Early detection and intervention are essential for preventing clinical deterioration in patients. We are developing a two-tiered clinical warning system designed to identify the signs of clinical deterioration and provide early warning of serious clinical events at general hospital units. The first tier of the system automatically identifies patients at risk of clinical deterioration from existing electronic medical record databases. The second tier performs real-time clinical event detection based on vital sign data collected from on-body wireless sensors attached to those high-risk patients. Wireless sensor networks play an important role in clinical warning by collecting real-time vital signs for clinical decision support. This talk presents the architecture of, and our experiences with, a large-scale wireless clinical monitoring system. Our system encompasses portable wireless pulse oximeters, a wireless relay network spanning multiple hospital floors, and integration with electronic medical record databases. We report our experience and lessons learned from a 14-month clinical trial of the system in six hospital wards of Barnes-Jewish Hospital in St. Louis, Missouri. Our experiences show the feasibility of achieving reliable vital sign collection using a wireless sensor network integrated with hospital IT infrastructure and procedures. We highlight technical and non-technical elements that pose challenges in a real-world hospital environment and provide guidelines for successful and efficient deployment of similar systems. The convergence of wireless sensors, mobile computing, data mining and electronic medical record in clinical warning systems will lead to enhanced quality of care for patients in hospitals and potentially outpatients in their everyday lives.
Awards and Honors:
- Outstanding Paper Award, Euromicro Conference on Real-Time Systems (2013)
- Best Student Paper Award, IEEE Real-Time Systems Symposium (2012)
- NSF CAREER Award. National Science Foundation (2005)