Delphi Enhances COVIDcast With Change Healthcare Claims Data
Unique Data Helps Researchers Track and Forecast Pandemic Patterns
- School of Computer Science
Since April, Delphi has been collecting real-time data on self-reported COVID-19 symptoms and other disease indicators nationwide. This county-level information about the coronavirus pandemic is updated continuously and shared with both the public and health researchers. Now, COVIDcast is taking a further step by adding de-identified COVID-19 claims from Change Healthcare to its unique combination of survey, testing and mobility data.
"Tracking and forecasting the spread of a novel disease such as COVID-19 is a challenging task that requires new types and sources of data," said Ryan Tibshirani, an associate professor of statistics who co-directs Delphi with Roni Rosenfeld, head of the Machine Learning Department. "We are always evaluating our data streams and looking at ways of filling gaps in our knowledge. Change Healthcare has stepped up in the biggest way possible to give us crucial information for understanding the current state of the pandemic. We are extremely appreciative of their contributions to our effort."
As a preeminent research university, Carnegie Mellon leverages its leadership in machine learning, statistics and data science, as well as its partnership with the U.S. Centers for Disease Control and Prevention, to offer a more complete and real-time picture of COVID-19 than previously possible. Its Delphi Research Group works to advance the theory and practice of epidemic forecasting. Now enhanced with Change Healthcare's de-identified claims, COVIDcast helps researchers better understand the pandemic's impact at the community level by providing comprehensive, geographically detailed, continuously updated indicators of pandemic activity. Understanding the impact of the pandemic on individual communities helps researchers learn patterns and identify trends that will make it possible to improve patient treatment within each community. The data used by COVIDcast is de-identified in accordance with HIPAA privacy regulations.
"Delphi group's vision is to make epidemiological forecasting as universally accepted and useful as weather forecasting," said Tim Suther, senior vice president of Data Solutions at Change Healthcare. "The addition of de-identified COVID-19 claims from Change Healthcare allows COVIDcast to present a more complete, multidimensional picture of the pandemic and its impact."