DENVER (KDVR) – People in Colorado did not stray far from home in accordance with the governor’s stay-at-home order, according to cellphone mobility data analyzed by a group of COVID-19 modeling researchers.
“Broadly, people in Colorado changed their behavior, and I think we see that pretty definitively, and that seemed to have a significant impact on the transmission of the disease,” said Jude Bayham, an assistant professor at Colorado State University and a member of the state’s COVID-19 modeling research team.
“The mask-wearing…(and) remaining vigilant is probably more important now than it was even during the stay-at-home phase,” he said.
Bayham said the group analyzed aggregated, anonymized cellphone data supplied by a company called Safe Graph. The data is collected from people whose location services were turned on over the past several months.
“There is absolutely no way we can identify an individual in here. What we are looking at is broad trends and usage. Even then, the data seemed to capture about 10 percent of the population,” he said. “It is certainly not comprehensive. It’s not tracking everybody all the time.”
Bayham described a “variation” in the rate in which people in different counties responded to stay-at-home orders and noted that the research team is still investigating whether the rural data might be less accurate than other areas.
Denver and Silverthorne, however, were two municipalities that seemed to reduce their activity at a faster rate than Fort Collins and Durango, he said.
“So we see a slower reduction (in Fort Collins and Durango) even though we saw reductions across the entire state,” he said.
Bayham and other research team members are hoping to use the data they analyzed to better understand how COVID-19 is transmitted and what role mobility plays in spreading the illness.
The data does not track whether people are using masks, coughing into their arms or practicing other preventive measures.
“We aren’t trying to predict the future, what we’re trying to do is say, ‘If people behave in a certain way, what would the future look like?’ And, in all of the cases, we expect the future to change,” said jimi adams, an associate professor at the University of Colorado Denver and a modeling team member who said his name is spelled with lowercase letters.
According to adams, the mobility data indicated that people seemed to anticipate the various state health orders both by staying at home prior to the order going into effect and by leaving home more often prior to the expiration of the stay-at-home order.
“On some levels, it’s worrisome because it is people starting to engage in behaviors that have been trying to be reduced…but the flip side of it is… just because people are changing their behavioral patterns in terms of how much time they’re spending away form home or in public places doesn’t mean they’re doing so in ways that are necessarily violating social distancing aims,” he said. “They still could be keeping space. They still could be wearing masks.”
He said the team is in the process of incorporating the mobility data into the epidemiologic models of disease progression that have been created. He hopes the data will provide more clarity on how people who might be susceptible to a disease might encounter an infectious person.
“We know that those are not happening at random. Those are patterned in certain ways, and these mobility data help us to get some empirical estimates of those patterns rather than having to assume everybody is equally likely to come in contact with everybody else ,” adams said.