Ohio’s governor says his flawed virus test shouldn’t undercut new, rapid methods
By Katherine J. Wu
Gov. Mike DeWine of Ohio, who last week tested positive for the coronavirus, then negative and then negative again, said on CNN on Sunday that his roller-coaster ride should not be reason for people to think “that testing is not reliable or doesn’t work.”
His first test result was positive, when he was screened with a rapid testing method Thursday before President Donald Trump arrived in Ohio for campaign appearances.
DeWine was given an antigen test made by Quidel, one of two companies that have received emergency use authorization from the Food and Drug Administration for coronavirus antigen tests.
These tests, while fast and convenient, are known to be less accurate than PCR tests, which were used to retest DeWine twice Thursday and once more Saturday. All three PCR tests turned up negative, confirming that DeWine was not infected with the virus.
“I don’t think that DeWine’s results were surprising, per se,” said Andrea Prinzi, a clinical microbiologist and diagnostics researcher at the Anschutz Medical Campus in Colorado. “We know that the performance of antigen testing is not as accurate as PCR testing.”
The Ohio governor’s experience, however, may raise concerns about how much states will rely on antigen tests as they seek to augment the forms of testing, like PCR, that are in short supply or that are mired in laboratory backlogs, unable to generate results in a timely fashion to help assess caseloads and dole out treatments.
DeWine, a Republican, is one of seven governors who announced last week that they were banding together to buy 3.5 million rapid coronavirus tests, including antigen tests, to ramp up production.
Daniel Tierney, the press secretary for DeWine, noted in an email that the states involved were considering “multiple companies and multiple testing types,” but did not specify further.
On Sunday, DeWine said he had already been in touch with Gov. Larry Hogan of Maryland, a Republican, to talk about the states’ agreement to use their collective “purchasing power” for testing and other supplies.
“If anyone needed a wake-up call with antigens, how careful you have to be, we certainly saw that with my test,” DeWine said. “And we’re going to be very careful in how we use it.”
A spokesman for Hogan, Michael Ricci, echoed that sentiment: “We are taking this one step at a time.”
Accurate test results are crucial for curbing the spread of disease. False positives, like the one DeWine received, can set off an unnecessary period of self-isolation, depriving people of access to their workplaces or their own families. False negatives, on the other hand, can hasten the spread of disease from unwittingly infected people.
PCR tests like the ones used to determine DeWine’s health status are often the best bet for avoiding incorrect results. But these tests are in short supply nationwide as manufacturers and laboratories struggle to meet the increase in demand that has accompanied recent surges in infections. Turnaround times for results have stretched past two weeks in some parts of the country, rendering the information useless for anxious people who need to know their status immediately so they can self-isolate as needed and stop the virus from spreading further.
“Honestly, PCR tests were not designed for this type of mass screening/testing,” Prinzi said. PCR tests, she added, function best in laboratory environments that are well-stocked with chemicals, high-tech machines and specially trained personnel. Their high-maintenance ingredient lists and relatively hefty price tags are not terribly compatible with quickly getting answers to large numbers of people.
Rapid tests, on the other hand, could catch a majority of active infections if administered frequently, even if they’re less accurate, many experts have argued.
Compared with PCR tests, antigen tests are more likely to return a false negative result, mistaking an infected person as virus-free. Quidel’s test, for instance, can miss up to 20% of the cases that PCR detects.
Notably, DeWine’s antigen test produced the opposite error: a false positive that incorrectly indicated he had been infected.
But DeWine might not have been the ideal candidate for an antigen test, said Karissa Culbreath, scientific director of infectious disease, research and development at TriCore Reference Laboratories in New Mexico. Such tests usually perform better on samples that contain high levels of virus, which tend to come from sicker patients and people at higher risk of transmitting the infection. When given within the first five days after coronavirus symptoms start, Quidel’s false negative rate may drop below 5%, according to the company’s intended use statement.
DeWine, however, had not experienced symptoms, aside from a headache.
“If we’re testing outside of that intended use, we might expect false positives or false negatives,” Culbreath said, referring to the five-day window that follows the onset of symptoms.
Allocating tests to people who fit that criteria, she added, will also eliminate the need for scores of follow-up tests, especially while many suspected cases across the nation remain undiagnosed.
“Tests are not interchangeable in their usefulness,” Culbreath said. “We need to look at this as a tool belt and identify the right tool for the job.”
On Sunday, DeWine did note that antigen tests function especially well as “screening” tests, expediently delivering information to people while their results are confirmed — if necessary — by the more accurate PCR tests.
He added that it was incumbent upon the companies developing the tests to demonstrate their accuracy, and that the experience would not deter him from expanding testing in his state.
“We could use additional money for testing,” said DeWine. “We have doubled our testing the last four weeks. We need to double it, and then double it again.”
Ohio was among the first states to reopen in May, but as cases ticked up in mid-June and July, DeWine signed a statewide mandatory mask mandate and asked several counties to limit gatherings of any size. There have been at least 99,969 cases and 3,668 deaths in Ohio since the beginning of the pandemic, according to a New York Times database.
The status of testing in the United States is far from ideal, Prinzi said. But for now, it is time to make do with the materials we have, she said. “We can argue about diagnostic accuracy all day, but this is a huge public health crisis right now,” she added.
Flaws and all, antigen tests are “a necessary part of our management of the pandemic,” Culbreath said. “But we have to be very intentional about how we use these tests.”