UT Launches New Flu Forecasting Center With Support From CDC
The center, COMPASS, aims to improve flu forecasts to save lives and reduce strain on the healthcare system.
Every winter, hospitals brace for the flu without knowing exactly when it will hit or how hard. A new flu forecasting center at The University of Texas at Austin aims to give them a head start.
The U.S. Centers for Disease Control and Prevention (CDC) has awarded a team led by Lauren Ancel Meyers, an integrative biology professor and director of the Center for Pandemic Decision Science, a five-year cooperative agreement to create the Center for Outbreak Modeling, Prediction, Analytics and Surveillance Science. The center known as COMPASS is one of six new centers in the CDC’s Network of Influenza Modeling and Forecasting Centers.
Building on two decades of collaboration with the CDC and public health agencies, COMPASS will improve flu forecasting models — both for seasonal flu and for a possible future flu pandemic — to help save lives and reduce the burden on the healthcare system.
“It’s the same thing we do for weather, but for flu,” Meyers said. “Fortunately, unlike the weather, we can actually change the course of an epidemic. If we know a major surge is coming, we can take steps to slow the spread and prevent hospitals from being overwhelmed.”
Flu forecasting models can help predict when the next season will start, how fast cases will rise, when it will peak, how big the peak will be, and how hard it might hit hospitals, schools and businesses. In response, the healthcare sector can surge staffing and shift the timing and availability of vaccines, antiviral drugs and at-home tests. Community leaders can encourage behavioral changes to slow the spread and take special precautions for elderly people and others at higher risk.
COMPASS is co-led by Meyers and Spencer Fox, an assistant professor at Northern Arizona University, formerly a student and researcher at UT.
Seasonal influenza causes tens of thousands of deaths in the U.S. each year, while the threat of another deadly influenza pandemic remains ever present.
“It’s a significant threat to older people and those with underlying health conditions. It imposes a huge burden on our hospital systems,” Meyers said. “Better forecasts can help us act earlier to protect people and reduce that burden.”
Meyers and her team will improve their flu forecasting models in three ways. First, the team will incorporate more complex dynamics than earlier models do — including multiple viral strains circulating at the same time; continuously evolving viruses, including immune escape and antiviral resistance; rising and falling immunity in individuals and groups over time; who gets vaccines and antivirals and when; and geographic factors influencing transmission. Second, the researchers will integrate novel data streams. And third, Meyers and her collaborators will explore the use of artificial intelligence to find previously overlooked patterns in flu data that can help inform the models — potentially even creating hybrid models that both follow the strict mechanistic rules of a traditional forecasting model and use AI to detect subtle patterns that can help modulate forecasts.
“Models can help us design better interventions because they allow us to virtually conduct experiments that would be unethical or impractical in the real world,” Meyers said. “They allow us to compare different strategies for using vaccines, antivirals, and other interventions to identify the most effective ways to slow transmission and save lives.”
The COMPASS team will also expand beyond seasonal flu forecasts to explore different possible scenarios for a future flu pandemic and conduct “war gaming” to find the most effective strategies for mitigation, essentially developing a flu pandemic playbook.
After the challenges of the COVID-19 pandemic, Meyers shared a hopeful perspective on a possible future flu pandemic. COVID taught us important lessons about the benefits and costs of interventions such as masking, isolation, and school and business closures — knowledge that can help us prepare a more effective, less disruptive response.
“If a new influenza pandemic emerges, it could take months before an effective vaccine is widely available,” Meyers said. “But unlike at the start of COVID, we already have antiviral drugs, and our models can help us plan how best to use them. That could give us a head start — a better chance to save lives while keeping schools and businesses as open as possible.”