Dascena presents at annual Congress of European Society of Intensive Care Medicine
Unplanned ICU readmission is associated with increased mortality, morbidity, and healthcare costs. Though early readmission is difficult to predict, Dascena collaborated with Cambridge University Hospital to apply machine learning techniques to this problem. Transfer learning was applied to publically available datasets used to train Dascena’s proprietary prediction model and tested in two adult ICUs between October 2014 and August 2016. This resulted in better performance than the Stability and Workload Index for Transfer (SWIFT) score. Cambridge University Hospital is a tertiary hospital with ICUs that admit approximately 2,000 patients a year.