Some of the nation's leading public health experts say that the most commonly used type of coronavirus test in America could be generating positive results for large numbers of people who are carrying low levels of the virus and therefore may not be contagious, Apoorva Mandavilli reports for the New York Times.

The 3 biggest questions about Covid-19 testing, answered

However, according to Mandavilli, those experts also say that doesn't necessarily mean the country should be testing less.

Why coronavirus tests may be too sensitive

The most commonly used coronavirus tests in America are polymerase chain reaction tests, or PCR tests. According to Mandavilli, PCR tests provide a yes or no answer to whether a patient is infected with the novel coronavirus.

"We've been using one type of data for everything, and that is just plus or minus—that's all," Michael Mina, an epidemiologist at the Harvard T.H. Chan School of Public Health, explained. "We're using that for clinical diagnostics, for public health, for policy decision-making."

But other types of tests can also provide information on how much of the virus a person is carrying, Mandavilli reports. According to Mandavilli, PCR tests look for genetic matter from the new coronavirus using amplification cycles. The fewer cycles needed to detect genetic matter from the virus, the higher a patient's viral load and the more likely that person is contagious, Mandavilli reports.

However, the number of amplification cycles that was needed to detect genetic matter from the virus, which is referred to as the cycle threshold, typically isn't included in test results sent to doctors and patients, Mandavilli reports.

Mina said that information is more telling, as it indicates a patient's viral load, and patients carrying minuscule amounts of the virus may not be contagious. Therefore, information on a patient's viral load should be used to determine the patient's next steps, Mina said.

"It's really irresponsible, I think, to forgo the recognition that this is a quantitative issue," he added.

But further complicating the matter, according to some experts, is that many commonly used tests for the novel coronavirus are too sensitive, so they generate positive results when patients are carrying low loads of the virus, Mandavilli reports.

According to Mandavilli, many coronavirus tests have fairly high cycle thresholds, with most set at 40 and some set at 37. That means a number of patients who aren't carrying much of the new coronavirus are still testing positive, even though they may not be contagious, Mandavilli reports.

Mina explained that tests with high cycle thresholds could be detecting genetic fragments of the virus, or pieces of the virus that are leftover from a previous infection that don't pose any current transmission risk.

Juliet Morrison, a virologist at the University of California-Riverside, said she believes any test with a cycle threshold over 35 is too sensitive. "I'm shocked that people would think that 40 could represent a positive," she said.

And according to Mandavilli, a review conducted by the New York Times of three sets of coronavirus testing data from Massachusetts, Nevada, and New York found that up to 90% of patients in those data sets who tested positive for the coronavirus had very low viral loads.

Ashish Jha, director of the Harvard Global Health Institute, said of the Times review's results, "I'm really shocked that it could be that high—the proportion of people with high [cycle threshold] value results." He added, "Boy, does it really change the way we need to be thinking about testing."

How should testing change, according to the experts?

Morrison said a better threshold would be 30 to 35, while Mina said he'd recommend setting the threshold to 30 or less.

According to Mandavilli, CDC data shows that it's very difficult to detect significant loads of live coronavirus within a patient sample that tests positive at a cycle threshold above 33, meaning patients who test positive at a cycle threshold of 33 or higher likely are not carrying enough of a viral load to transmit the virus to others. Mandavilli reports that CDC has said it's looking into using cycle threshold measures "for policy decisions," and it will work with FDA and device manufacturers to make sure the measures "can be used properly and with assurance that we know what they mean."

Mina said there's a simple solution for weeding out which patients are and aren't contagious: "Test them again, six hours later or 15 hours later or whatever."

Mina explained that using a rapid coronavirus test, which can be less accurate than PCR tests but provide quicker results, could be useful for this type of re-testing. Those tests may be less sensitive, but they'd catch patients whose viral loads are rising quickly, indicating they've recently been infected with the novel coronavirus and are not just carrying leftover fragments of the pathogen, he said.

Sensitive PCR tests made sense at the start of America's coronavirus epidemic, but with outbreaks persisting throughout the United States, tests that are faster, lower cost, and more abundant are what's now needed to get the epidemic under control, Mina said.

"It might not catch every last one of the transmitting people, but it sure will catch the most transmissible people, including the superspreaders," he said. "That alone would drive epidemics practically to zero" (Mandavilli, New York Times, 8/29).