Advertisement

SKIP ADVERTISEMENT

‘It’s Kitchen Sink Time’: Fast, Less-Accurate Coronavirus Tests May Be Good Enough

Experts are revising their views on the best methods to detect infections, setting aside long-held standards so that the spread of the virus can be more quickly tracked and contained.

Health workers with a rapid antigen test in Kolkata, India. These quicker tests are less precise than other types in use, but are much faster at producing results that help spot community spread.Credit...Avishek Das/SOPA Images/ZUMA Wire, via Alamy

For months, the call for coronavirus testing has been led by one resounding refrain: To keep outbreaks under control, doctors and researchers need to deploy the most accurate tests available — ones reliable enough to root out as many infections as possible, even in the absence of symptoms.

That’s long been the dogma of infectious disease diagnostics, experts say, since it helps ensure that cases won’t be missed. During this pandemic, that has meant relying heavily on PCR testing, an extremely accurate but time- and labor-intensive method that requires samples to be processed at laboratories.

But as the virus continues its rampage across the country and tests remain in short supply in many regions, researchers and public health experts have grown increasingly vocal about revising this long-held credo. The best chance to rein in the sprawling outbreaks in the United States now, experts say, requires widespread adoption of less accurate tests, as long as they’re administered quickly and often enough.

“Even if you miss somebody on Day 1,” said Omai Garner, director of clinical microbiology in the U.C.L.A. Health System. “If you test them repeatedly, the argument is, you’ll catch them the next time around.”

This quantity-over-quality strategy has its downsides, and is contingent on an enormous supply of testing kits. But many experts believe more rapid, frequent testing would identify those who need immediate medical care — and perhaps even pinpoint those at greatest risk of spreading the disease.

Such a considerable shift would likely be a welcome change for a country where the status quo of testing was just described as “unacceptable, period” by Dr. Anthony Fauci, the director of the National Institute of Allergy and Infectious Diseases, in an interview Wednesday on CNN.

“If you had asked me this a couple months ago, I would have said we just need to be doing the PCR tests,” said Susan Butler-Wu, a clinical microbiologist at the University of Southern California. “But we are so far gone in this country. It is a catastrophe. It’s kitchen sink time, even if the tests are imperfect.”

Of the dozens of coronavirus tests that have been granted emergency use authorization by the Food and Drug Administration, most rely on complex laboratory procedures, like PCR, to detect the coronavirus’s genetic material.

Only a handful are quick and simple enough to be run in what is called a point-of-care setting, like a doctor’s office or urgent care clinic, without the need for lab equipment. And these tests are still relatively scarce nationwide. Government officials have pledged to astronomically scale up the number of point-of-care tests by fall, increasing by millions the weekly tally of tests conducted.

But many of the rapid tests at the center of this buzz still rely on specialized machines that are neither cheap nor easy to produce in bulk. And while some rapid tests, like the Abbott ID Now, were quickly adopted into the White House, most are unlikely to get much play in the larger community.

“We can’t have these in every household,” said Michael Mina, an epidemiologist at Harvard University’s School of Public Health and a vocal proponent of speedy testing.

Image
An Abbott Laboratories ID Now rapid testing machine.Credit...Carlos Osorio/Associated Press

A better option, Dr. Mina said, might be antigen testing, which identifies pieces of protein. Two such tests, made by BD and Quidel, have received emergency authorization from the F.D.A. Both require instruments to run, but even simpler versions of this technology could provide a fast answer in the same way as pregnancy tests, allowing users to swab their mouths or noses or spit into a tube, then read the results as colored bars on a strip of paper within minutes.

These tests could be done at a doctor’s office, or even at home — no fancy machines or specially trained personnel required — and cost just a few dollars a pop, perhaps even less. And there would be no delays of a week or longer. Companies like the Massachusetts-based E25Bio are currently cooking up tests that might fit this need.

Such convenient setups could resolve some of the supply shortages that have plagued testing laboratories across the nation for months and caused a national embarrassment over inadequate, inaccessible testing. In many cities, people are still experiencing turnaround times of over a week, sometimes three or more — as people did at the beginning of the U.S. epidemics — for results from PCR-based tests, effectively rendering them useless for themselves and the people around them.

Hamstrung by a long delay, even an accurate test can morph into a useless one.

Natalie Magnus, who got tested in Winnebago County, Ill., on July 14, still didn’t have results 22 days later. Her brother and sister-in-law, who were each tested twice at separate facilities in Colorado on July 7 and July 8, have so far received only one set of results after a 17-day wait. One of them was positive for the coronavirus.

Ms. Magnus, who has already completed a two-week quarantine at home, no longer cares if she gets her results. “By now, that doesn’t tell me anything,” she said.

Antigen tests, on the other hand, can be low-tech and easy to manufacture en masse. Distributed weekly or even daily, they could painlessly screen people headed back into offices, schools and universities, delivering peace of mind to parents, teachers and employers. Everyone — not just those feeling ill — would have an easier way to assess their health status on a regular basis.

“The goal here is to detect as many infections as possible,” said Daniel Larremore, an applied mathematician who models infectious diseases at the University of Colorado, Boulder. “That means taking as many shots on goal as we can.”

