'Radical' COVID Testing Plan Could Offer A 'Return To Normal,' Journalist Says

TERRY GROSS, HOST:

This is FRESH AIR. I'm Terry Gross.

There's a plan that might be able to give us our lives back even before a COVID-19 vaccine. That plan is the subject of the latest article by my guest, Alexis Madrigal, in The Atlantic magazine, where he's a staff writer. The plan involves massive numbers of daily tests with a simple saliva test using a paper strip similar to a home pregnancy test - no waiting on long lines, no swabs up your nose, no relying on reagents and swabs, which have often been in short supply, and no long waits for results. His article describes how these tests work, their advantages and disadvantages and what it would take to make available tens of millions of these tests a day. He says implementation would require changing how we think about tests and what we expect of them.

Madrigal is the co-founder of The COVID Tracking Project, along with three other people, including Robinson Meyer, who co-wrote the new article. The project is doing what Madrigal says the federal government hasn't effectively done - collecting data from each U.S. state and territory, compiling the latest numbers on tests, cases, hospitalizations and patient outcomes. The project has also partnered with the Boston University Center for Antiracist Research to track racial and ethnic data from every state that reports it and pushing states that don't to start. That project analyzes the data to uncover the impact of the virus on vulnerable communities.

There's one more credit I should mention. I am proud to say Alexis used to be FRESH AIR's tech commentator.

Alexis Madrigal, welcome to FRESH AIR. It is so good to have you back on our show.

ALEXIS MADRIGAL: Thank you for having me.

GROSS: So I'd like you to explain why massive numbers of quick and easy tests might help open up America again.

MADRIGAL: Because the key problem in the pandemic is we don't know who's contagious. And unfortunately, with SARS-CoV-2, the virus that causes COVID-19, many people are asymptomatic but spreading the disease. And because of that, it is very hard to really grind transmission down to nothing. And so what you need, really, is some tool in the toolkit that lets you say, OK, this person's contagious, and this person's not. That person needs to stay home, and this person can go about their life.

We have a lot of tests in the United States now, but they're pretty expensive. And they actually take a while, generally speaking, to come back. You know, we're talking something on the order of days, not on the order of minutes. And what we really need is something that's on the order of minutes so that you know before you leave your house, am I contagious?

GROSS: So before we get to how these tests would work and how they compare to the tests we are using now, describe a typical scenario. Like, take - the kids are going back to school now and to college. How could they be tested in a way that we might know who is infected and who needs to stay home?

MADRIGAL: Right. So it's really about the testing system. You know, our tests - what we've really optimized for so far has really been about, you know, how incredibly sensitive and accurate we can make these tests. But you're only going to be tested every once in a while.

So let's take a scenario of a kid going back to school or going back to college, let's say. You know, they get tested, say, a week before they go back to school. Maybe they get their results back, you know, from two to seven days from then. By the time they actually get to school, they've had, likely, other exposures, possibly from flying on a plane, possibly from other students. And so even though the test itself can say very precisely this person did or did not have the virus a week ago, it's kind of hard to say how that should play into the public health scenario of that student showing up on campus and going into a dorm.

And some of that just has to do with the features of the virus - how long people are contagious for, when they show symptoms, when they don't. And some of that has to do with the testing system that we have, which just takes a while to return results. And so in many cases, by the time you find out, you're already not contagious (laughter) or you've already spread it to the people you would've spread it to. And it's that kind of mismatch of the kind of tests we have and the kind of testing system that we're using and what we're trying to use it for, which is this sort of public health screening.

GROSS: So compare that to if we had massive tests, where you can test yourself at home and find out in 15 minutes if you have the virus.

MADRIGAL: Yeah. So if we had tests, hypothetically hundreds of millions of these tests so that they were a dollar or two and you - it was just a regular part of life to test yourself, you could know with a pretty high degree of certainty - perhaps not 100%, but let's say you could know 80% - (laughter) I do - I am not contagious, or I am contagious.

And that's - wouldn't catch everyone. I mean, to be clear, the kind of test that we're talking about wouldn't have the accuracy of the tests that have largely been employed. But it would catch a bunch of people who are asymptomatic. It would catch a bunch of people who are presymptomatic. And it would stop a lot of transmission chains that otherwise would have gotten going. And that's really the promise of this kind of testing regime. You test a lot. And actually, testing frequently helps make up for some of the lower sensitivity of the test.

