The TRACE model shows that relaxing social distancing with current national capacity for testing and contact tracing “flattens the curve,” while a more gradual relaxation of social distancing combined with expanded testing and quarantine policies can yield true suppression. (Figure courtesy Ross Hammond/Brookings Institution)

Several months into the biggest public health crisis in a century, policy-makers in the U.S. and around the world are trying to figure out when and how to begin reopening businesses and public spaces. But with so many uncertainties remaining around the coronavirus, including how many people have it, how contagious asymptomatic people are, and whether those who have recovered are immune, it can be difficult for officials to make informed decisions. And the stakes could not be higher: Reopening too soon, or without adequate precautions, could cause a surge of contagion and more deaths.   

SFI External Professor Ross Hammond, a senior fellow in Economic Studies at the Brookings Institution, along with collaborators at Washington University in St. Louis, Missouri, has developed a new agent-based computer model that helps policy-makers wrangle with that uncertainty. The simulation model, called Testing Responses through Agent-based Computational Epidemiology, or TRACE, incorporates both knowns and unknowns to produce multiple variations for any particular policy option.  

“Unlike many models, we really try to embrace that uncertainty,” Hammond says. “We’re trying to determine which policies are a good idea to pursue, and we want policies that are robust to that uncertainty.”

As an agent-based model, TRACE can incorporate variations in age, contact networks, activity patterns, and likelihood of infection, all critical factors in determining how to contain COVID-19. 

For example, a state that’s considering focusing on testing and contact tracing to contain the virus and ensure a safe reopening may want to know how extensive testing would need to be if asymptomatic carriers of the virus are contagious, versus if they are not. Or officials may want to know how accurate tests would need to be, or what kinds of social distancing measures might still be needed to minimize risk as businesses reopen. Whatever the question may be, the model, which simulated 10,000 different combinations of factors, has a simulation for that. 

With such precise modeling of complex policy scenarios, the TRACE project, which Hammond describes as a “ policy laboratory,” can help decision-makers set effective policies that go beyond “flattening the curve” and instead actually suppress the virus while allowing for a gradual relaxing of social distancing. 

The next step for the new model, unveiled in May, is applying it to real-world decisions. 

“We’ve been in touch with policy-makers ranging from U.S. states to the White House committee on reopening to foreign governments around the world,” Hammond says. “They’re all interested in this model.”    

Read more about the TRACE model, and view the interactive dashboard.

Read Hammond's blog post, "Developing policies for effective COVID-19 containment: The TRACE model" (May 15, 2020)