In 24 hours, the testing landscape of the United States has transformed.
Yesterday morning, all of the tests for COVID-19—traditional or rapid—that had received emergency authorization from the Food and Drug Administration required an expensive machine and cost around $40 or more. In the afternoon, the health-care company Abbott announced that it had received FDA authorization to distribute a new type of test. This test requires only a coated-paper card and a small swab, and the scale of its production is stunning: Abbott says it will begin manufacturing 50 million of these tests a month in October. The tests will cost just $5 apiece.
That wasn’t the only news. This morning, the Trump administration announced that it would be purchasing 150 million of these tests from the company. For comparison, states have reported fewer than 75 million tests total over the past six months, according to the COVID Tracking Project at The Atlantic.
Many experts have hailed Abbott’s new test as a huge milestone, and a rapid acceleration toward a plan that could give Americans back some sense of normalcy. Deployed widely and often enough, tests like Abbott’s might allow for kids to return to school, office workers to head back to cubicles, and essential employees who have been working throughout the pandemic to have a greater degree of safety. At a time when so many people are desperate to escape the pandemic, this kind of testing holds the tantalizing promise of a pre-vaccine way out of the quagmire.
Abbott’s test itself is not quite the fulfillment of that grand vision, though—at least not yet. Michael Mina, a Harvard epidemiologist who has pushed for the deployment of this kind of rapid test, was not ready to declare that the new era had begun. “This is the type of test that we’ve been waiting for, but may not be the test,” Mina told me.
The fine details of the FDA’s emergency use authorization, the regulatory sign-off that allows a test to be sold, are significant. Antigen tests such as this one detect viral protein and are less sensitive than the gold-standard PCR tests, which detect viral RNA. So, to gain acceptance, Abbott showed data confirming that its test was able to classify a pool of samples in nearly the same way as PCR tests of the same samples. Although its performance was impressive, the number of samples was far smaller than would be tested under normal circumstances—just 102.
Also, the FDA’s authorization allows Abbott’s test to be used only on people with symptoms of COVID-19 within seven days of the onset of their symptoms. It cannot be used to test people at any time, an approach that is often called screening. Mina said that limits the test’s potential, because it can’t be used to detect people who carry the virus but don’t show symptoms, and asymptomatic and presymptomatic people are thought to transmit a substantial percentage of infections.
Kristian Andersen, an infectious-disease researcher at Scripps Research, noted that Abbott developed an app to go along with the test that would provide proof that a person had recently tested negative. To Andersen, this is clear evidence that “Abott didn’t just have symptomatic diagnostics in mind here, but also screening of asymptomatic people.” Symptomatic people are much more likely to have the virus, decreasing the chance of a false positive. For asymptomatic people, even in an area with lots of infections, Andersen worried that Abbott’s test isn’t accurate enough, and would generate a substantial number of false positives. (In May, Abbott was embroiled in a controversy over the sensitivity of another one of its rapid tests, Abbott ID Now.)
Andersen and Mina both offered up prospective solutions to increase the test’s specificity. For example, a future version of the test could look for two viral targets. Or two tests could be packaged together, one of which would be a slightly more expensive test that you’d take if you got a positive on the first. That is to say, many questions remain about the test, about Abbott’s strategy, about the Trump administration’s distribution of the tests, and more.
To process the potentially big news, I reached out to Abbott’s lead scientist on the project, John Hackett, to address some of the questions about how the new test might be deployed, improved, and criticized. I’ve condensed and lightly edited our conversation for clarity. Abbott’s FDA authorization might be limited right now, but the company’s vision is clearly bigger than its current regulatory status.
Alexis C. Madrigal: How do you see these tests being used?
John Hackett: This project was built around the concept: What do we need to start to move our society back to normal? How do we get people back to work? How do we ensure these workforces will be as safe as possible? How do we get people back to school? That was the goal.
So we really did want to have tests that could be as broadly distributed as possible.
Madrigal: Fifty million tests a month is a huge number. That’s more than twice the number of tests the U.S. completes in a month. How did you ramp up production so massively?
