Rapid antigen tests for coronavirus will no longer be labeled as 'inconclusive' by Iowa

Iowa will start counting the results of a relatively rapid type of coronavirus tests, which will increase the reported rates of infection in some counties, Gov. Kim Reynolds announced Thursday.

Antigen tests, which look for molecules on the surface of the virus, can produce results in hours. They aren't quite as accurate the tests used by the state's Test Iowa program, known as PCR tests, which look for pieces of the virus but can take several days to produce results.

"It does have a lower level of reliability at this point," but federal experts have changed their guidance on how to use the antigen test's results, Reynolds said at a Thursday news conference. 

The Iowa Department of Public Health had been classifying the results of antigen tests as "inconclusive," so people who came up positive on the quick tests were not being included in official reports of the virus' spread. That is changing, and some counties' reported coronavirus rates will increase because of it, Dr. Caitlin Pedati, the department's medical director, said at the news conference.

The most dramatic statistical affect will come in Van Buren County, where a relatively large number of antigen tests have been performed. That county's reported 14-day positivity rate would increase from 6.6% to 11.6% because of the change, Pedati said. The positivity rate shows how many positive test results were returned among all people who were tested for the virus over a given time period.

More Iowa clinics, hospitals and employers have been using antigen tests in recent weeks, Reynolds said. Pedati said about 10,000 have been performed in Iowa, with about 1,000 of those coming up positive. About 8,000 of the tests have come in the past 12 days, she said. 

Last week, Pedati told the Dubuque Telegraph Herald that the state had been counting "inconclusive" test results in its positivity rate calculations. Health department spokeswoman Amy McCoy said Thursday that antigen test results were included in that category.

Professor Joseph Cavanaugh, chairman of the University of Iowa’s biostatistics department, told the Des Moines Register last week that antigen test results should not be discounted, especially if they’re positive. Doing so could skew infection rates so they appear lower than they are, he said. 

The main issue with antigen tests is they have less “sensitivity” than PCR tests, meaning they are a bit less likely to pick up low levels of the virus if it’s present in a person being tested, he said.

But antigen tests have “nearly perfect specificity,” meaning that if a person comes up positive on the test, they very likely have the virus.

“Discarding the positives is therefore logically contradictory, since the positive results are nearly always accurate,” he wrote in an email to the Register.

Cavanaugh noted that federal experts writing about antigen tests have determined that “false positive results are unlikely.” 

Pedati acknowledged Thursday that false negatives can be an issue with antigen tests, incorrectly indicating an infected person doesn't have the virus. She said if health care providers suspect an infection in a person with a negative antigen test, they should consider ordering a follow-up PCR test. Regardless of test results, she said, anyone who had close contact with a person with a confirmed infection should go into quarantine. 

Pedati said that all along, public health officials have been recommending isolation for people who have positive antigen tests for the coronavirus. 

Tony Leys covers health care for the Register. Reach him at tleys@registermedia.com or 515-284-8449. 

Your subscription makes work like this possible. Subscribe today at DesMoinesRegister.com/Deal.