RUTHFULLY YOURS

The story of how the country went from nothing to more than half-a-million COVID tests on some days is a tale of inspired private–public cooperation. After bad initial stumbles when the Centers for Disease Control fouled up the initial test (at a time when it had a monopoly on testing) and when a Food and Drug Administration regulatory bottleneck stymied the development of testing in the private sector, the administration found its footing.

It used cooperative data, governmental authorities, relationships with the private sector, and improvisation-on-the-fly to work through supply shortages and other problems. Giroir’s team, a group of trouble-shooting aides around White House adviser Jared Kushner, the FDA, and the Department of Health and Human Services all played important roles. Meanwhile, private companies — often working hand-and-glove with the administration — quickly innovated and scaled up their production of everything from swabs to test kits.

At the beginning of March, there were fewer than 1,000 tests in the country on most days. By the end of March, there were more than 100,000 a day. The number climbed again before hovering somewhere around 150,000 a day for much of April.

Then it bumped up again, to more than 200,000 at the end of the month. It surpassed 400,000 for the first time on May 17. On May 29, it was nearly 500,000. It’s been in the range of 400,000 to 500,000 most days since, with a high of more than 580,000 just this just Friday.

More than 23 million tests have been run altogether, more than 3 million over just the last week. In April, 32 million tests were produced, and in May, 39 million, with the trajectory further upward. Giroir expects to be able to perform 50 million tests per month by September.

It’s obviously not an apples-to-apples comparison, but the hard-hit states like New York (151,000 per million), New Jersey (124,000 per million), and Massachusetts (111,000 per million) have higher rates of per capita testing than the major countries — including Spain (96,000), the United Kingdom (100,000), and Portugal (96,000) — with the highest levels of testing per capita.

The positivity rate of tests, a key indicator of whether enough people are being tested, has been dropping nationally since mid April and is now below 5 percent. The World Health Organization has set a rough benchmark of 5 percent. (If you’re above it, you aren’t testing enough.)

Stephen MacMillan, the CEO of the diagnostic manufacturer Hologic, says of the roughly 10 million tests that were conducted in May, “there had never in the history of diagnostics been a test to develop faster or used in this volume.”

“I know for the outside world,” he adds, “they look at it and say, ‘Why can’t you do this?’ But when you truly understand the magnitude, the industry or the world has never seen volume of one diagnostic test conducted in this quantity — never close.”

The problem with testing nationally is becoming less a shortfall of availability of the tests and more a shortfall of people showing up to get tested. An insider in the diagnostics industry says that laboratories are reporting that they are “sample starved” — i.e., they aren’t getting enough specimens. He notes, “We have all seen stories about sample-collection sites in some regions not seeing that many patients.”

An HHS official says that in May there was the capacity to do twice as many tests as were actually performed, calling it a function of “allocation and efficiency, but more just demand.” Says Giroir, “We really see areas in the country now that there’s more tests available than people who want to get tested or the need for testing.”

How this came to be is a crucial part of the story of the response to COVID-19, although one that has been neglected.

The Media Focused Solely on FailureThere has been a lot of media attention, and justifiably so, on the initial CDC and FDA foul-ups and the initial overpromising on testing. But as the administration moved beyond that and pursued a considered but urgent approach to testing, the media persisted in harsh critiques that betrayed a misunderstanding of what was really happening.

The media often insisted that there was no testing plan, simply because the plan wasn’t centralized enough, a line of attack pursued by Democrats as well.

The Guardian in early May ran a piece titled “No leadership and no plan: Is Trump about to fail the U.S. on coronavirus testing?” Noting that testing had increased, it warned that “declarations of false victory by the Trump administration and a vacuum of federal leadership have undermined the endeavor.” It asserted that the U.S. had been left “without a unified, coherent strategy for testing and contact tracing to contain a virus that does not respect state borders and has already killed more than 60,000 Americans.”