And many other - many - because many people would be testing themselves, you'd gain the advantage of widespread testing covering large chunks of the population rather than fairly narrow testing capturing, you know, just a tiny slice of the possible number of infected people each day.

GROSS: So let's get back to the school model. You explained why the tests we have now aren't that effective in screening out people who are carrying the virus. Say we had these rapid home-based tests. How would that work in the model of returning to school or to college?

MADRIGAL: Sure. I mean, you would just take a test each day along (laughter) that progression. So let's say, you know, you'd take a test a couple days or a day before you went to school. You'd fly to the college. You'd take a test before you interacted with other people. In the, you know, most intense versions of this plan, you might take a test every day before you left the dorm. And that would allow you to catch all kinds of contagious people. I mean, this really - that it requires volume. We've got to, like, sort of out-volume the virus. That's kind of the idea behind the plan.

And then basically, you would go about your daily life. You'd go, you'd do all the stuff that you normally do. Perhaps you'd still wear a mask. Perhaps there'd still be some social distancing measures in place, but they would be nothing like what we're encountering now. I mean, maybe you'd even take it before you went to a party (laughter), you know? I mean, there's - there would be ubiquitous testing, basically.

GROSS: And there's even a model where the tests would be time coded so you could show it to the person who'd be asking at your school or a college. You could say, here's the results. Here's the time code. I took the test this morning.

MADRIGAL: Sure, yeah. I mean, there's all kinds of systems that you could build around it once you had this type of test available in large numbers.

GROSS: Let me reintroduce you here. If you're just joining us, my guest is Alexis Madrigal, a staff writer for The Atlantic magazine and co-founder of the COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on tests, cases, hospitalizations and patient outcomes. We'll be right back. This is FRESH AIR.

(SOUNDBITE OF TODD SICKAFOOSE'S "TINY RESISTORS")

GROSS: This is FRESH AIR. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine. He's been reporting on the spread of the coronavirus and efforts to contain the spread. He's also co-founder of The COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on test cases, hospitalizations and patient outcomes. His latest article in The Atlantic is about how new rabbit saliva-based tests could help us reopen the country.

So the test we're using now - at least, the test most people are are using now - is very accurate. But it's expensive. It's time-consuming. Tell us what it's called and a little bit of how it works.

MADRIGAL: Sure. These are PCR tests. They are one of the wonders of science, really, of the last, like, 30 years. It allows you to detect even very small amounts of genetic materials, kind of like a big amplifier for genetic material in a particular sample. And so what these tests are doing is actually looking for little snippets of the actual SARS-CoV-2 viral genome. And they are very sensitive. And they're very specific. So these are these kind of two markers of, you know, you're very unlikely to get false positives, and you're very unlikely to get false negatives.

And these tests are expensive. Let's call them, say, hundred or $150 for a test. Most of these tests have been done in big, commercial laboratories. So you kind of have to send them out to the lab. And the turnaround time for a lot of these tests is, say, between one day and, you know, seven days or, in a lot of bad circumstances during the pandemic, even up to two weeks. So these tests are this really specific tool, you know? They are extremely accurate along the dimensions that we noted. They can pick up tiny amounts of viral genetic material.

But they also come with these downsides. They take a while to return results. They're dependent on these supply chains for chemicals called reagents, as well as swabs, as well as particular kinds of test tubes. And they tend to run on these things that are called platforms, which are produced by big pharmaceutical companies. Those machines themselves are pretty expensive. And so you have the cost problems. And you have the turnaround time problems. So they're just absolutely amazing tests. But they have these downsides. They have a very specific shape, I think, is what I'm trying to say.

GROSS: And one of the problems of the test taking so long to come back is that you're most contagious before you even have the symptoms.

MADRIGAL: Oftentimes, yeah, you're contagious before you have symptoms or right when you're getting them. And so let's say you get a little bit sick or you detect some symptom. You wait a few days before you get a test - or you have to wait, because oftentimes that was - that's been happening. By the time you even get the test, you might already be out of the most contagious phase of the disease. So that's really why this niche mismatch of timelines is, really, a big problem.

GROSS: So compare that test, the PCR test, to the new tests that are being developed and are being used in limited quantities right now. And these are tests that work through the saliva, as opposed to a swab up your nose. Tell us how they work and how they compare in what they're testing for to the tests that are widely used now.

