Dascena publishes on a sepsis randomized clinical trial
Dascena carried out the first randomized controlled trial of a machine learning algorithm for sepsis surveillance that resulted in statistically significant decreases in hospital length of stay and in-hospital mortality. The outcomes are published in BMJ Open Respiratory Research.
In this single-center study conducted in the ICU at UCSF Medical Center in Parnassus Heights, patient outcomes in the experimental group, which was monitored by the machine learning algorithm, were compared to those in the control group, which was monitored by a pre-existing sepsis detector. At the end of the study, the average length of stay in the experimental group was 20.6% lower than that in the control group (p=0.0421). In-hospital mortality decreased by 58.0% in the experimental group (p=0.018). These results suggest that the machine learning-based sepsis prediction techniques may lead to improved patient outcomes and earlier clinical interventions.