Paul Reider of Princeton University told the paper, “If Jared Kushner wanted to do something decent, and Vice President Pence, they could try to standardize and distribute nationally a global test.”

A New York Times headline on the administration’s testing strategy referred to “the hunger games.”

In mid April, the headline of the Washington Post newsletter the Daily 202 said, alluding to the administration’s reopening guidelines: “Trump’s new guidance makes lack of national coronavirus testing strategy more glaring.”

A Washington Post news report at the time took a dim view of “the lack of a federal strategy,” saying that “the government has been unable to compel test manufacturers to dramatically increase the number of tests produced.”

Vox deemed April, just as the country was on the cusp of an enormous increase in testing, a “lost month.” The publication interviewed an expert who insisted, “Testing is not going to be resolved until the federal government takes ownership of expanding it; as long as they keep kicking it to the states, we will remain stuck.”

(The day the piece was published, the U.S. performed 273,000 tests; a little more than a week later, it performed more than 360,000.)

Democrats struck a similar note. In early May, Chuck Schumer and his Senate colleagues wrote the administration that they were “deeply troubled by the lack of detail and strategy in your testing blueprint, and we fundamentally reject the notion that the federal government bears this little responsibility in increasing testing capacity. The Trump Administration must not shirk its responsibilities and leave states and tribes to fend for themselves.”

In late May, Nancy Pelosi and Chuck Schumer  for choosing to “dump the burden onto the states.”

Schumer has continually agitated for use of the Defense Production Act (DPA), sometimes sounding as if he’s unaware that it’s actually been employed, if not as often or as sweepingly as he’d like. “The president will invoke the Defense Production Act for meat plants, but not for testing,” he complained on MSNBC’s Morning Joe in early May.

He predicted that “there will be no way to ramp up and get the number of tests, and the auxiliary things like the swabs that we need, unless the federal government takes over.”

The minority leader apparently hasn’t stopped to consider why he was wrong, nor has much of the media.

A Complicated EcosystemBefore delving into that, some basics on testing.

Testing is reliant on a complicated ecosystem prone to disruption at any number of places. If there aren’t enough swabs, samples can’t be collected. If there isn’t enough personal protective equipment, personnel can’t be protected while collecting samples. It there aren’t enough tubes or transport media, the samples can’t be sent out to be tested. If there aren’t enough reagents — the chemicals necessary to detect the virus — the tests can’t be run. It doesn’t matter if most of the components are in ample supply if there’s a shortage in one of them.

There are a couple of types of tests. Molecular tests, the workhorse on COVID-19, identify the genetic material of the virus during an active infection; antigen tests, not nearly as prevalent, look for specific proteins on the surface of the virus, also during an active infection; and serological tests identify antibodies to the virus in the blood, indicating a past infection.

On top of that, there is a variety of different molecular tests. There are lab-developed tests, which are what they sound like — tests developed by commercial labs like Quest or LabCorp. The labs process these tests themselves from samples coming from hospitals, physician offices, urgent-care clinics, and mobile sites. These were the first tests available other than what came from the CDC.

Then, there are the diagnostic manufacturers, companies like Hologic, Roche, Abbott, and Thermo Fisher, that develop test kits that can be run on their specific machines, which are at hospitals and labs.

There are also point-of-care tests, where the sample is taken and the test is run, at the same place.

And, finally, at-home collection kits have been developed.

The diversity of platforms is ultimately a strength, but it makes for a mosaic where everything has to be matched — e.g., it does no one any good to have a Roche machine if there isn’t a Roche test available to run on it.

Working through this complexity was one of the challenges of boosting testing to another level.

The Strategy that Supposedly Didn’t Exist The Trump administration’s general approach was to catalyze and support the private sector while working with the states to identify the testing capacity available to them and to secure the necessary supplies to meet their goals. The FDA worked to approve new tests and technologies as rapidly as possible, which was enormously important to nearly every aspect of testing. The Defense Production Act was used, but sparingly, and as way to buttress companies rather than take them over.

