Event prediction


Intelligent monitoring in intensive care units

Overview

Critical Care Medicine, especially that provided in the Intensive Care Unit, is expensive and extremely complex. While intelligent monitoring and decision support has the capability to synthesize physiological complexity into a set of clinically informative states, these systems are not currently widely available throughout the world. This research aims to develop accurate prediction metrics (for events such as outcome, readmission and infection) using as few physiological parameters as possible. This allows for reduced instrumentation, using only low cost alternatives, with no significant impact on accuracy. This research goal is exemplified by the newly developed OASIS (Oxford Acute Severity of Illness Score), which allows for accurate tracking of patient health with far fewer variables than are traditionally required. The score has applications in a variety of clinical care scenarios where there are insufficient resources to fully monitor an individual patient.