In a pregnancy test, a woman pees on a cartridge that contains a strip of absorbent paper. That paper is coated with tiny Y-shaped antibodies that recognize and clamp onto a hormone that women make when they are pregnant. In a life-altering bit of chemical drama, if the hormone is detected, the strip changes color.
A new batch of COVID-19 tests in development makes use of essentially the same technology. If they are deployed widely and frequently, some experts say, they could change the course of the pandemic, even before there’s a vaccine.
“We have a workable solution today,” says Michael Mina, MD, PhD, an expert in diagnostic tests at the Center for Communicable Disease Dynamics at Harvard University.
Mina says that with enough government funding -- meaning billions of dollars to be on the same scale as the investment in vaccines or the economic stimulus -- millions of these tests could be printed and distributed to every American at the cost of about $1 per test. He thinks that if everybody got a box of 50 tests and took one every day, “just like they put in their contact lenses,” the U.S. could see its outbreak disappear. Mina says he is helping some of the companies that are developing the rapid tests, but he doesn’t have a financial interest in any of them.
“If everyone is doing it, or even a just a majority of people are doing it, then it will stop the vast majority of transmission and it will cause these outbreaks to disappear in a matter of weeks. We don’t have to wait for a vaccine,” he said on a recent call with reporters.
How Would it Work?
A person spits on a test strip, or mixes their saliva with a bit of saline solution and applies it to a paper strip. As the solution is sucked through the absorbent paper, it flows past antibodies that recognize and hook only to a specific piece of the virus that causes COVID-19. If the virus is present in high amounts -- which is also when a person is thought to be most infectious -- the antibodies bind to it and cause an indicator in the test to change color.
No painful swab up the nose. No trip to a doctor’s office or testing site. And crucially, no long wait for results.
When a person waits days for COVID-19 results, the delay renders a test useless for controlling the pandemic.
Here’s why speed counts: When a virus invades the body, it hijacks our cells to turn them into miniature virus factories. Very quickly, it scales up production of itself this way, so that within days, the body is teeming with copies of the COVID-19 virus.
After studying the dynamics of COVID-19 disease transmission over the last several months, researchers now believe that a person’s viral load -- that is the number of copies of the virus in their bodies -- is probably highest early in the infection, starting a day or 2 before they get symptoms, if they get any symptoms at all. This is when they’re most capable of spreading the disease.
In order to control the spread of COVID-19, infected people have to be found and isolated right as their viral loads are at or approaching this peak, and ideally, before they have the chance to infect anyone else.
If a person has to wait 4 or 5 days or even a week to get their test results back, they don’t get results fast enough to act on them. They don’t know they need to stay home and limit their contact with others. Public health workers aren’t alerted in time to contact and warn the people they might have infected.
Infected kids go on to infect their classmates and teachers. Workers bring the virus into their offices. The spread of the virus continues.
A recent study in The Lancet Public Health journal found that once test results are delayed more than 3 days, even a perfect contact tracing system can’t stop the growth of the outbreak. The same study found that with a more modest 2-day delay in test results, you need swift and thorough contact tracing to put a meaningful dent in virus transmission. That’s something most of the U.S. doesn’t have right now.
That’s the reason Microsoft founder Bill Gates recently said on CNN that the American testing system was delivering “the most worthless results of any country in the world.”
The Promise and Pitfalls of Daily Testing
Several companies have these kinds of rapid tests in development. Mina says he has some of their products sitting on his desk.
They include big companies like 3M, which announced in a July 14 news release that it was collaborating with scientists at the Massachusetts Institute of Technology to develop a rapid test. They also got $500,000 in funding from a technology accelerator program at the National Institutes of Health to help them speed their test to market.
There are also smaller companies in the mix: E25Bio and Sherlock Biosciences, which are based in Cambridge, MA.