The emphasis was on improvisation and innovation. “It was very clear that the normal institutions of government under any administration wasn’t going to solve this and could not solve it, particularly in a country as vast as America,” says Giroir.

In early March, the White House called leaders of the diagnostic manufacturers and the commercial labs to talk about scaling up. Dr. Deborah Birx, the White House coronavirus-response coordinator, describes it as “really a call to action by the president and vice president to say very clearly, ‘We need enhanced and greater diagnostic laboratory capacity. You are the private sector that have the technical ability to do this.’”

It coincided, weeks later, with the first big bump in testing.

Scott Whitaker, the CEO of the medical-technology association AdvaMed, explains the progression, from the first lift in numbers in March to the plateau in April to the recent sharp increase: “You get a handful of companies that were quick to the gate, and you started running all those tests on the machines that they had available with the labs. The next round comes in and scales up even more dramatically on the high-throughput machines,” machines capable of running a lot of tests quickly.

“And,” he continues, “that’s when it started to take off again.”

A key event was the FDA’s acting to clear away in late February the hurdles for commercial labs to begin offering tests. Within days, labs were ready to accept specimens, and they have ramped up ever since. By March 22, the labs that are part of the American Clinical Laboratory Association, encompassing the big players in the industry, had collectively performed 200,000 tests; by the end of May, they were consistently performing more than 200,000 a day.

“Even for the larger players,” a lab industry insider explains, “the LabCorps and the Quests, they were able to bring it online at one of their labs, say. And then it’s just gradual, that expanding out.” As the shortage of supplies and backlogs from the pent-up demand for testing were worked through, the testing reached a different level, and “then the issue became actually unused capacity.”

Meanwhile, the Trump team worked to see that every component necessary to testing was being manufactured.

It put in big orders to assure companies that they’d have a customer if they drastically scaled up production. It went to new players to get them to pitch in — Oak Ridge National Laboratory started making metal molds for plastic tubes, and Coca-Cola started making tubes. It cleared the way to use sterile saline, much easier to procure than traditional transport media, for some tests. It got a so-called fill-finish company to begin putting transport media, instead of its usual injectable vaccines, in vials.

And so on, all across the supply chain.

The DPA was a tool, but not the main event. The act can be usefully deployed to rate orders and move the government to the front of the line, and to make investments in companies to increase their production. No one thought that using it to nationalize industry or take over supply chains made sense.

Giroir calls that idea “absolutely insane.” He asks, “Why would we buy it all up, bring it to a warehouse in FEMA, and try to redistribute it, when they’ve got mature distribution chains that they use every single day throughout the country?”

Besides, companies were willing partners. “It’s unclear,” a senior administration official says, “what’s necessary when you go to a company like U.S. Cotton, or Honeywell [which have been making swabs and N-95 masks, respectively], or some of these other companies that are basically like, ‘Yeah, we’re crazy-producing as much as possible.’”

The official adds, “Our job is to remove any barriers, but then you let the company run free and do things aggressively. The United States government taking over a company? Why the hell would you do that?”

Similarly, the decision to defer to states on many details of testing was based on a modest view of how much the federal could possibly know about the particular conditions in each state. “It would be impossible or near impossible,” an HHS official says, “for the federal government to decide who in each state needs a test, where those testing locations should be, which labs should process which tasks.”

Giroir notes that “New York City is different from Montana and it’s really different from El Paso,” so it wasn’t practical to make a single national testing plan apply to everyone. He recalls one instance in Texas when someone came on horseback to a drive-by testing site.

But the administration did go out of its way to make sure that hot spots were getting more tests. Birx says of the state-level numbers as of late May, “It is not a coincidence that New York was able to test 9 percent of their population, or Massachusetts 8 percent of their population, or Louisiana 7 percent of their population, or Rhode Island 13 percent of their population. That was possible because tests were allocated very much by need.”

