New modeling and data from a community testing program are providing a clearer picture of the prevalence of novel coronavirus infections in King County, as well as additional evidence that transmission rates continue to decline thanks to social distancing.

Researchers say the number of confirmed cases in the county — about 6,300 as of Friday — represents the tip of a much larger iceberg. In reality, about 46,000 people have likely been infected since the local epidemic began, according to information released Friday by Public Health – Seattle & King County. That translates into about 2% of county residents.

But the estimate, in a new report from the Bellevue-based Institute for Disease Modeling, comes with an uncertainty range between 15,000 and 108,000.

Another model, from researchers at MIT, estimates more than 105,000 people statewide have been infected, compared with about 14,600 confirmed cases.

The discrepancy between confirmed and total infections isn’t a surprise. Testing has been limited, and data from other areas suggests the actual number of infections could be 10 to 20 times higher than the number of diagnosed infections. But this is the first time scientists have tried to put a ballpark number on total possible infections locally.

The modelers used death data and information about the virus’ lethality to simulate the hidden transmission of disease. They also incorporated information from SCAN, a surveillance program that allows volunteers to swab their noses at home and send the sample to a central lab for analysis.


Since being launched in late March, SCAN has tested nearly 8,500 samples, almost two-thirds from people with COVID-like symptoms, and a third from people who feel fine. Sixty-seven of the sick people tested positive, along with five people with no symptoms.

“Thanks to the more than 8,000 volunteers who’ve enrolled in SCAN to-date, we’re identifying cases that would not be captured through traditional testing methods,” Dr. Jay Shendure, scientific director for the Brotman Baty Institute at the University of Washington and SCAN leader, said in a statement.

Also on Friday, some of Shendure’s colleagues published findings from the project that pioneered the use of at-home swabs.

The Seattle Flu Study had been using the technique for nearly two years to track the spread of flu and other respiratory diseases when the coronavirus first landed Washington. They began testing swabs for the new pathogen, and found it in about 1.1% of their specimens, UW virologist Dr. Helen Chu and her colleagues reported in the New England Journal of Medicine.

The results revealed that the virus had been circulating for weeks, undetected by traditional public health surveillance programs.

“The key take-away is that a strategy of home-based swab collection and testing is a way to find community cases of coronavirus and is something that is potentially scalable for testing in the community,” Chu said.


The UW team and others have been working to further validate the home-swab kits with the hope of gaining federal approval for more widespread use. Some of the work is funded by the Bill & Melinda Gates Foundation, which also wants home testing expanded.

“We’re at the point now where we’re hoping to reopen our society,” Chu said. “One of the key things is being able to find cases early and identify places where there might be clusters and to stop that transmission.”

Another take-away from Friday’s updated modeling results is that transmission rates have continued to fall, reflecting the impact of social distancing measures, said Mike Famulare, of the Institute for Disease Modeling. But if it’s true that 2% of King County residents have been infected so far, that means the vast majority of people remain susceptible to the virus.