Sherlock’s test, called INSPECTR, was developed at Harvard’s Wyss Institute. The company says they hope to have it on the market by the first half of 2021. It works in a slightly different way. Instead of using an antibody to look for the virus, it uses a fragment of a synthetic gene. This bioengineered genetic fragment attaches to a snippet of genetic material from the virus. When the probe combines with the virus, it turns on the gene--almost like completing an electrical circuit. The gene then starts making copies of a protein which can be detected on the paper test. strip
Many of these companies all have a version of these tests ready. The problem, says Mina, is that they aren’t yet accurate enough to get the go-ahead from the FDA.
Sherlock says its test is highly accurate and should meet FDA requirements, but they haven't been able to complete clinical studies to take to the agency.
FDA requires over-the-counter diagnostic tests to be sensitive enough to accurately catch nine out of every 10 people infected with the virus that causes COVID-19. If a doctor has prescribed the test, it has to be sensitive enough to accurately report positive results for eight out of 10 people who are infected. The FDA’s requirements for doctor-prescribed tests are less strict than they are for over-the-counter options. The agency thinks that since doctors will be interpreting the test results, they’ll be more likely to know when they need to double-check the results against the gold-standard test that’s sent off to a lab for analysis.
In order to keep the price of the quick, at-home tests low, they have to use basic, easily available materials. Since they use lower-grade materials, the tests aren’t as good at catching people with COVID-19 when the level of virus in their bodies is low. When virus levels in the body are higher, the tests become more accurate, Mina says.
That may not be the kind of test a doctor might want to use during flu season to try to figure out what’s causing a person’s fever and body aches, he says, but these tests could work for pandemic control because they find some people right at the point they’re most likely to pass the infection to someone else.
Mina says the U.S. needs a new regulatory structure that would allow cheaper, faster tests, to be used as public health tools, rather than diagnostics.
“I think we’ve kind of lost our minds in terms of this real focus on analytical sensitivity,” Mina said on a recent episode of the This Week in Virology podcast.
“We’re so focused on high-end, expensive tests that we’re not testing anyone because we’re trying to focus on clinical-grade tests.”
Mina estimates that because COVID-19 tests in the U.S. have to be accessed through a doctor’s office or sent off to a lab for results, we’re probably only catching about 5% of people when they’re actively spreading the infection. He thinks if people could take a test every day or every other day, at home, we might be able to catch as many of 85% of people before they spread their illness.
“Maybe we only need a really crappy test if it’s cheap enough to use it very frequently,” he says.
Other experts agree with him.
“I think it’s a very, very good idea,” says Arthur Caplan, PhD, director of the Division of Medical Ethics at New York University’s Langone Medical Center. “We are sinking under the weight of approved testing. We don’t have enough tests. They’re too slow. Chemicals are running out. Return times are way, way too long, and it’s not working.”
Caplan says he was recently consulted by a school system in Seattle that wants to try rapid testing like this with its students.
“It's time to go to a system that while less accurate and more prone to errors, still could give us some information that would be useful, particularly useful and affordable, I should also add, particularly if you do it, you know, over long periods of time,” he says.
But the FDA is not convinced.
“In evaluating a test’s known and potential benefits and risks, FDA considers both the likelihood of a false result and the consequence of a false result. For example, a false result in a home setting for a non-prescription test is not mitigated by consultation with a healthcare professional who considers other signs and symptoms and can advise the individual based on the healthcare professional’s medical judgement. As such, for example, an individual with a false negative result from an OTC diagnostic test may be less likely to quarantine despite symptoms, putting others in the community at risk,” the agency says in an emailed statement.
They aren’t the only ones on the fence about this approach.
Peter Lurie, MD, is the president of the Center for Science in the Public Interest. Before taking that post, he was an associate commissioner for public health strategy at the FDA.
He thinks it could be disastrous for the FDA to lower its standards for testing accuracy.
Lurie points out that even a very accurate test can deliver a high percentage of incorrect results. That’s because the performance of the test depends not only on the characteristics of the test, but also on how widespread a disease is within a population.
Let’s say you have a test with a 95% sensitivity and 95% specificity. In medical terms, the test sensitivity is the ability of a test to correctly detect a person who truly has the disease. The test specificity is the ability of the test to rule out people who aren’t infected.