MADRIGAL: Sure. So the main way that people can imagine doing a lot of rapid testing is through antigen testing. So that basically means it's looking for a protein that the virus produces and not, like, the actual genetic material. So it's kind of a different target. It's less sensitive. It means that - that means it's more likely to tell you you don't have the virus when, in fact, you do. And these tests are simpler. That's probably the best way of putting it. And there's two different kinds.

One is sort of the Nespresso model. There's, like, a little machine. And you do the test. And the machine sort of pops back the result. And those are normally done, you know, say, at a doctor's office. Or a nursing home is a big location for this kind of testing. It's called the point-of-care testing. And that's kind of what we have now. There's a couple of those tests approved by the FDA for use. And the federal government has bought a lot for nursing homes.

The other kind of test is - would be a simple strip test. So the thing that most people are probably most familiar with would be a pregnancy test. Or there's also tests like this for dengue fever. So about 100 million of those are done each year. And these kinds of tests are just extremely simple, relatively speaking. And you'd be, really, doing them at home. They wouldn't be in a doctor's office. They wouldn't be in any kind of institutional setting. You'd either do them at school or you'd do them in the workplace. You'd do them at home.

And they are just quite different (laughter). There's, really, sort of almost no relationship between how sort of complex and expensive and centralized the system is for doing PCR testing and this kind of paper-strip test. It's just, you know, kind of worlds apart in the sort of testing that it is. And there's been a lot of people who've called for large-scale screening. You know, Paul Romer has been talking about screening, doing lots of tests, rapid tests, for a long time. More recently, Michael Mina, an epidemiologist at Harvard, has become a big proponent of this kind of plan and has convinced a lot of people that we need to go in this direction.

And it's just - it's really about finding the right tool for handling the precise public health crisis that we have now. And that might be these sort of rougher, less sensitive tests but that can be produced in vast quantities and distributed for a lot less money.

GROSS: Where are we now in terms of developing these rapid tests?

MADRIGAL: There's not really a technical issue as much. These tests do exist. There's companies like E25Bio, which is a Cambridge startup that has been working on these kinds of tests. But they haven't really been produced at scale, and they haven't been approved by the FDA. And one reason is they're being compared to sort of the gold standard PCR tests, the gold standard look-for-the-viral-genetic-material test.

And there's a couple problems with that. One, you'd be using them in quite a different way. And, two, one of the issues with PCR testing is that it actually picks up a lot of people who are no longer sick anymore. It's almost, like, so sensitive that it picks up particularly on the backside of the infection. So, you know, you get sick, and then you get better, and then you're no longer contagious, nor are you sick. And yet a PCR test will keep giving you positive test results.

GROSS: So you think that the downside of the rapid test is far outweighed by the speed and by the number of people who can be frequently tested.

MADRIGAL: That's right. I do. And there are clearly - there are downsides. You know, there are going to be situations in which someone takes one of these rapid tests, gets a negative, goes and sees someone they love, and it turns out they were contagious, you know? That is going to happen. The thing is, that already happens now. And we're not catching a huge percentage of the infections.

I mean, this is really kind of a crucial thing to understand. We don't even really know how to estimate the ratio between sort of cases we can confirm with a PCR test and the number of people who've actually been infected. You know, maybe we're catching 5% of the cases. Maybe we're catching, you know, 10%. Maybe we're catching 20%. Maybe it's 50% in some places and 2% in other places. What we do know is there's a lot, a lot more infections than there are confirmed cases. We just - that is absolutely clear at this point based on the antibody testing that's been done in hard-hit places in New York and in other areas.

GROSS: So, you know, one more question on the rapid tests that are in the works. Not all epidemiologists support the emphasis on these saliva tests. Why not? Like, what are some of the problems they see, problems that you take seriously?

MADRIGAL: Yeah. I mean, you know, the truth is the tests are not as sensitive as PCR tests, as the gold standard tests. There's also a kind of behavioral theory embedded in the, you know, widespread saliva test plan, which is just that people would, in fact, you know, take the test every day and self-isolate if they got a positive. And that may not be true. People may not do that. There may be manufacturing problems.

And I think the other big question is, how many people would have to get a negative and then go infect somebody and how many of those stories would have to come out before people decided that this was a bad plan, even if the population-level public health arguments are very strong? And that's really the issue, you know? It's - there - it's really asking people to think about the population level and not just about their individual test result, but the system as a whole.

GROSS: And, also, if you want to track the results, people have to report it. If they're doing a test at home, what would the reporting system be like?

