When the coronavirus first started spreading on US shores, no one knew what to expect. But hidden signals — Google searches, virus levels in waste water, ER visits — foreshadowed the coming onslaught.
Now the US Centers for Disease Control and Prevention is seeking to marshal the researchers who detected these signals and develop a national network that can forecast infections just like the National Weather Service warns about hurricanes.
“This is one of the great things that the CDC is investing in. It is so, so deeply needed,” said Dr. Megan L. Ranney, dean of the Yale School of Public Health, who is not involved in the project. “In public health, we don’t currently have systems that are the equivalent of satellites and radar. I’m hoping this will lead to that kind of equivalent capacity for infectious disease outbreaks.”
The CDC, through its new Center for Forecasting and Outbreak Analytics, is doling out some $262.5 million over five years to 13 centers around the country, including to researchers in Massachusetts.
The recipients will work to refine and test the algorithms and methods that harvest data from a variety of sources and craft projections of what the future holds. Significantly, however, some of the money is going to state and local public health authorities, including the Massachusetts Department of Public Health, to build relationships with academia so that information can flow seamlessly in a crisis.
The goal is not necessarily to detect the next new microbe that will cause a pandemic, but to have the tools ready to inform the response when that new pathogen arrives.
During what the researchers call “peacetime,” the lull between pandemics, they hope to hone the systems that will make accurate forecasting possible. Meanwhile, they expect to make themselves useful in predicting how bad the flu season will be in the winter, or how quickly ticks are spreading disease in the summer.
Working closely with public health officials will be a big focus of the project, said Nicholas Reich, professor of biostatistics at the University of Massachusetts Amherst. UMass and the University of Texas at Austin are together receiving $27.5 million to jointly serve as one of the 13 centers.
Reich runs the US COVID-19 Forecast Hub, which provides weekly short-term forecasts of data such as hospital admissions and deaths at the state and county level, based on a variety of sources such as waste water and mobility data.
The new money, Reich said, funds both the academic teams and the public health officials, enabling them to have dedicated staff who can work with the researchers and who understand the models.
Reich said he has collaborated with the Massachusetts Department of Public Health in the past but this new funding will deepen the relationship. “One of things we’re actively working on is thinking about looking at ways to evaluate forecasts in the context of the decisions they’ll be used for ... trying to figure out what are the specific questions these models can help inform,” he said.
Northeastern University will receive $17.5 million to establish “EPISTORM: The Center for Advanced Epidemic Analytics and Predictive Modeling Technology.”
Alessandro Vespignani, director of the Network Science Institute, will lead the project at Northeastern.
His team will coordinate a consortium of 10 research institutions including Boston University, health care systems, and private companies that will draw data from hospital admissions, waste water surveillance, social media, and other sources, as well as individuals’ travel patterns, and develop algorithms to project where and how fast an infection is spreading.
“As soon as we get a signal, something suspicious, then we can spin the algorithm and understand what kind of threat it is, what is the risk,” and thus inform decisions about how to contain the outbreak, Vespignani said.
Vespignani said that long before anyone knew how fast COVID-19 would spread, he looked at the number of cases that had appeared outside of China in the first few weeks of the pandemic, and calculated that the virus had already spread to thousands in major US cities.
At the same time, Vespignani’s colleague, Mauricio Santillana, noticed a spike in social media searches for things like “sore throat” as well as an increase in doctors browsing physician-specific information sources such as UpToDate. Digital signals like these often precede by weeks the recognition that an outbreak has occurred, said Santillana, who directs Northeastern’s Machine Intelligence Group for the betterment of Health and the Environment.
“We had confidence that COVID was in our communities despite the fact that we were only testing people who came from China,” he said.
Another group in the network is based at the Bloomberg School of Public Health at Johns Hopkins University, which received $23.5 million.
Among the Johns Hopkins group’s projects will be recruiting a random sampling of citizens to answer surveys and provide blood samples, enabling researchers to cast a wider net as they search for evidence of infections in people other than just those who have sought medical care.
The goal is to improve the ability to make both short-term and longer-term projections, to guide decisions on such matters as infection-control measures and vaccine distribution, said Shaun Truelove, an assistant scientist in the Departments of International Health and Epidemiology at Johns Hopkins.
“This network is a really important step toward building the capacity in the US to be able to be more prepared and to develop the tools and the capacity to be better at responding to these health emergencies,” Truelove said.
Dr. William Schaffner, professor of preventive medicine at the Vanderbilt University School of Medicine, agreed. “This is exactly the sort of innovative, outside the box — or push the edges of the box — initiative that the CDC ought to be doing and I’m so glad they are,” said Schaffner, who is not involved in the project. “We would always like to know at the earliest possible time when something new is happening, whether locally or nationally. That just gives you more time to prepare and to muster your resources.”