Blood samples to test for covid-19 antibodies sit on a table at a clinic in Moscow on May 15, 2020.

Blood samples to test for covid-19 antibodies sit on a table at a clinic in Moscow on May 15, 2020.
Photo: Getty Images

The U.S. Centers for Disease Control and Prevention, along with several states, are muddling the statistics on covid-19 testing, according to reporting from The Atlantic and other media outlets. They have combined the results of two different types of tests in their tallies released to the public, only one of which is used to confirm an active infection. The conflation, experts say, could make states appear safer to reopen than they actually are.

PCR tests are the standard diagnostic test for covid-19. These tests, which require a swab or spit sample, look for the presence of RNA left behind by the new coronavirus. Another type of test looks for antigens produced by the virus, but it has only recently come into existence and isn’t widely in use yet. Though these diagnostic tests have their limitations, they’re expressly designed to confirm an active case of covid-19.

Scientists have also created tests that look for the antibodies our immune system produces in response to the coronavirus. Antibodies specific to the virus can be found while someone is still sick, but they shouldn’t be used to diagnose an active case. Rather, a positive result likely indicates that someone survived an infection in the recent past.

While both types of testing are crucial to tracking covid-19, they have very different purposes. Antibody testing can be used to judge how widespread covid-19 has become in a population over time and may in the future be used to determine someone’s immunity to it. Diagnostic testing, on the other hand, helps us figure out the size of a current outbreak in an area; on a smaller scale, it can also help public health workers contain clusters of illness before they go out of control.

On Wednesday, Florida public radio station WLRN first reported that the CDC was including both antibody and diagnostic test results in its publicly released tally of testing, without distinguishing between the two; it also received a confirmation of the practice from the CDC. Local media outlets also reported a similar trend in test counting in Virginia, Texas, Georgia, and Vermont. On Thursday, The Atlantic reported that Pennsylvania and Maine have been doing the same.

Save up to 54% off Some of Your Favorite Dragon Ball Blu-

At this point, the level of antibody testing performed daily in the U.S. does not appear to be close to the level of diagnostic testing. But combining the results into a single indicator is likely to obscure the reality of the pandemic, experts have said. That’s partly because states plan to reopen segments of business based on meeting a certain threshold of daily testing and a declining rate of residents testing positive for covid-19. Since antibody testing is meant for the general public, not only those who suspect they’re sick, it might further skew the picture in favor of reopening.

“How could the CDC make that mistake? This is a mess,” Ashish Jha, a public health researcher and the director of the Harvard Global Health Institute, told The Atlantic.

Virginia stopped combining its tests last week, following reporting of the practice by The Atlantic and others. Vermont similarly stopped as well, though its governor Phil Scott claimed that officials were only reporting the data they had gotten from testing sites and were unaware of the conflation. New York State, the hardest hit by covid-19 so far with over 300,000 reported cases, does not include antibody test results in its public covid-19 tracker, a representative for the New York State Department of Health told Gizmodo.

The CDC, however, told The Atlantic that it is still working on separating the different types of testing in its reporting and hopes to do so within the next few weeks.