ADVERTISEMENT

Better, Faster Testing Is the Path to an American Comeback

America Needs Rapid Covid Testing to Get Something Like Its Life Back

As undergraduates returned to the University of Arizona for the fall semester, many of the new precautions were hard to miss. Plexiglass dividers were affixed in front of lecterns and between lab benches. Giant tents were set up so students arriving early for a class could wait outdoors, sheltered from the punishing sun. Roving teams of student “health ambassadors” tooled around in golf carts, handing out masks and politely chiding their peers for standing less than 6 feet apart.

But the first thing students had to do was visit one of the university gyms for a rapid SARS-CoV-2 test. The results took an hour. Negative, and you could move into your residence hall right away. Positive, and you were sent to a special isolation dorm, where you spent the next 10 days taking classes online.

Many of these safety measures emerged from discussions headed by Richard Carmona, a former U.S. surgeon general who’s a professor of public health at the university. Carmona’s medical career stretches back to the Vietnam War, where he served as a special forces medic after dropping out of high school. Over the years he’s developed an expertise in emergency preparedness and disaster response. As news of the novel coronavirus began making its way out of China last winter, the university’s president asked Carmona to form a working group and propose ways to keep the institution functioning during a pandemic. “The question was, what were the metrics we’re going to look at and how are we going to be able to make a reasonable determination that this is somewhat safe?” Carmona says. “We didn’t have a playbook for this. Nobody had a playbook.”

Better, Faster Testing Is the Path to an American Comeback

In early spring the team reached out to Ian Pepper, an environmental microbiologist on the faculty who had experience with wastewater testing. “It used to be called sewage surveillance,” says Pepper. “Now the preferred term seems to be wastewater-based epidemiology.” The technique traditionally has been used in population-level studies of illegal drug use or viral infections. Carmona’s team, however, wanted to go upstream from the treatment plant to determine not just if the coronavirus was on campus, but where. The sampling would provide an early warning system—people ill with Covid-19 can start shedding the virus in their stool as many as seven days before showing symptoms. “That gives you seven precious days for intervention,” Pepper says.

Over the summer, Carmona pored over university blueprints with the facilities department, tracing decades-old sewage pipes back to the dorms they drain, while Pepper worked on filtering samples and identifying trace amounts of the virus. By the time students returned in August, the program was in place. Each morning at 8:30—an hour chosen to give residents a chance to visit the bathroom—a faculty engineer working with an undergrad assistant would make the rounds, at each stop prying off a manhole cover and lowering an open Nalgene water bottle at the end of a long aluminum pole into the river of waste below. The pair would place the samples in a cooler and drive them back to the lab, where Pepper’s crew would turn around the results by the afternoon.

On Tuesday, Aug. 25, a week and a half after students started to return, Pepper got his first positive, from the pipe leaving a dorm called Likins Hall. The next day, all 311 Likins residents took Covid tests. Two students, both without symptoms, came back positive and were relocated to the isolation dorm. Thanks to the head start, no other residents would contract the virus.

Since then, the university has stemmed outbreaks in at least eight other dorms using the same method. Pepper is scaling up his lab, training more teams of samplers, and expanding the effort to off-campus apartments and Greek houses. He’s fielding requests from around the world for help setting up similar programs. Carmona, for his part, is careful not to get ahead of himself. “It’s a tool that’s just evolving,” he says, “but we see value.”

Ever since the novel coronavirus began circulating in the U.S., the country’s response has been crippled by a failure to see the spread in anything close to real time. Half a year after the first wave of U.S. lockdowns, with more than 200,000 Americans dead from Covid, we’re still playing catch-up. The Trump administration’s botched rollout of its first tests, and the supply chain shortages that followed, helped the disease spread unchecked. Today, processing bottlenecks still render many test results worthless by the time people get them. This blindness has left public-health officials only the crudest measures of containment, such as broad social distancing mandates and lockdowns. It’s turned the loosening of restrictions on restaurants, sports, offices, gyms, and schools into terrifying leaps of faith. And it has surely killed people.