That means that 95% of the time, the test correctly catches people who have COVID-19, and it correctly identifies people who aren’t infected 95% of the time. That sounds pretty trustworthy, but in a population where only 5% of the people have the infection, that can be a problem.
Here’s why: Let’s say you have a population of 100 people. Five of them actually have COVID-19. If you give a test with a 95% sensitivity to all 100 people, it will probably catch all 5 of the true positives, which is great. But with a specificity of 95%, the test is also going to make some errors.
Out of 100 people who take it, 5%, or five people, will get a false positive result. A false positive is when the test says they have an infection, but they really don’t. In this example, half the positive results -- five out of 10 total positive results -- returned by the test would be wrong. A test like that could generate 50,000 false positives a day for every million people who take it. There would be an equally large chunk of people who get falsely negative results.
And most at-home tests currently can’t meet that mark of 95% sensitivity and specificity.
“Performance of most at-home tests have been well below the 80% threshold,” Adm. Brett Giroir, MD, assistant secretary for health at the U.S. Department of Health and Human Services, says in a statement to WebMD. Giroir is also the COVID-19 testing czar.
Lurie points out that all those positives would need to go on to get an additional test -- the slower kind that is sent off to a lab -- to confirm their results.
“It just seems like it creates a whole bunch of additional work,” he says. “You’re going to spend a lot of time chasing your tail.”
Lurie says he’s also concerned about the anxiety those false positives would create for people before they could be straightened out. And there’s the danger of people acting on wrong information: Going to work when they thought they were negative, for example.
“A wrong test is not something with a trivial outcome,” he says.
And then there’s the precedent it would set. Lurie thinks that if the FDA relaxes its standards to accommodate one kind of test, the market could be flooded with low-quality tests.
“If you allow tests on the market with poor performance characteristics, it takes away the incentive for people to be good, because the incentive is for manufacturers is to be just good enough to get to market,” he says.
Lurie points out that already happened with antibody tests. The FDA lowered its standards to allow manufacturers to get their tests to market more quickly, and many of them didn’t work. The agency has now pulled more than 100 of these tests off the market.
“It’s going to be the Wild West out there of people with very poor tests. That’s what happens,” he says.
'Test, Test, Test'
In a May 16 news briefing, the director-general of the World Health Organization warned that without an effective vaccine, there’s only one way to stop the spread of COVID-19.
“You cannot fight a fire blindfolded. And we cannot stop this pandemic if we don’t know who is infected,” said Tedros Adhanom Ghebreyesus. “We have a simple message for all countries: test, test, test.”
From the beginning, the United States has struggled run an effective testing program. Mina thinks rapid tests could give people in the U.S. a way to get back to more normal lives.
He acknowledges that these cheap, rapid tests could be wrong. A lot. But the tests would also catch far more people, and catch them while they are contagious. And he says there are ways to make this kind of testing better.
For one thing, people should be advised to treat the results as preliminary, “the same way that we treat HIV tests. We don’t actually go and tell all of those people that they’re positive” after they take a rapid HIV test. “We run a confirmatory test first.”
Mina says that’s a good reason to allow several different kinds of rapid tests onto the market. Since they all work in a slightly different way, they could backstop each other. “Maybe there’s some protein that their body happens to make that’s sticking to the test in a way that makes it look positive,” Mina says. “So what somebody could do is they would open their package of 30 tests that they got at the drugstore, and along with those tests there would be five confirmatory tests,” he says.
Other experts think it could work.
“These tests just have to be good enough to detect people with a lot of virus and really identify individuals who are at greatest risk of transmitting the infection, so they don’t have to be perfect. They just have to be good enough,” says Isaac Bogoch, MD, who specializes in infectious diseases at the Toronto General Hospital Research Institute and the University of Toronto.
And that means some of the tests will be wrong. People will be missed, and people will get positive results when they aren’t infected. Bogoch says this kind of testing will require us to think a little harder about the results and what to do with them.
“Like anything else, if these types of tests are to be integrated into real-world settings, we have to think about smart systems and smart protocols on how to do that appropriately,” he says.