MADRIGAL: Exactly. I think we would lose a lot of data if this happened because, exactly, people themselves would be responsible for reporting it, and not a lab and not a public health department, and that would be tough. It's basically only worth it if, in the execution, it was able to stop the pandemic, which is the promise. Otherwise, the real downsides of data loss and the lower sensitivity of the tests and all these things are quite real, and we obviously have to take that seriously.

GROSS: Well, let's take a short break here, and then we'll talk about how soon these tests might be available and what it would take to get them mass produced on a huge scale and distributed around the country. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine and co-founder of The COVID Tracking Project. We'll be back after a short break. I'm Terry Gross, and this is FRESH AIR.

(SOUNDBITE OF MUSIC)

GROSS: This is FRESH AIR. I'm Terry Gross. Let's get back to my interview with Alexis Madrigal, a staff writer for The Atlantic magazine who's been reporting on the spread of coronavirus and the efforts to contain it. His latest article is about rapid saliva tests being developed that could possibly help contain the virus and allow us to open up the country again even before a vaccine is available. He's also co-founder of The COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on tests, cases, hospitalizations and patient outcomes.

So, you know, we've been talking about these rapid saliva tests that are being developed, kind of similar to a pregnancy test where there's a strip. I guess, you spit on it, and then it quickly tells you, maybe in 15 minutes, whether you have the virus or not. But it's only, like - what? - 70 or 80% accurate. So there's an issue with that. How close are we to actually having these tests developed and tested in a way that they're usable?

MADRIGAL: Well, on a technical level, some of these tests already exist. Basically, a half-dozen companies are working on something like this. The big problem is in order to scale up like a lab technology to something that would be distributed, you know, to hundreds of millions of people all the time, you kind of need the government's muscle. And, you know, Michael Mina - who I've mentioned a few times - Harvard epidemiologist, you know, his version of this plan is essentially - and the Rockefeller Foundation, which has also been heavily involved in sort of testing strategies. Their version of this plan, too, is basically to have the government come in and kind of go on a World War II-style production blitz.

So, you know, I'm - here in Oakland, there were a bunch of ships that were produced here in the Bay Area, you know. And it was - the government just moved money and capital and worked with companies to build tons of ships, you know, Liberty ships that helped, you know, win World War II. And they're basically saying we need a similar kind of effort - the government to come in, basically nationalize the companies that are working on these technologies and scale up their manufacturing, you know, take over factories and just go. And, you know, it's a huge - it would be huge effort, huge effort.

But this a national emergency. And if you compare it to things like the Fed did, say, in 2008, that's actually not as extreme as what we did to sort of get our banking system alone, you know, up and running versus, you know, this would be a technology that might get our entire country back up and running. And so the argument by proponents of this plan is, you know, we're spending, you know, trillions of dollars to keep the country running forward when let's - why wouldn't we put 10 or 20 or $30 billion towards the creation of this other kind of system - or even a billion dollars? (Laughter) You know, I mean, we haven't really put more than a few hundred million dollars into testing technologies in the United States. And yet, they could be a key part of getting us out of this quagmire.

GROSS: It sounds like things are getting a little bit better in terms of government help with funding for these new rapid tests. The Centers for Medicare and Medicaid have started to ensure that demand will exist for an experimental test by promising to buy tests from two of the companies for every nursing home in the country. Can you describe what they're doing?

MADRIGAL: Sure. So, you know, they're - a couple of companies are producing these point-of-care antigen-rapid tests. And they basically said, what you make, we'll buy (laughter). And this is basically what the government did in vaccine development as well. They said, you produce it, we'll buy it. And they - it's a smart strategy. Basically, it takes the risk out of the investment that you're asking these companies to make. And it's something that seems to work quite well. The real limit is really just, like, the production side of this, you know. How many of these machines? How many of these tests can you produce?

When it comes to these, you know, rapid tests that are done in a little, you know, point-of-care machine, the estimates that have come out is maybe there'd be 4 million tests a week available by the end of October, which would be quite substantial. That would be a large percentage of the tests that are done in the United States right now on a weekly basis. PCR test - that is the sort of traditional tests that we've been using. And so it's pretty promising, I think. That said, again, that doesn't get us to everyone taking a test all the time and a return to normal life. That may provide substantial protection for vulnerable communities and nursing homes and other long-term care facilities. But it's not you wake up in the morning and take this test.