“The testing we are doing today is mostly just keeping score for the virus,” says William Hanage, an epidemiology professor at Harvard. Even that grim framing gives the U.S. system too much credit. The Centers for Disease Control and Prevention’s sampling of Americans with SARS-CoV-2 antibodies shows that the true number of infections is as many as seven times higher than the official count.

Over time, though, the problem of testing has attracted new focus, new thinking, and new money. Experimental viral screening technologies have taken big steps forward, and researchers have found ways to retool existing procedures. Some of that work isn’t likely to pay off in time to change the course of the pandemic, but some of it already has. And the outbreak’s global scale has spurred epidemiologists and policymakers to seek better answers to fundamental questions about the management of a modern plague: not only how to test but whom to test, and why.

Better, Faster Testing Is the Path to an American Comeback

Those debates are particularly vital now. The number of new U.S. Covid cases, which peaked this summer before dropping significantly, is climbing again—the CDC’s seven-day moving average rose from 34,371 on Sept. 12 to 44,307 on Sept. 26. Colder weather and classroom reopenings threaten an explosion of cases at a time when the public has tired of social distancing’s heavy costs. And we’re still, at the earliest, months away from a working, widely available vaccine.

Until then, testing can help close the gap between normalcy and where we are. If Americans want to safely send kids to school, eat in a cafe, go to a basketball game, or get on a plane, the U.S. needs to test a lot more people a lot faster. Faster, cheaper testing may not flag every new case of Covid, but that shouldn’t mean settling for the current level of blindness, with its torturous drip of preventable deaths. Even world-class testing won’t rid us of the virus, but it can allow us to live our lives in the meanwhile.

Well before this year, the U.S. had ample warning about the health-care system’s vulnerability to a pandemic. An industry long kept as lean as possible lacked the personnel and hospital beds needed to respond to a national crisis, let alone the protective gear and ventilators the coronavirus requires. Testing was the one thing the U.S. had supposedly handled. The CDC’s early warning system for epidemics, based on reports from medical professionals all over the country, was designed to spot new diseases in time to head off bigger problems. And the technology to create tests for new viral diseases is so straightforward that it can generally be done in a few days.

When the time came, though, the apparatus worked against itself. After its Chinese counterparts published SARS-CoV-2’s genetic sequence on Jan. 11, the CDC did indeed create its own test in a matter of days—but because of regulatory requirements and limited access to the relevant viral samples, the agency’s Atlanta headquarters was the only place it could be done. In early February, the CDC produced tests that state labs around the country could use, but they were faulty and returned inconclusive results. Public and private clinical labs rushed to fill the void, only to be stymied by the Food and Drug Administration, which demanded extensive data and a regulatory submission before authorizing alternatives.

By early March, the U.S. was struggling to run 1,000 tests a day, while South Korea, which confirmed its first case the same day the U.S. did, was running 10,000. By the end of that month, lab giants Abbott Labs, Hologic, LabCorp, Quest Diagnostics, and Roche had rolled out U.S. tests of their own. But at $100 to $150 apiece, they were expensive, and labs were short on technicians trained to run them. Long, complex supply chains led to further shortages, from the swabs needed to gather nasopharyngeal samples to the reagents used to process the tests.

The tests in use at the time were molecular diagnostics known as polymerase chain reaction tests. Performed by medical professionals and processed in specialized labs, PCR tests identify segments of a virus’s genetic material in secretions swabbed from the back of a patient’s throat or nose. The process starts with enzymes being added to the sample, which is then repeatedly heated and cooled to create billions of copies of the viral genome. Special “probe” molecules bind to the copied genetic material, releasing fluorescing nanoparticles whose glow signals the virus’s presence.

Better, Faster Testing Is the Path to an American Comeback

Used in everything from food safety to DNA profiling, PCR tests take several hours and are extremely sensitive; if the virus is present in a sample, even in infinitesimal amounts, PCR will find it. To public health officials in the early days of the pandemic, reliability seemed like the most important quality. But the hospitals, doctors’ offices, and pop-up testing facilities where PCR tests are given often don’t have the equipment to process the results on-site—the thermocycler that incubates the samples costs at least $100,000—so the swabs are usually sent out. That’s resulted in massive backlogs and severely limited the tests’ usefulness.

