Project Summary Local solutions rarely scale to multi-institutional settings, and don’t realize the vision of CTSAs catalyzing data- driven translational research. Data, Software, and People are Largely Decentralized and Unconnected in CTSA. CTSAs hold a wealth of data that is neither accessible nor interoperable; this hinders opportunities to innovate algorithms and tools. Tools in turn, suffer from myopic focus or blind duplication. It is extremely difficult to identify expert partners whether for functionally validating a candidate variant for a rare disease, finding clinical cohorts, or navigating the IP odyssey of starting a company. There are the challenges in building a truly learning health system where high quality clinical and research data can effectively and efficiently be reused in clinical care. The CD2H has developed the National COVID Cohort Collaborative (N3C) with COVID data from clinical institutions across the country. This publicly available data resource includes data on adults and pediatrics and can address the questions related to vaccine effectiveness, treatments and predictions of severity of illness. The Pediatric COVID-19 DREAM Challenge will ask community participants to predict severity status in children, using the National COVID Cohort Collaborative (N3C) Electronic Health Record (EHR) data. The desired outcome for this challenge is to produce trained and validated pediatric COVID-19 severity prediction models that can be implemented into a clinical workflow in an EHR to help facilitate appropriate treatment.

Project Narrative The National Center for Data to Health (CD2H) is an informatics coordination effort to generate methods, technologies and approaches to support network research across the CTSAs. The Pediatric COVID-19 DREAM Challenge will ask community participants to predict severity status in children, using the National COVID Cohort Collaborative (N3C) Electronic Health Record (EHR) data. The desired outcome for this challenge is to produce trained and validated pediatric COVID-19 severity prediction models that can be implemented into a clinical workflow in an EHR to help facilitate appropriate treatment.