Hackett: This was the challenge of this program. We needed some sort of reliable testing that could be affordable and that doesn’t require instrumentation. You need scale. The more frequently you could test people, frankly, even tests with lower sensitivity would be very effective at identifying people quickly and slowing the spread. As we were developing the test, there were people working in parallel looking at supply chain and logistics. Abbott took a lot of risk—hundreds of millions of dollars were spent building two new manufacturing facilities focused solely on those tests. We hoped we could come to a solution that would be where we needed it from an overall accuracy perspective, but if you weren’t building capability simultaneously, there was no way it could be the answer.
Madrigal: With the current sensitivity of the test, there are some questions about it as a screening test. The [emergency use authorization] doesn’t allow that yet. What would you have to do to roll this out as an FDA-authorized screening test?
Hackett: The FDA is starting to create templates for over-the-counter-type testing. We’re looking at that. And trying to explore the challenges. When you think about going to home testing, there are a lot of factors involved. This test is very easy to use; there’s no question. But you’d have to worry about how people store them. Are they sitting in their hot car? This test will be a major step, but we do have to continue to explore others.
Madrigal: So right now this test still needs to be done in some kind of a facility. But when I look at the number of tests you’re able to produce, it makes me think that at-home testing is where you’re ultimately driving, even if you aren’t allowed to do screening in general yet. Is that true?
Hackett: This is really a critical first step. The first primary thought was being able to do tests where people gather: a workplace is a good example, or a school. There, you can have a health-care provider—it doesn’t have to be a doctor, could be a school nurse, an occupational health specialist—who would use [the test] with very limited training. Any facility that has a CLIA certificate [in this case, a kind of waiver] for testing in a new-patient setting would be capable of using this.
We recognize there could be value in home testing, and we’ll continue to look at this. There are concerns. When you go to home testing, how effective is the sampling a person does? That’s the importance of why we chose a nasal swab [which just goes in the nostril] instead of nasopharyngeal [which goes way up in the nose]. It’s night and day in terms of pain. And you do need a trained health expert to do nasopharyngeal. We’ll explore other pathways to getting more testing to the people who need it.
Madrigal: One complication with widespread antigen testing is that it’s blurring the national picture of the pandemic. These tests aren’t really being reported, so even though this type of test is valuable, they’re currently creating a hole in the data. Have you given some thought to the data-reporting side of this?
Hackett: People testing are supposed to report positives. For example, if you were an employer using a test like this, that is the responsibility of the health group. We have created the Navica app, a complementary mobile app that goes with the test. We weren’t using it for that purpose, but it is one that allows the person to get a digital health pass.
We were thinking about: How can you lock in testing a test result and be able to verify that you were, in fact, tested? If you were an employer, you could know the employees have done the testing at whatever interval you wanted to do the testing. The default setting for that is seven days, but that could be modified.
Madrigal: Take my kids’ school as an example. It wants to do rapid testing. How would the school go about getting your tests and doing this testing? How would it actually work?
Hackett: We at Abbott do testing of critical staff. This is the sort of thing that can be set up. You need some health-care providers and to figure out what is the interval of testing that would achieve what your goal is overall or what you think is appropriate. Then it is really just organization. Again, I think that’s where the Navica app would be a real plus. If you went into a school building, they could just check; it would be like a boarding pass. And if, unfortunately, you were positive, you’d get a message to quarantine and see your doctor.
Madrigal: But that sounds like broad screening to me, not diagnostic testing of someone with symptoms.
Hackett: Well, that comes back to the health-care provider. For this test, you still need, in essence, a prescription, saying that these individuals are suspected or they have symptoms. This test, the indication is for seven days or less after the onset of symptoms.
Madrigal: How do you improve a test like this? How do you take the basic components and say “This is how we’re going to get to this sensitivity and this specificity”?
Hackett: Well, this is a situation where, obviously in this outbreak, there has been very little time to react. We haven’t had the normal time frames that we would to develop tests like these. There was a tremendous amount of work that went into the process, every level of this.
What protein of the virus should we target? Then, looking at the reality in this format, during the extraction process off of the swab, you need to be able to separate the proteins that are part of that virus in order to be able to detect them. So would you choose spike or nuclear protein? What type of swab? Abbott’s got a long history of this in the infectious-disease area. You can go back to 1985; we had the first FDA-approved test for HIV. This is part of our history.