Dascena publishes on a novel technique of homeostatic analysis
In this study Dascena applies an aggregate health score to patient stability prediction. The findings are published in Journal of the American Medical Informatics Association.
AutoTriage is a patient stability prediction algorithm that generates patient stability scores from diverse patient measurements. The most common clinical decision support (CDS) risk scores can misidentify the stability of up to 60% of patients, preventing timely patient transfer to appropriate care. When compared to the Medical Emergency Team (MET) calling criteria and the Modified Early Warning Score (MEWS 4) scoring system, both of which are in clinical use, the AutoTriage discharge recommendation tool identified three times as many stabilizing patients than the existing CDS tools. AutoTriage also achieved an accuracy of 93% while maintaining 94.5% specificity. Improvements to current CDS risk scoring systems were made by considering the interdependence between risk factors, as well as their changes and trends over time.