Virginia Tech team building infectious disease models, research to help guide decisions
August 31, 2020
A multidisciplinary team of Virginia Tech faculty, students, and analysts has spent the summer building models to help the university understand how COVID-19 could impact the campus.
The group’s short-term goal is to apply their expertise in infectious disease modeling to help inform decision-making about how Virginia Tech could deliver academic programs this fall. The group’s longer-term goal is to expand the university’s research and education programs in emerging and zoonotic infectious diseases.
Led by Laura Hungerford, head of the Department of Population Health Sciences, and Ron Fricker, professor of statistics and senior associate dean in the College of Science, the modeling group is helping university and community leaders better understand and plan for effects of the COVID-19 pandemic through institution-specific network and epidemiologic models, outbreak simulations, and impact predictions on hospital resources.
“It is important that we understand that in the absence of nonpharmaceutical interventions (NPIs), such as physical distancing, wearing face coverings, and contact tracing, COVID-19 is projected to spread rapidly,” said Executive Vice President and Provost Cyril Clarke. “The models developed by our team of experts confirm this widely held assumption. On the other hand, the models also show that if NPIs are implemented effectively and applied consistently, it is possible to slow the spread of COVID-19 to manageable levels. This would allow Virginia Tech to continue offering students classroom instruction and campus experiences to the greatest extent possible while mitigating the risk of a future outbreak.”
Initial model outputs show that without extensive mask wearing, social distancing, and other NPIs, both on and off campus, even just a few infected students could result in a widespread outbreak this fall on campus and in the surrounding community.
“It is crucial that we come together as a community if we are going to be successful,” said Hungerford. “Places like New York City have shown that this can be done, but only if everyone does their part. But as we have seen at some other universities, it only takes a few bad choices to cause major outbreaks.”
In building their models, the group had to confront the problem that there is still a lot unknown about COVID-19, which results in a lot of uncertainty in the models.
“We have incorporated the lastest scientific information in our models,” said Fricker. “What we have learned is that a variery of outcomes are possible, including having a successful semester. However, the models are consistent in one thing: success is hard – we all have to work together and everything has to go just right – and failure is easy. But that doesn’t mean we should accept failure. Just the opposite – let’s do the hard work to maximize our chances of success.”
According to the group and its predictive models, a successful fall semester is critically dependent on being able to do very fast testing, tracking, and quarantine at scale. Models show that in the absence of fast testing and effective contact tracing, the university is likely to have a large-scale outbreak and that it should be ready to shift to all online (and back) as needed if there is an uncontrolled outbreak. The university has used these insights from the group to work to improve its testing, tracking and quarantine capabilities. Virginia Tech is also closely coordinating with the New River Valley Health District to establish effective contact tracing and case management capabilities.
One thing the models clearly show is lowering the reproduction number, which is the average number of secondary cases produced by a single infection in a susceptible population, is critical, and this requires close adherence to NPIs. To be successful, Virginia Tech will need complete buy-in by all members of the university community whether on or off campus.
“The science is becoming clearer by the day – wearing a mask and social distancing works in decreasing the spread of COVID-19,” said Fricker.
“Wearing a mask is Ut Prosim in action,” Hungerford said. “It says, ‘I care about my fellow Hokies.’”
The COVID-19 Modeling Group will continue to provide updates and model outputs to university leaders through the fall semester.