There are alternatives. One, a lateral-flow test, is a disposable strip of cellulose or woven fiber—picture a home pregnancy test. Instead of looking for viral DNA or RNA, it identifies proteins called antigens that protrude from the virus capsule. It exploits the lock-and-key relationship between those molecules and the antibodies the human immune system makes to identify them.

Lateral-flow tests can be done on nasal swabs, but some are in development to work with saliva. The sample is deposited at one end, doused with liquid, and then spreads through the fiber by capillary action, like a spill infusing a sponge. As it advances, the solution flows over antibodies embedded in the fiber with nanoparticles of gold or dye grafted onto them as tags. If the virus is present in the sample, the antibodies latch on and are carried along until the virus-antibody pair encounters a second set of antibodies anchored in a line. Those, too, grab onto the target viruses, activating the tags on the antibodies that hitched a ride earlier. As more virus-rich liquid reaches the line, the tags, like millions of microscopic pixels, form the telltale stripe of a positive result. The whole process can take as little as 15 minutes.

OraSure Technologies Inc. has long sold lateral-flow tests for hepatitis C and HIV. In 2017 it created the only FDA-approved Ebola antigen test. The company is based in Bethlehem, Pa., but sells many of its fast, cheap, over-the-counter diagnostics in sub-Saharan African nations that lack the medical infrastructure for PCR testing. At the beginning of this year, with the U.S. struggling to create reliable tests and its most advanced labs overwhelmed, Chief Executive Officer Stephen Tang saw the parallels.

“We have to repatriate our experience from low- and middle-income countries,” he says. “You have a test that is rapid, that you can take by yourself, that doesn’t require an instrument to read or a medical professional to administer. What we have learned from experience is that people will test themselves more frequently in the privacy of their own homes.” Tang plans to build on the company’s HIV test platform to bring an at-home Covid test to market before the end of the year. While the HIV test costs $40 to $50 in a U.S. pharmacy, the company promises its coronavirus test will be cheaper.

Other strip-test makers have also prioritized the pandemic. E25Bio, a startup born in an MIT lab two years ago, has temporarily shifted its focus from epidemic fever viruses such as dengue and Zika. “With Covid, there have been maybe 50 million infections worldwide,” says CEO Bobby Brooke Herrera. “Dengue causes 400 million every single year.” Like Tang, Herrera has been trying to apply lessons from countries with threadbare public-health systems to those like the U.S. He’s in discussions with U.S. regulators to authorize a coronavirus strip test made to be processed in a lab. Once it reaches the market, he says, a home version will be next.

Established testing companies are already bringing lateral-flow tests to market. On Aug. 26, Abbott Laboratories announced that the FDA had awarded it an emergency use authorization for BinaxNOW, a 15-minute disposable antigen test priced at $5. A so-called point-of-care test, it must be given by a health professional. But since it requires no lab equipment, an unlimited number can be conducted and processed at any given time. Abbott is already producing 50 million of them a month.

A test’s accuracy is a matter of sensitivity (how often it correctly spots a disease) and specificity (how often it correctly rules the disease out). A test with low sensitivity will miss lots of cases; a test with low specificity will mistakenly tell people they have a disease when they don’t. Lateral-flow tests are neither as sensitive nor as specific as PCR tests. They’re more likely to miss an infection during the incubation period, as the invader has just started replicating, or weeks in, when the body’s immune response has managed to eliminate many of the virus particles.

Lateral-flow tests’ lower sensitivity has hampered efforts to bring them to market. There are concerns that tests done at home, where people swab their own noses instead of having a professional do it (or at least supervise), could further lessen the sensitivity. Herrera says he doubts E25Bio’s at-home test will be approved without a change in expectations from the FDA, which requires accuracy levels difficult for him to achieve outside a lab. (OraSure says it’s confident it can clear the bar.)

