Launched earlier this year in response to the coronavirus pandemic, the EHR COVID-19 DREAM Challenge is designed to use electronic health record (EHR) data to to answer the question: Of patients who had at least one encounter/visit at University of Washington Medicine and who had a COVID-19 test, can we predict who tested positive? Due to the importance and emergent need for a better understanding of the virus and for the development of patient-specific risk scores and early warning tools, the Center for Data to Health (CD2H) developed a platform to support analytic and machine learning hypotheses on clinical data to rapidly discover and implement approaches for care
The COVID EHR dataset was updated on August 6th with additional patient data from UW’s EHR. (Models previously submitted still function in the updated infrastructure.) Several useful resources, including a synthetic dataset and a data dictionary, are available for scientific collaboration. The data will be a valuable part of the National COVID Cohort Collaborative (N3C) Enclave that is scheduled to open for community research this summer. The DREAM Challenge’s goal is to make biomedical research more transparent, more reproducible, and more accessible to a broader audience of scientists. CD2H and N3C thank all participants for their interest in this continuous benchmark project!