Broad distribution of antigen tests might also ease the demand for PCR, which still has an important role in hospitals and vulnerable communities like nursing homes. As things stand, many doctors still can’t get their patients’ results within 24 hours.

In those settings, accuracy is crucial, said Melissa Miller, director of the Clinical Molecular Microbiology Laboratory for U.N.C. Hospitals. “You don’t want to miss that diagnosis.”

There are drawbacks. Antigen tests will miss some people who would test positive by PCR, which amplifies coronavirus RNA and picks up even tiny amounts of it. With antigen tests, proteins can’t be copied in the same way, making it easier to mistake some people who are carrying minute levels of the coronavirus in their bodies as pathogen-free. Some antigen tests used in the past miss up to half the infections they look for.

Early on, many experts balked because of these shortcomings. “As laboratorians, we wanted the most sensitive test, the most specific test, the most accurate test,” Dr. Miller said. “End of story.”

But Dr. Mina argues that false negatives might not be as bad as they seem.

Virus levels vary from person to person, and can wax and wane in an individual over the course of an infection, typically peaking around the time symptoms first appear. People carrying — and thus probably shedding — gobs of germs will most likely turn up positive using every test on the market, Dr. Mina said. Those harboring less virus in their bodies might get more mixed results. Many of these individuals, however, probably aren’t the cases of most concern.

It’s another way to look at testing accuracy, Dr. Mina said: “Detecting the most infectious people.”

Researchers don’t yet know how much virus a person needs to carry in their body to actually transmit it. But the range in which the accuracy of antigen tests starts to drop off is probably far below that level, Dr. Mina said.

Testing frequency can also be a formidable foe to disease transmission. In a recent paper that has yet to be published in a peer-reviewed scientific journal, he and Dr. Larremore showed through mathematical models that a relatively insensitive test, rolled out twice a week, vastly outperformed a more accurate test, administered once every two weeks, in curbing the spread of disease. Other studies pitting speed against sensitivity have come to similar conclusions.

The upshot here is a practical one, Dr. Garner said. “You’re not trying to find every single person who has the virus,” he said. “You’re trying to mitigate outbreaks.”

That approach is a substantial departure from the way that many lab researchers have traditionally tackled infectious disease testing.

“We’re in sort of a new era of testing,” said Esther Babady, a clinical microbiologist at Memorial Sloan Kettering Cancer Center. “Usually we use tests to diagnose disease states in patients, not look for disease states. Now, with Covid, we are starting to look for this virus everywhere.”

Image
A coronavirus drive-through test site in Miami-Dade County last month.Credit...Joe Raedle/Getty Images

A testing rethink this substantial will inevitably come with snags. The success of the speedy testing strategy hinges heavily on availability — something the United States has utterly failed at since the virus first made landfall within its borders. Ramping up antigen testing may only add strain to an already sputtering supply chain, potentially hampering plans for widespread use.

“If you test people all the time, you can account for lack of sensitivity,” Dr. Butler-Wu said. “But are you really going to test people all the time? If the answer is no, then that lack of sensitivity becomes more of a big deal.”

And many experts are still hesitant to trust negative antigen results, which may need to be backed up with a more sensitive test. That might not be very burdensome in the midst of an outbreak, when positivity rates are likely to be high, Dr. Babady said. In spots where infection rates are especially low, however, “that is not going to be the best approach to testing,” she said.

Cheap tests also aren’t the same thing as free tests. Even $1 tests can start to rack up quite a bill, especially for large families hoping to do daily checkups or nursing home aides required to get tested repeatedly. Left unregulated, the testing market could end up exacerbating the socioeconomic split that’s already disproportionately burdened some sectors of the population with the worst effects of Covid-19.

Concerns over accuracy bogged down the approval process for simple, speedy tests. F.D.A. guidelines stipulate that any new coronavirus test vying for emergency clearance from the agency must perform nearly as well as the gold standard of PCR.

The F.D.A.’s rules frustrate Dr. Mina, who noted that several companies on the verge of debuting antigen tests have found the regulatory hurdles daunting. “Many of them are not even bothering,” he said. “‘Our product might not be good enough. So what’s the point?’”

That’s left the onus on the few companies who already have the F.D.A.’s green light. In hopes of encouraging a speedier, ramped up production, the governors of seven states announced this week a joint bid to purchase a total of 3.5 million antigen tests from BD and Quidel.

There probably isn’t one way to grapple with all these issues — and certainly not an obvious one, Dr. Butler-Wu said. What’s clear to her and others, though, is that the current situation is untenable.

“Our backs are against the wall, and it’s Hail Mary time,” Dr. Butler-Wu said. “We have to try something different.”

Katherine J. Wu is a reporter covering science and health. She holds a Ph.D. in microbiology and immunobiology from Harvard University. More about Katherine J. Wu

A version of this article appears in print on  , Section A, Page 5 of the New York edition with the headline: ‘It’s Kitchen Sink Time’: Fast, Less-Accurate Tests May Be Good Enough. Order Reprints | Today’s Paper | Subscribe

Advertisement

SKIP ADVERTISEMENT