Virginia Tech professor models the role of testing for COVID-19 outbreaks
To Lauren Childs, assistant professor of mathematics, in the Virginia Tech College of Science, it’s what we don’t know that makes it difficult for communities to prevent the spread of the coronavirus.
One big unknown is that evidence indicates that people who do not have symptoms can transmit the virus for several days before symptoms show up.
“If symptoms are associated with transmission, one can avoid anyone with symptoms and be nearly risk-free,” Childs said. “If symptoms are not a good way to identify who is transmitting, then testing becomes much more important.”
Childs had been following reports released by government health agencies around the world, and she realized that communities would need much more information about who was being tested, who was not, and how many of the untested could be carriers of the virus if we are to have any hope of containing the pandemic. The National Science Foundation awarded Childs, a mathematical modeling expert focused on public health, with a $180,000 Rapid Response Research (RAPID) grant to try to fill in the gaps.
“A key aspect of all the interventions we are using – whether it’s contact tracing, quarantine, travel restrictions, or social distancing – is knowing who is infected,” Childs said, “and that’s something that can only be determined by testing.”
Childs has extensive experience building computational models to analyze infectious disease and will work with mathematics doctoral student Melody Walker to develop, analyze, and simulate the COVID-19 models. Initially, they will work with a publicly available dataset from Iceland, which is updated daily to include such information as the number of tests performed, positive tests, and proportion of tests from people in quarantine.
Childs’ aim is not to reveal rates of the virus’s spread, but to help a country, state, or other jurisdiction determine the most effective ways to staunch the spread of the virus based upon their particular testing strategy and rate of testing. For instance, if a community is testing only individuals with symptoms, what actions should it take in terms of school closings, social distancing, and usage of masks, and which combinations would be most effective? If it takes five days to get back test results, as opposed to one or two days, how should a community adapt its strategy?
Though Iceland’s test results are significant, she will also layer in daily case counts and deaths available through Johns Hopkins University, which includes such sources as the World Health Organization and Centers for Disease Control.
A new source of information about who may be infected with the COVID-19 virus without knowing it has recently emerged: people undergoing medical procedures that were previously determined to be “nonessential” and postponed.
“Now that nonessential medical procedures are being rescheduled and in most cases a COVID test is required beforehand,” Childs said, “we are seeing more positive results from asymptomatic people.”
One end result of the NSF RAPID grant will be to create a publicly accessible dashboard where you can piece together the types of testing, sensitivity and specificity, results, turnaround time, and other factors. One region may have tests that are 99 percent accurate, yet take five days for results. Another may have a test that is 80 percent accurate, but gets results in 15 minutes. How should their mitigation strategies differ? A health district could look at Childs’ model and “use that to make informed decisions.”
The NSF grant has a one-year timeframe, and Childs and Walker plan to work quickly.
“Obviously,” she said, “it’s important that we get going on it as soon as possible.”
In immediate response to the COVID-19 pandemic, Virginia Tech faculty, staff, and students have initiated numerous research projects with local and global salience. Learn more from the Office of the Vice President for Research and Innovation.