CU Boulder researchers study sniff-test as possible COVID-19 screening alternative

Posted at 6:56 PM, Dec 12, 2020
and last updated 2020-12-12 20:56:31-05

BOULDER, Colo. — University of Colorado Boulder researchers developed a COVID-19 test that could be cheaper and more effective than the the traditional PCR test.

According to Roy Parker, a professor of biochemistry and director of CU’s BioFrontiers Institute, a scratch-and-sniff test may be a better way to control infections and would only cost 50 cents per test.

"One of the things we've been doing is trying to think about ways that you could do screening for people who are affected by COVID, but using symptoms instead of testing for the virus," Parker said.

Parker and Dan Larremore, an assistant professor of computer science, teamed up with CU Boulder alumnus Derek Toomre, a professor at the Yale School of Medicine to develop an index-card sized test that interacts with a cell phone app to assess sense of smell.

"We're all familiar with the idea of testing for fever, but fever doesn't really work because not that many people who are infected with COVID actually get fevers, maybe only 20-25%," Parker said. "Maybe 40-50% of people notice that they lose their sense of smell, but if you actually test people for their sense of smell, then 80% of people roughly actually have defects in their ability to smell when they're infected with COVID."

Fever is associated with many diseases and may only last a day or two. Loss of smell without a stuffy nose is highly specific to COVID-19 and often lasts for a week or more.

“A lot of people have joked about this idea, but this is the first effort to ask in a rigorous, mathematical way: Could screening for loss of smell actually work?” Parker said. “We were surprised by how good the results were.”

Toomre, who graduated with a bachelor’s degree in biochemistry from CU Boulder in 1990, recently launched a company called usmellit to commercialize the test. It features five scratch-and-sniff squares and asks users to identify the five scents and enter the answer into their cell phone. It either tells them they passed or instructs them to get a test.

The team used mathematical modeling to predict the effect of smell-testing in several different hypothetical scenarios, like on a college campus of 20,000 people where individuals were tested once a week, every three days or daily.

In general, testing for sense of smell every three days worked better than weekly PCR tests in curbing infection, according to the group. The sniff-test isn't necessarily a definitive diagnostic test, but can be used as a screening tool.

The test isn't available commercially, but Toomre applied for Emergency Use Authorization from the Food and Drug Administration in October. The company could potentially produce hundreds of millions of tests per week, if approved. He'd also donate one card to a charitable nonprofit for every one purchased.

This way of testing would be much cheaper as well. Even at a cost of $20 per PCR test, it would cost nearly $3 million to suppress an ongoing outbreak in a population of 20,000 people using weekly PCR tests. Providing smell tests every three days for that same population would cost $200,000 and catch more infected individuals before they spread the virus, they found.

“We hope that it would allow for a fast and easy test that anyone can take anywhere and can be applied to anyplace that is using temperature screening as an inexpensive way to help avoid shutdowns, keep businesses open and decrease the terrible human toll,” he said.

The next step, Parker said, is finding a group of people to run a trial. They'll ideally need thousands of people they can test every day to see if it would lead to the prevention of COVID-19 transmission within that community, using their sniff-test. Parker is hoping it will happen in the next few weeks.

The paper on their research has not been published in a peer-reviewed journal yet and more research is necessary before the sniff-tests could be broadly rolled out as a screening tool.