But the administration obviously wanted to see the tests as easily and widely available as possible — hence a focus on mobile sites.

From Nothing to 1,500Kushner’s team made mobile sites a priority, driving a dash in mid March to get them up and running within days.

The effort started from almost zero. “I think what you saw was when we started that work,” the HHS officials says, “I want to say there were only two or three mobile testing sites across the country. I think they’d all opened within a day or two of us starting that work.”

And they immediately ran into supply constraints. “As we started to dig in,” he says, “which I think happened on a Thursday or a Friday, we quickly realized probably half of the supplies that we needed to operate the sites were not in this strategic national stockpile, including things like swabs. And so literally within about a 48- to 72-hour period, we built an operational plan for all of the sites. We had gone out and secured all the supplies we needed. And then I want to say by Tuesday night we had shipped them out to all the locations across the country that we were starting to set up.”

The several dozen initial sites would serve as prototypes. “Right before we launched, we did a test drive where we did a mock site up in Maryland, and we had a number of private-sector retailers from across the country come in just to observe a workflow of us working,” the HHS official says.

“As we originally went out in the very beginning, it was a partnership, privatepublic partnership,” recalls Kathleen Yeager, who heads pharmacy care for the National Association of Chain Drug Stores. “The public-health officials were helping our members set this up, and our members were volunteering and bringing their resources and assets to the table and they set up the mobile sites. So they were learning as they were going.”

These sites were relatively unsophisticated. Given the way the swabbing was done at the beginning, says a senior administration official, the sites “were manned with like a dozen to two dozen people, because everybody had to be in PPE and they all had to collect it.” The operational workflow was steadily honed not to require so much PPE, and then came the FDA-approved self-swabbing, which was transformative. “You just drive up,” says the senior administration official, “and they give you a swab, you can self-administer.”

“That was really absolutely key,” says Birx.

As retailers grappled with creating mobile sites, they were hampered at first by shortages of PPE, by logjams at labs, and by how long it would take to get a result once a specimen was sent to a lab. The press pronounced the mobile-testing initative an abject failure, but the problems were getting worked through and obstacles removed.

HHS issued emergency guidance in early April to allow pharmacists to order and administer COVID-19 tests. Besides self-swabbing, the FDA also approved point-of-care testing, opening up the possibility of getting a test result back in hours rather than days or even weeks.

The pharmacies were a natural to play a big role. The last several years they have begun doing point-of-care testing for flu, strep throat, HIV, and hepatitis C. They already had the testing machines in place, and they are embedded in almost every community in the nation.

Sites have been set up in outside locations all over the country, often in the parking lots of schools and banks. (The idea is to avoid having someone who might be infected with COVID actually come into a store.) Some sites can offer point-of-care testing, while others are still sending samples to labs, although results come back more quickly than at the outset.

About two months after the press left the mobile-testing initative on the ash heap of history, retailers are running about 1,500 such sites around the country.

A 50-State Approach A lot of the action on testing has been at the state level, so the administration has devoted considerable effort to supporting the states, which has been widely misunderstood.

“It just gets me a little fired up when people say we’ve thrown it all to the states,” says Giroir, “because I spent ten weeks working 20 hours a day with a team of about 100 people making this happen, and that doesn’t define it to me of throwing it back to the states.”

A major initiative was gathering data from manufacturers and getting it to the states so they understood what testing capacity they had — what kind of machines and at which labs. Birx  calls it “eye-opening that, really, we had ten to 15 times the capacity that we were utilizing.”

In the ordinary course of business, there’s no reason for a state to know the number and whereabouts of testing machines. Scott Whitaker of AdvaMed explains that orders for machines are “based on labs requesting them, or hospitals requesting them, or doctor’s office or others requesting them. So you fill an order for a machine. You send that to location X. They say, we need X number of test kits to run these tests. Then you send those to a location.” The state government has nothing to do with it.

