer management. We found that this system can stimulate managers to access the CIS events by creating meaningful indicators. It could help governance when it is necessary to obtain some information immediately, or to create statistical charts or gauges.