GROSS: Meanwhile, this week, the CDC modified its guidelines for who should get tested. And it now excludes people who don't have symptoms even if they've been exposed to the virus. Isn't one of the major reasons for testing to prevent people who don't know they're infected from spreading it to others?

MADRIGAL: Yeah. And there's a lot of questions in the public health community right now about how this is going to be interpreted at the state level, where a lot of public health really happens. And there's a lot of questions about why exactly to do this. To me, it doesn't really make sense not to test people who are asymptomatic, particularly those who've had close contact with people with the disease. One thing that has happened, though, repeatedly in the pandemic is that once a transmission really gets going in a particular community, it completely overwhelms the testing system no matter kind of what the scale of it has been.

And so, perhaps, that's one reason. And the other thing is that the tests that we're doing are in fact extremely, extremely sensitive, the type of tests that have done - been done to date, you know, some so-called PCR tests, which are, you know, detecting the genetic material of the virus. And so they pick up a lot of people who are probably no longer contagious but still have some virus in their system.

And so what that has tended to do - I mean, it's people who, say, got sick 10 days ago are no longer or, you know, 12 days ago, they're no longer infectious. But now those people need to quarantine at home for two weeks. And now we're going to send contact-tracing resources to talk to the people that they've been talking to for the past couple days when they probably actually weren't contagious during that time.

So there's a lot of waste that's kind of baked into the system, and that waste actually creates problems because it overwhelms the system with testing demand. And so that may be what the CDC is thinking here. But I would say that a lot of the people that I talked to around testing and my reporting partner, Rob Meyer, talked to around testing, screening - that is to say, testing people who are asymptomatic - has got to be one of the strategies that we're employing. And so these new guidelines are quite confusing.

GROSS: If you want to be really skeptical, you can wonder, is the CDC changing its guidelines and suggesting limiting testing because President Trump knows that the more you test, the more positive results you'll get, the numbers will go up. He does not want to see the numbers go up, especially before the election. Are you hearing that from people?

MADRIGAL: I think there's a lot of concern about that, about juking the stats in various ways. I think the problem is, if you're not testing asymptomatic people who can spread the virus, who may be spreading the virus, you'll just end up getting more positive cases of symptomatic people. I don't actually think it's possible to hide a really serious outbreak in the United States. And you might be able to shave off the top of the case numbers, but eventually, people are going to show up in the hospital. Eventually, people are going to start dying.

And we've just seen that so many times in so many different places here and abroad. That may be something that's in play. I have no knowledge of that. But it also just doesn't make sense because, eventually, people really do get sick, and it shows up in other kinds of data aside from just the number of cases.

GROSS: Let's take a short break here, and then we'll talk some more. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine and co-founder of The COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on test cases, hospitalizations and patient outcomes. We'll be right back. This is FRESH AIR.

(SOUNDBITE OF MUSIC)

GROSS: This is FRESH AIR. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine who's been reporting on the spread of the coronavirus and efforts to contain the spread. He's also co-founder of The COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on test cases, hospitalizations and patient outcomes.

It sounds from your article like there have been pharmaceutical companies, tech companies, economists, philanthropists kind of bonding together to do what they think needs to be done to develop some kind of solutions for the pandemic and to fill in the gaps from the government, things the government isn't doing.

MADRIGAL: That's right. I mean, the federal response has been extremely spotty. So, you know, it gave credit on Operation War Speed and vaccine development, but a lot of the parts of the response have been completely uncoordinated and bounced onto the states. You know, the federal government multiple times has kind of tried to wipe their hands of an actual, real national response and has just said, hey, it's the responsibility of the states to do X and Y.

And that has led to, you know, massive waste and chaos at different times and for different types of pandemic responses, whether it was the personal protective equipment, the PPE, whether it was, you know, the development of testing and, you know, rolling out of testing locations. There's just been a wide variety of things that the federal government has not really done well and has even just said, well, we're just not going to do it.

And so what you've seen is sort of civil society groups trying to step into the breach - coming up with testing plans, creating coalitions between states, you know, dreaming up ways of doing testing in new ways, coming up with ways of using our existing, you know, PCR testing capacity more efficiently. And it's - you know, it's a loose confederation. These people are not all working, you know, in precise concert with each other.

But when you just take a step back, you just see the way that so many people have tried to step into the vacuum created by the spotty, overwhelmed and sometimes just terrible federal response (laughter). And it's not ideal that those are the people, right? Because philanthropists and technologists and economists and, you know, reporters like us, we're not the people who should be filling the gaps that the federal government presumably should be filling. And yet that's what has happened, you know, time and time again during this pandemic.

