Integrating humans, devices, and events in clinical workflow processes

Abstract

Clinical workflow processes do not only strongly rely on human activities but may also depend on events that trigger the execution of work items. We describe a novel approach that integrates human tasks, devices, and events into a single process model. As a sample workflow, we discuss the patient's admission to an intensive care unit. Given an infrastructure of loosely-coupled disparate systems, such as auto-ID devices and clinical information systems, process integration is accomplished by adopting as well as extending an open-source messaging middleware that is known as JCoupling. We argue that JCoupling in conjunction with auto-ID technology allows for the replacement of manual, computer-related tasks by semi-automatic message-driven interactions. Our approach is based on the integration of JCoupling with the forthcoming release of the workflow management system YAWL. This ensures explicit support for resource-centric workflow patterns and thus lends additional power to our proposal. © 2009 Springer Berlin Heidelberg.