A new federal center aims to help the nation better forecast future waves of the COVID-19 pandemic as well as outbreaks of other diseases.
Marc Lipsitch, professor of epidemiology and director of the Center for Communicable Disease Dynamics at Harvard T.H. Chan School of Public Health, is director for science for the new Center for Forecasting and Outbreak Analytics. The creation of the center was announced in August by the Centers for Disease Control and Prevention.
In a December 15, 2021 interview on WBUR’s “Here and Now,” Lipsitch discussed the center’s aims—and the challenges it faces.
“In the long run, we want to take the vast amount of data we have, add to that to design ways to improve the flow and quality of data, and to use what are known as transmission dynamic models and forecasting tools to understand what may lie ahead, and to help decision-makers target interventions so that they are as effective as possible, while minimizing the disruption that they cause,” he said.
Lipsitch said that the disease forecasting models will be based on various kinds of data, such as case numbers, hospitalizations, human mobility, and drug prescriptions.
He noted that one of the big challenges in forecasting how outbreaks will play out is modeling human behavior—such as whether people will travel, wear masks, or get vaccinated. “Anything that depends on future transmissions is dependent on human behavior that we can’t be certain about,” Lipsitch said. “That is why forecasts … have an inherent limit to how long they can go.”
Many health departments in the U.S. and around the world have been using forecasting tools during the COVID-19 pandemic to inform their policy decisions, Lipsitch said. Now, he noted, “We are trying to supercharge that by putting it on a scale that’s suitable for a world-class system.”
Listen to the WBUR interview: New CDC outbreak forecasting center could do for disease what weather service does for meteorology
Learn more
Epidemiologists Marc Lipsitch and Rebecca Kahn to help establish new CDC center (Harvard Chan School feature)