GROSS: Oh, I'm glad at least it's happening. I mean, it's so frustrating 'cause you think, like, we live in America. We have brilliant scientists. We have brilliant epidemiologists. We have brilliant tech people. How come we don't have more tools to deal with a pandemic?

MADRIGAL: Yeah, it's really been frustrating, I think, and quite an emotional experience, I think, for those of us who've been close up watching these things happen precisely for that reason. You know, the administrative capacity of the American state seems to be a real question right now, and that's not something that I kind of expected. You know, you know that, you know, the federal government has been hollowed out in certain ways over the years. But I kind of always thought, particularly on this topic, I'm like, you know, public health infectious disease response, this was, like, an American strong suit. Countries pattern their national infectious disease response after the CDC, you know, for decades. We were, supposedly, the experts on this. And here comes, you know, a quite serious challenge to, you know, that system in the U.S. And it just crumbled right before our eyes. And, you know, how much of that is due to the Trump administration? How much of that is due to the hollowing out of the CDC? How much of that is due to the just totally unprecedented nature of this crisis?

My own view - or at least one kind of central plank of my view - is that we had a lot of pandemic preparedness plans, you know, something people have been working on for a long time. But they all kind of assumed that the data underlying these plans would be good. They assumed that, you know, we'd have testing data that was high quality and comprehensive. They assumed that that data would flow through states and into the federal government and back out in ways that were useful and that would kind of set off all the proper alarms to sort of trigger the right responses within the system. And that's not what happened.

The testing system broke down immediately because of the problems at the CDC, then was entirely overwhelmed because the pandemic had spread so far by the time we really started to see what was going on through testing. And then the data systems that were created have just been a total scramble. And so decision-makers and leaders at all levels of government have not really had access to the information that they needed to make the decisions that were required, which was how many people were sick. And absent that crucial piece of information, the system collapsed. And so here we are, you know, six months in and with a response that sort of pales in comparison to many, many, many countries around the world.

GROSS: Let's take a short break here. And then we'll talk some more. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine and co-founder of the COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on test cases, hospitalizations and patient outcomes. We'll be right back. This is FRESH AIR.

(SOUNDBITE OF PETE YORN SONG, "ON YOUR SIDE")

GROSS: This is FRESH AIR. If you're just joining us, my guest is Alexis Madrigal, a staff writer at The Atlantic magazine who's been reporting on the spread of the coronavirus and efforts to contain the spread. He's also co-founder of the COVID Tracking Project, which has been collecting data from each U.S. state and territory, compiling the latest numbers on tests, cases, hospitalizations and patient outcomes.

Alexis, why did you feel a need to create the COVID Tracking Project, which is collecting data? Are you collecting data that the government is not collecting?

MADRIGAL: It's an interesting question. We collect state government data. And the reason we needed to do it was we just didn't know how many people had been tested back in March, you know? My reporting partner Rob Meyer and I wanted to know, well, you know, we know there's very few cases in the U.S. You know, this was in in early March. But how many people have we tested? And when we tried to find those numbers from the CDC, we weren't able to do so. We didn't know how many people had been tested. And so we decided that we would look at every state, you know, call up their health departments, look at their websites and begin to compile a number of people.

And basically, right after we published the story, I heard from a friend of mine, who, actually, you know, guy I've known since college, named Jeff Hammerbacher, who'd been doing the same thing. And we decided to join forces with a woman named Erin Kissane and create an entity that could track this through time. And, you know, to be honest, we thought we'd just be doing it for a few weeks because it seems so obvious that the CDC would provide this information to the American public. And it seemed like the best place for it would be the CDC. And, you know, that was now (laughter) almost six months ago and we're still doing it. And it is still the best source of data for a lot of testing information.

GROSS: Is the government actually relying on your project?

MADRIGAL: That's an interesting question. I think, through time, I'm pretty sure the government has, in fact, relied on it. You know, we were cited in a coronavirus taskforce document, for example. And we know that states rely on the data in order to make decisions about how they deal with other states' residents. So yes, it's become sort of part of this infrastructure, which, again, from our perspective, we're doing the best we can. But it sometimes feels ludicrous that, you know, a team of volunteers and some, you know, people being paid stipends, (laughter) basically, are the ones responsible for, you know, compiling this data from the states and making it available in usable forms that, you know, modelers and states and, you know, city officials can use.

