(Updated: 1:45 p.m., 5/22/20: Hours after this story was published on Friday, May 22, Health Department spokesman Nate Wardle told the Capital-Star that the statements he made about serology testing on Thursday were inaccurate. “The negative results being reported do not include negative serology tests,” Wardle said in an email. He added: “I was under the impression that negative results included all negative results. That is not correct. Negative results are only negative [diagnostic] results. I apologize for that mistake.” Read more here.
Pennsylvania finally appeared to break a lengthy slump in its COVID-19 testing efforts this week, when it reported more new test results per day than at any other point in the pandemic.
But since state officials changed the way they’re reporting testing data, an unknown number of Pennsylvania’s 300,000 negative results come from a testing method that experts say is unreliable.
A Department of Health spokesman confirmed to the Capital-Star Thursday that the agency has been lumping data from two different types of COVID-19 tests into its daily public reports: diagnostic tests, which determine whether or not a patient is infected with COVID-19; and antibody tests, which can assess, with limited accuracy, whether or not someone may have been infected in the past.
The practice was first reported by The Atlantic magazine on Thursday morning. It defies the advice of public health experts, who say that antibody tests – also known as serological tests – are still too unreliable to inform the state’s public health response to COVID-19.
It also complicates efforts to measure Pennsylvania’s testing capacity, which experts say must expand as counties reopen.
“Serological testing could be useful in the future, but right now I don’t see its value” Dr. Ping Du, an epidemiologist at Penn State Hershey Medical Center, told the Capital-Star. “It’s really not very helpful for disease control.”
Du said diagnostic tests remain the “gold standard” to “understand the magnitude of a [COVID-19] outbreak within a community.” The tests allow public health officials to track and contain outbreaks in real time by identifying and isolating infected patients.
In many states, a positive diagnostic test is the only way a patient will be counted as a confirmed COVID-19 case, according to The Atlantic.
Since Pennsylvania reported its first two COVID-19 cases in March, less than 1 percent of the state’s 65,392 confirmed COVID-19 patients have received presumptive diagnoses based on serological testing. And that was only after they also showed symptoms of the disease or had a high-risk exposure to another patient, health department spokesman Nate Wardle told the Capital-Star in an email.
The state counts those patients as “probable” cases, Wardle said.
Antibody tests account for an unknown number of the 303,514 negative test results that Pennsylvania has collected to date, Wardle said. He said it’s impossible to isolate diagnostic tests from the total figure, thanks to the way that labs and health systems report negative test results to the state.
Health secretary Dr. Rachel Levine said Thursday that the state only considers confirmed and probable COVID-19 cases when deciding whether or not a county can lift stay-at-home orders. And since the state can distinguish between serological tests and diagnostic tests that are positive, officials believe “there are no issues with how we are reporting this data,” Wardle said.
But the the influx of serological testing data makes it hard to judge whether the state is meeting the goal it set in its statewide reopening plan, which is to administer 8,500 COVID-19 tests per day.
Pennsylvania appeared to hit that goal three times in the last week. But experts like Du were under the impression that the benchmark was based on diagnostic tests.
Evaluating the state’s testing strategy is much more difficult when its negative test results – which account for the vast majority of the data Pennsylvania reports each day – include an unknown number of results from unreliable, non-diagnostic tests.
The reporting practice also appears to contradict the state’s official testing plan, which as of Thursday says serological testing is not yet widespread in Pennsylvania since officials are still reviewing available models.
Though there are more than a dozen COVID-19 antibody tests on the market, Du said their reliability is “still under discussion.”
Antibody tests detect the blood proteins that patients produce while fighting a COVID-19 infection, which can indicate whether or not a patient may have carried the disease in the past.
But early research suggests that COVID-19 patients don’t all produce antibodies at the same rate. Since some tests can only detect antibodies in a limited window of time, Du said, a patient who had COVID-19 can still return a negative serological test.
The U.S. Food and Drug Administration (FDA), which regulates pharmaceutical products and vaccines, says some patients may need to be tested twice to get reliable results from a serological test.
Wardle told the Capital-Star that Pennsylvania health officials began to combine diagnostic test results with antibody tests once the state started getting serological testing data from labs. That was “sometime after” April 13, he said, when the federal Centers for Disease Control and Prevention issued its first formal guidance on the use of antibody tests to classify probable COVID-19 cases.
That means the change may have taken effect even as state officials were still expressing doubt about the reliability of serological testing.
Gov. Tom Wolf told reporters on an April 23 conference call that there were “problems” with the speed and reliability of available antibody tests.
On April 21, Wolf said that officials at a state-run lab in Exton, Pa. were still evaluating serological tests “to make sure that they actually do what they are intended to do. But so far, the jury’s out.”
According to Wardle, that evaluation is still ongoing.
He said Thursday that Pennsylvania does not yet administer antibody tests at its state-run lab, since officials are still determining which test to use based on the accuracy of different models, their equipment requirements, and their authorization from the FDA.