HOUSTON–When health officials quietly removed nearly 3,500 COVID-19 cases from the official Texas total on Wednesday, it launched a deluge of conspiracy theories about inflated and unreliable data in the midst of a surging pandemic.
The 3,484 removed cases were diagnosed using FDA-approved antigen tests. The FDA has said positive results from antigen tests are “highly accurate,” and can be used to diagnose current COVID-19 infections. But state health officials pointed to the definition of a coronavirus case the CDC published in early April to explain why the cases were removed.
“The case data on our website reflect confirmed cases, and cases identified by antigen testing are considered probable cases under the national case definition,” said Chris Van Deusen, a spokesman for the Texas Department of State Health Services.
Under that definition, the CDC only considers cases “confirmed” if they are diagnosed using a molecular, often called PCR, test. Cases that are detected using antigen tests are classified as “probable.” If someone is diagnosed with an antigen test, Texas will not count their case among the state total.
The removed cases were from Bexar County, which includes San Antonio. The city’s mayor said Thursday that San Antonio was one of three cities in Texas that tracks antigen tests—and that the tests help local health officials “see the full picture” of COVID-19 in the area.
“The state wants an apples-to-apples comparison with all cities in their reports, so they're removing antigen counts,” said Mayor Ron Nirenberg. “It now means the State of Texas will not be counting thousands of FDA-approved positive COVID-19 tests in their reports. That's troubling—to say the least.”
The FDA approved the first COVID-19 antigen test in May. When you see a picture of COVID-19, you typically see a white ball with a bunch of red spikes coming off of it. A molecular (PCR) test looks for the virus’s genetic material in the white ball. An antigen test looks for proteins that make up the red spikes.
“Each category of diagnostic test has its own unique role in the fight against this virus,” according to the FDA. “PCR tests can be incredibly accurate, but running the tests and analyzing the results can take time. One of the main advantages of an antigen test is the speed of the test, which can provide results in minutes.”
While some jumped to the conclusion that the state’s removal was proof the record-high case numbers in the state were inflated and the whole virus scare overblown, it actually suggests the opposite. In a state that’s already struggling to keep up with testing demands and rising hospitalizations, Texas runs the risk of undercounting cases—and undermining trust—by not recognizing positive antigen test results, experts said.
Dr. Sarah Bezek, assistant professor of emergency medicine at Baylor College of Medicine in Houston, said the most transparent way for the state to report cases would be to include data from each type of coronavirus test.
“Just say, ‘These are the number of positives from the PCR tests, these are the number of positives from the antigen tests, and these are the number of positives from the serological studies (antibody tests),” said Bezek, who works on the front line in Houston-area emergency departments. “That would be complete transparency of data.”
A positive antigen test result is reliable, Bezek argued. The two COVID-19 antigen tests the FDA has approved are very specific, meaning they can distinguish between COVID-19 and other respiratory viruses.
The legit knock on antigen tests is that they are less sensitive than PCR tests, which means they return more false negatives. But Bezek pointed out that even PCR tests aren’t 100 percent accurate, further hindering the ability of local health officials to conduct accurate contact tracing.
Depending on how and when a PCR test is administered, it, too, can return a false negative.
“We have patients that are having negative test after negative test,” Bezek said. “There’s certain constellations of symptoms that, after you see enough patients with coronavirus, that when you see somebody who’s testing negative you can say, ‘Well, I’m pretty sure this is coronavirus.’”
Those cases, she added, also aren’t counted by the state.
On Thursday, outrage over the case removals spread like wildfire online. But many of the responses weren’t from people concerned about reduced disease surveillance. Instead, they were conspiracy theorists suggesting the removals somehow indicated the crisis was overblown.
“It’s all been a lie,” one Twitter user said. “They locked us down and destroyed the economy on lies.”
Texas Sen. Ted Cruz got in on the action, sharing a misleading tweet suggesting that those 3,484 people whose cases were removed were never tested and that the San Antonio health department had made a mistake. Cruz added the comment, “Troubling.”
Local health officials were adamant that was not the case.
“Probable cases do not mean ‘maybe’ cases of COVID-19,” said Colleen Bridger, interim director of San Antonio Metro Health. “Antigen tests are FDA-approved, and positive tests are highly accurate.”
Dr. Seema Yasmin, director of the Stanford Health Communication Initiative, said moving to erase the cases like Texas did was the “worst thing” they could’ve done in a climate where mistrust is soaring.
“The last thing you need when you are seeing a surge is for people to suddenly think that the numbers are inaccurate and actually things aren’t as tragic and as at-a-crisis-point as they really are,” Yasmin said.
Texas reported 10,256 new cases on Friday, and hit a new daily record for hospitalizations (10,632) and fatalities (174).
According to the COVID Tracking Project at The Atlantic, many states report confirmed and probable cases in their state totals. The CDC did not respond to a request for comment about whether they planned to issue an updated guidance regarding counting positive antigen tests as confirmed rather than probable.
Because antigen tests return more false negatives than molecular (PCR) tests, there’s a good argument to be made for keeping molecular and antigen testing data separate so researchers can best determine the positivity rates of each (how often tests come back positive).
Joseph Petrosino, chairman of the department of molecular virology and microbiology at Baylor College of Medicine, believes the data sets should be kept separate for the purposes of epidemiological research.
“You want to compare apples to apples,” Petrosino said. “SARS-CoV-2 tracking started with counting positive PCR tests. If additional testing methods of different levels of sensitivity are added, it can hinder tracing efforts and epidemiological studies such as where the virus is spreading the fastest and what individuals are at greater risk.”
Yasmin agrees that, behind the scenes, it makes sense to keep the two data sets separated. But when it comes to informing the public about infections, the total number of positive cases–diagnosed by molecular and antigen tests–should be reported, she argued.
“We make decisions about our lives and whether kids will go back to school or whether we will go to the grocery store often based on community transmission,” Yasmin said. “You need that transparency of data. When it comes to antigen tests, they’re a good indicator of if somebody is a case or not.”