Some companies are trying to address this shortcoming technologically. Mammoth Biosciences Inc., working with a microbiologist named Charles Chiu at the University of California at San Francisco, is using the DNA-splicing technology Crispr to improve the accuracy of strip tests. Crispr’s potential for precise gene editing rests on its use of sequences of “guide” RNA that zero in on target genes. Chiu and Mammoth are turning that search-engine property to the task of finding the telltale genome of SARS-CoV-2.

Better, Faster Testing Is the Path to an American Comeback

The effort is taking two parallel paths. One that the FDA cleared this summer runs on the same equipment as PCR tests, but it copies the viral genetic material without the time-consuming heating cycles. The other, in development with GlaxoSmithKline Plc, manages to replicate this approach, with the attendant sensitivity, in a self-contained and disposable kit not unlike a lateral-flow test. “You’re getting the same accuracy as in the lab, but it’s much easier to use,” says CEO Trevor Martin.

Another school of thought, though, holds that worries about the sensitivity of strip tests miss the point. “The way we’ve been approaching testing so far in this epidemic is to try to shove a public-health need through a diagnostic pathway,” says Michael Mina, the Harvard epidemiologist who is the leading proponent of this idea. Mina argues that we should essentially test everyone all the time, everywhere. It’s a form of what public-health experts call surveillance—rather than diagnostic—testing. In this model, speed and ubiquity are all that matter.

Mina likes to talk about what would happen if we had $1 self-administered saliva tests. Restaurants and bars would have cups of strips on the maître d’ stand by the toothpicks and the mints. Theaters would have them at the door. Air travelers would be tested at their departure gate before boarding. (The International Air Transport Association recently announced its determination to put a system like this in place.) Coronavirus testing would become a habit—if not like brushing our teeth, then at least like flossing. These tests would be much more reliable than temperature checks, which can miss carriers with mild or nonexistent symptoms.

At a large enough scale and a high enough frequency, the thinking goes, testing is all you need to do. Tests at the entrance to offices and schools would visibly ensure the safety of those spaces, encouraging people to return to them. People who test positive could immediately isolate themselves, breaking the chain of transmission. In this world, a strip test’s relative insensitivity would be less of an issue—positives could be double-checked to make sure they’re correct, and false negatives would be caught a day or two later by the next test, still early in the disease course. Mina goes so far as to characterize lower sensitivity as an asset, because a PCR test taken in the late stages of Covid-19 will still show positive even if the patient is well past the contagious stage and no longer needs to be isolated. In that sense, he argues, strip tests can better gauge an individual’s risks to the broader population.

“I don’t want to see these kinds of tests being promoted as passports to party or to go out to dinner—that misses the big picture,” the epidemiologist says. “These aren’t passports. These are actually the intervention, like a vaccine, to suppress population spread.” According to his modeling, testing just half the population every three days would bring Covid under control within weeks.

To make this a reality, the FDA would need to relax its testing standards down from PCR-level sensitivity, and someone would need to manufacture and pay for the untold billions of tests. Mina believes it should be Congress. “This is a national emergency,” he says, and a $50 billion or $100 billion price tag to stop the virus is easily worth it given that the U.S. has already spent $3 trillion in Covid-related stimulus money and its economy has lost trillions more.

At least one of the two men running for president has unequivocally committed to more testing. The first point in Joe Biden’s seven-point Covid plan is to “ensure all Americans have access to regular, reliable, and free testing.” The plan mentions at-home and instant tests and proposes a testing board modeled on Franklin D. Roosevelt’s War Production Board, which oversaw the manufacture of weapons and supplies that helped win World War II.

President Trump has publicly called for less testing, and his CDC recently said asymptomatic people who’ve had contact with a sick person don’t need to be tested. (Most public-health experts strongly disagree, and the agency has since reversed that guidance.) On the other hand, the day after Abbott unveiled its lateral-flow kits, the U.S. Department of Health and Human Services announced a deal to buy and deploy 150 million of them. And the agency is already distributing a different type of antigen test to nursing homes across the country. Of course, all of this is a fraction of what Mina’s plan would require.

