What does ‘negative’ on a coronavirus test really mean?

Katherine J. Wu, The New York Times

Posted at Oct 17 2020 12:42 AM

A healthcare worker performs a rapid Covid-19 antigenic test at a coronavirus disease (COVID-19) testing centre in Nice, France, October 15, 2020. Eric Gaillard, Reuters

In a flurry of memos released this week and last, the White House physician, Dr. Sean Conley, stated that President Donald Trump no longer posed a transmission risk to others — an assessment, he noted, that was largely precipitated by the results of a bevy of “advanced diagnostics.” The declarations have helped clear Trump to return to the campaign trail, including a town-hall-style event hosted by NBC News on Thursday evening.

Outside experts have also said that Trump, who reportedly began feeling sick about two weeks ago, is probably no longer infectious. But most have based such assessments on the trajectory of the president’s symptoms — not the results of his tests.

There exists no test that can definitively determine whether someone who caught the coronavirus is still contagious. “We do not have a test for cure, and we do not have a test for infectiousness,” said Omai Garner, a clinical microbiologist at the University of California, Los Angeles.

Experts have criticized the administration’s overreliance on tests to keep the coronavirus out of Trump’s inner circle. Now, they said, the White House appears to be leaning too heavily on tests to break the president out of isolation.

Not all coronavirus tests are designed to detect the same parts of the virus. And a negative on one test does not necessarily guarantee a negative on another.

“We don’t just look at these tests in the context of ‘Coronavirus, yes or no,’” said Karissa Culbreath, a clinical microbiologist at TriCore Reference Laboratories in New Mexico. “Each test looks for a different aspect of the virus.”

As recently as Tuesday, for instance, a rapid test called the BinaxNOW was unable to detect the coronavirus in Trump. But results yielded from a laboratory test, which used a slower but more accurate technique called polymerase chain reaction, or PCR, showed he still carried genetic material from the virus at low levels in his body.

The White House also reported using two experimental approaches on Trump: a viral culture, in which researchers in a laboratory tried to grow the coronavirus from a sample taken from the president, and a test for subgenomic RNA, a proxy for the presence of actively replicating virus.

Many types of PCR and antigen tests for the coronavirus have received a green light from the Food and Drug Administration. But viral culture and probes for subgenomic RNA are not currently part of the standard testing tool kit, Culbreath said.

PCR-based tests are some of many among a suite of so-called molecular tests, which hunt for genetic material exclusive to the coronavirus. These tests include an amplification step, wherein the genetic material is copied over and over until it reaches detectable levels, revealing even very small amounts of the virus.

Some PCR-based tests can also give an indication of how much virus is lingering in the body — a number called the cycle threshold, or CT, which increases as the virus becomes more scarce.

A recent PCR test taken by Trump yielded a CT of 34.3, after steadily increasing for several days. Researchers have had trouble growing the virus out of samples taken from people whose PCR tests crest above the low 30s. But exceptions to this trend do exist, and cycle-threshold readings are often inconsistent among different types of PCR-based tests, and even among laboratories using the same test.

“We cannot say, ‘A CT value of 34 does not make someone infectious’ across the board,” Culbreath said. “It’s not a universal answer.”

Tests like the BinaxNOW, on the other hand, are antigen tests, which search for bits of proteins, or antigens, made by the coronavirus. Unlike PCR-based tests, antigen tests do not include a step in which they amplify their targets, making them faster and more convenient but less reliable at finding the virus when it is in low abundance. Some PCR-based tests are thought to be many thousands times more sensitive than antigen tests.

It’s thus very possible to be antigen negative but PCR positive, while still harboring the virus in the body, Culbreath said.

The BinaxNOW has emergency clearance from the FDA, but only when administered within seven days after the onset of symptoms. Researchers do not yet know how the test will perform in people outside this window. Trump received several of these tests during his second week of illness, all of which returned negative results.

“The tests have to be used when they’re supposed to be used,” said Andrea Prinzi, a clinical microbiologist at the University of Colorado, Anschutz Medical Campus. “That’s when they’re going to help you the most.”

Both PCR and antigen tests have another limitation: They look for hunks of the virus’ anatomy — debris from the pathogen — and cannot determine whether the virus is still active. Some people who have been infected by the coronavirus may register as PCR positive for weeks, even months, after they are no longer contagious or ill, simply because tests are picking up on harmless souvenirs of an infection long past.

“Neither is a measure of actual, live virus,” Garner said, of PCR and antigen tests.

That’s where viral culture comes in. Scientists can take a sample from a person’s airway and then try to coax the coronavirus into infecting cells in a lab. But these procedures are not widely available to the public, because they have to be performed in a high-containment facility by people specifically trained to work with deadly pathogens.

These restrictions have led several scientists to experiment with PCR-based tests that search specifically for subgenomic RNA as a possible proxy for active virus. Subgenomic RNA is a type of genetic material that is produced only after the coronavirus has infected a human cell. The compound can thus act as a sort of molecular beacon that can alert researchers to a virus that has started to reproduce itself, without the need to grow the dangerous pathogen.

Conley’s memos about Trump appear to indicate that the president is no longer producing samples with detectable subgenomic RNA. Researchers have also tried to culture the coronavirus outside of his body, although few details on this have been shared. Virus that can be cultured is not necessarily transmittable, or vice versa.

Neither viral culture nor subgenomic RNA tests are approved for widespread use, Garner noted. Researchers doing these studies might not perform their experiments in the same way, making any results difficult to interpret without more information how, and by whom, they were obtained.

Guidelines published by the Centers for Disease Control and Prevention stipulate that symptoms — not test results — should be the primary motivator for ending a person’s isolation. People with mild or moderate COVID-19 should isolate for least 10 days after their symptoms start. That timeline could extend up to 20 days if their symptoms are severe.

Before advising someone to end their isolation for COVID-19, “I would want to know about their symptoms,” said Dr. Krutika Kuppalli, an infectious disease physician in South Carolina. While information on tests wouldn’t hurt, that data should be considered secondary to how the patient is feeling. Otherwise, the tests might be not just unhelpful, but also a “waste of resources,” Kuppalli said.

“If the president meets CDC guidelines, he can come out of isolation,” said Dr. Alexander McAdam, director of the infectious diseases diagnostic laboratory at Boston Children’s Hospital. “But I don’t see testing as a get-out-of-jail-free card.”