Forecasting acute inflammatory condition and decision support tool

Abstract

A catastrophe-theoretic approach is provided for predicting an occurrence of an acute inflammatory condition or event (e.g., SIRS or sepsis) for a human patient based on a time series of monitored vital signs values measured from a patient, and in some instances, for providing advanced notice to clinicians or caregivers when such an acute inflammatory event is forecasted or modifying treatment for the patient, according to the predicted likelihood. In particular, an acute inflammatory condition management system is provided for determining a likelihood of near-term future significant acute inflammation in human patients. Embodiments of the disclosure described herein may provide a forecasted risk for future significant acute inflammation within a time horizon comprising a future time interval. In one embodiment, the future time interval is from 30 min to approximately 8 hours into the future, and may be dependent on the frequency of vital signs measurements.