GROSS: What are some of the trends you're seeing now in the data?

MADRIGAL: Well, you know, I think one of the good trends is that hospitalizations are way down from late July. We also see cases down. On the bad side, deaths have been pretty steady, as they are recorded sort of more slowly than the other metrics. And, troublingly, we've seen testing actually falling across the Sunbelt. So in the Northeast, as the cases were declining and the outbreak was easing, Northeastern states just kept testing more and more people. And that's one of the reasons, I think, why the Northeastern states have been pretty successful at containing the virus.

In the South, that's not what we're seeing. As cases decline, the states are testing fewer and fewer people. And that doesn't make a lot of sense. And you need to test people so that you know who has the virus, and you can contain outbreaks so that it doesn't get back out of control again. And that's probably the most troubling trend in the data right now.

GROSS: One of the things your COVID Tracking Project is tracking is trends in different racial and ethnic groups because there are certain groups that are more vulnerable to the virus for various reasons. Are there trends that you're seeing there now that we are more aware of how the virus is more prevalent in certain racial and ethnic groups and in certain racial ethnic communities for various reasons? Are you seeing any positive trends because of the awareness of this?

MADRIGAL: You know, at kind of a state or national level, I don't think the trends are actually better. You know, you can't fix the structural and systemic racism in this country overnight, and you certainly can't do it in the midst of a pandemic. There are some ameliorative measures that people can take. And I think, you know, increasing testing in poor communities and in vulnerable communities has definitely been a strategy that some cities are employing.

But, really, the numbers have been quite steady. Black people, Latins, Native American communities are being infected at higher rates and, in many cases, also dying at higher rates and dying younger. And that has been true all along, you know? And I think what, for me, it really points out is how systemic these issues really are because it's, you know, neighborhoods that, as a result of environmental racism, are polluted and polluted in ways that stress people's bodies. And it's, you know, food deserts. It's just the, like, raw racism of the health care system.

I mean, there's so many different components that, of course, it has had this disproportionate impact on communities of color. And you got to fix the system to get better outcomes. And my hope is that, you know, the combination of the Black Lives Matter protests that have happened this summer in conjunction with, you know, COVID kind of lead people to think about structural solutions to these problems for vulnerable communities.

GROSS: Well, Alexis Madrigal, it was great to talk with you. Thank you for the work that you're doing and thank you for being on our show.

MADRIGAL: Thank you so much for having me.

GROSS: Alexis Madrigal is a staff writer at The Atlantic magazine. His latest article is called "The Plan That Could Give Us Our Lives Back." He's also co-founder of The COVID Tracking Project.

We recorded our interview yesterday. After we recorded it, Abbott Laboratories announced that the FDA granted emergency use authorization for Abbott's new rapid antigen test for COVID-19. It's a small, portable unit, costs $5 and can deliver results in 15 minutes. But it's not a do-it-yourself, saliva-based home testing kit. It uses a nasal swab - less invasive than the PCR test - and requires a health care practitioner to administer it. It could be used to check people before entering large venues, including schools and workplaces. Abbott will begin shipping tens of millions of these tests in September.

(SOUNDBITE OF MUSIC)

GROSS: If you'd like to catch up on FRESH AIR interviews you missed - like this week's interviews with Brian Stelter, CNN's chief media correspondent and author of a new book about FOX News; or Jean Guerrero, author of "Hatemonger: Stephen Miller, Donald Trump, And The White Nationalist Agenda"; or Rick Perlstein, author of "Reaganland" - check out our podcast. You'll find lots of FRESH AIR interviews.

FRESH AIR's executive producer is Danny Miller. Our technical director and engineer is Audrey Bentham. Our interviews and reviews are produced and edited by Amy Salit, Phyllis Myers, Roberta Shorrock, Sam Briger, Lauren Krenzel, Heidi Saman, Therese Madden, Thea Chaloner, Joel Wolfram and Kayla Lattimore. Our associate producer of digital media is Molly Seavy-Nesper. Seth Kelley directed today's show. I'm Terry Gross.

(SOUNDBITE OF MUSIC)

Transcripts are created on a rush deadline, and accuracy and availability may vary. This text may not be in its final form and may be updated or revised in the future. Please be aware that the authoritative record of Fresh Air interviews and reviews are the audio recordings of each segment.