Dascena (Oakland, CA) received FDA breakthrough device designation for its Previse machine-learning algorithm to predict acute kidney injury (AKI). The Breakthrough Devices Program provides a prioritized review of a device submission to the FDA and lets manufacturers interact with agency experts to address topics as they arise during the premarket review phase.
The algorithm uses values of various heart rates, respiratory rate, temperature, serum creatinine, Glasgow Coma Scale score, and patient age to predict likelihood of AKI. The company demonstrated that its algorithm predicted AKI more than one day before patients would meet the clinical criteria for diagnosis, which is based on changes in serum creatinine, urine output, or both, according to the National Center for Biotechnology information. In validation studies, Previse demonstrated higher sensitivity and predictive value than a clinician’s assessment based on clinical criteria.