Among his fellow epidemiologists and public-health experts, Mina’s plan does have skeptics. Without reimbursement, the price of even very cheap tests might be a challenge for some families and organizations. (New York City’s school system has more than 1 million students.) The production of billions of new tests could create supply chain problems of its own, and the millions of false positives and negatives that would be produced at that scale could further undermine Americans’ shaky confidence in their public-health system, a dangerous prospect in a country where large numbers already refuse to wear masks. There’s also no mechanism to compel people who test positive at home to stay there, especially if they’re living paycheck to paycheck without paid sick leave.

It’s our failure to act on the test results we have that has been the problem, these skeptics argue. As Hanage, Mina’s colleague at Harvard, points out, South Korea succeeded in corralling the virus by relentlessly tracking down those put at risk by each discovered case, not by reinventing testing. “All of these extraordinarily smart people are coming up with extraordinarily smart things,” Hanage says, “and we haven’t done the basic things right.”

Whether or not Mina can sell the idea of regular tests for all, the importance of faster, more widespread surveillance testing has become clear. Such a model could use strip tests, if they can be scaled up, or it could rely on methods such as pooling—combining multiple samples to increase the throughput of labs. Hanage is particularly excited about the potential of wastewater.

Ashish Jha, a physician and the dean of the Brown University School of Public Health, is among the scholars who’ve tried to calculate how many tests the country actually needs. Early in the pandemic, his team came up with a range of 500,000 to 600,000 per day. Those estimates, though, assumed that positive results would be followed by South Korea-style contact tracing. So far, such programs have largely failed in the U.S., and Jha is reassessing. “I have increasingly come to believe that a large army of contact tracers may not work in our country,” he says. “And that we’ll need large-scale surveillance testing as a strategy for building a high degree of safety into a lot of our activities.”

Jha’s new approach focuses less on how many tests the nation needs than on where they’re needed. Nursing homes and hospitals are places where people should be tested all the time, he argues, with meatpacking plants, prisons, and schools close behind. “You could make the case for any place where you can get large numbers of people gathering indoors for extended periods,” he says. “My argument has been that we have to make decisions. If we don’t have tens of millions of tests a day—and we don’t—who do we want to leave out?”

The American health-care system, with its balkanized tangle of private and public providers, consumers, and often-reluctant reimbursers, is structurally resistant to nationwide interventions. “With a few exceptions, we don’t have a national or even a statewide public-health infrastructure that scales to administering many, many procedures directly to patients,” says Katherine Hempstead, a health-care policy adviser at the Robert Wood Johnson Foundation.

It’s possible that an imagined Pandemic Testing Board on a war footing could overcome some of those limitations. If not, and at least until then, it’s likely that the U.S. will see the uneven growth of the patchwork testing quilt we already have. Individual companies, schools, even states could put together effective testing programs while other efforts fail. In August, a group of governors announced a joint effort with the Rockefeller Foundation to purchase millions of antigen tests.

The patchwork model means gaps, blind spots, and inequity. But it will nonetheless embed the idea of disease surveillance in our lives long after this pandemic is over. One of the problems that’s bedeviled the U.S. response at every level is the nation’s short attention span. Public and private funding goes to swine flu or Zika or Ebola, then evaporates once the immediate danger has passed. “It’s shortsighted, the fact that we only deal with the outbreak of the day,” says UCSF’s Chiu. “Had we had ongoing funding to do diagnostics for Zika, maybe the tests that we’re now struggling to develop in real time would have already been developed.”

Even the terrible costs of Covid probably won’t curb these human tendencies, but more durable changes might help. In Arizona, Carmona has been meeting with architects trying to figure out how to design buildings with wastewater epidemiology in mind. One of their goals is to enable future sewage samplers to trace pathogens not just to a building but to individual floors, or even rooms. “It may be what we’re going to need in the future,” Carmona says. “We probably will have other biologic hazards—maybe not a pandemic, but it could be an epidemic, or just an outbreak—where you’d like to know where the sick people are.”
 
Read next: Germany Has Its Own Dr. Fauci—and Actually Follows His Advice

©2020 Bloomberg L.P.