To gather the information about the machines, Giroir says, the Trump team went “manufacturer to manufacturer, asking them. We got all the data.” Then they passed it along to the states.

The senior administration official explains, “So what we did is we basically had to have calls with all the governors and a whole bunch individually that basically was just saying, ‘This is where your testing locations are, and this is where you should send them and collect more.’” The official continues, “We literally gave them addresses. They’re like big books of addresses of their labs that we had to tell them, and tell them this is where their capacity is.”

A White House official recalls, “One of the funny stories that Abbott told us was that they had a governor that was screaming at them saying, ‘I need more tests. I need more capacity.’ They basically turned to him and said, ‘Look, you have all these Abbott machines in your state and you’re only using 10 percent of your capacity.’”

The Trump team had a meeting with New York governor Andrew Cuomo and says it pointed out the additional testing capacity in New York State. “The lights kind of went on,” says Giroir.

The better knowledge of the existing testing infrastructure made it possible to better utilize it. “For example,” the HHS official explains, “there may have been physician offices or folks who felt like, I don’t know where to send my labs. When I send it to LabCorp, it takes seven days to get back. But what they didn’t realize is that the university lab that was ten minutes from their office actually had testing capacity. And so by creating that transparency for states, they were able to work out those local strategies that would more efficiently utilize their existing resource.”

The data allowed for granular interventions. The official cites another example. Something called the Cepheid machine is common, he says, in rural areas. The Trump team looked for places where half the machines are Cepheids or there are no other high-throughput machines within 50 miles. Then it worked with Cepheid to make sure those areas were getting their share of its tests.

Another other major part of the work with the states was coordinating on basic supplies.

According to the White House official, the approach shifted. At the beginning, the emphasis was on trying to flood the private market with tests and supplies for states to draw on. It evolved into working more directly with the states and saying, “Okay, what do you need?”

“What we did,” says the HHS official, “is we looked at each state’s case-testing goal for May and then we looked at the sum total of what lab testing supplies they were getting, to make sure that their supplies were greater than their testing goal. In the states where there was a gap, we’ve worked with the state to figure out how to close that gap. For example, Thermo Fisher had a very high number of tests that had actually not been purchased. For the states that had the biggest gaps, we were able to connect the CEO of Thermo Fisher directly to the governor’s office so that they could procure those additional tests.”

Giroir is confident that every state has been getting the testing supplies that it needs.

The Swab MobilizationThe swab is one of them, a basic implement that we didn’t have nearly enough of at the outset.

There was no reason not to go big on swabs. The White House advisor explains, “We didn’t know what number we were going to have to solve for, so swabs are not very expensive. We basically said, just contract for a billion swabs. Just figure out how to get it going as big as possible knowing that, like with ventilators, if we’re satisfying the American demand there will be an export market for these things and we also want to have onshore ability to do it.”

One focus was the only facility making them before the crisis hit, Puritan in Maine. Using the DPA, the administration buttressed the company’s operation. The HHS official says that “it was used to help them actually build another facility and purchase the machinery that they needed to run additional lines.” Giroir notes, “We’ll get an extra 20 million swabs a month out of that from using the DPA.”

Then a new manufacturer came on line, U.S. Cotton. It’s parent company, Parkdale Mills, was already working with the government to make masks. “It kind of came up that, ‘Hey, you guys are U.S. Cotton also,’” John Nims, the CEO of U.S. Cotton, recalls, And, of course, the answer was, ‘Yes.’ And they said, ‘Can you do swabs?’”

The answer to that was “yes,” too, although it wasn’t so simple.

U.S. Cotton makes Q-tips, which, of course, have a cotton tip and a paper stick. That wouldn’t work for testing. The tip had to be polyester, and the stick plastic. These aren’t minor differences. “Polyester is a lot different to work with than cotton, from our perspective,” says Nims, explaining that it’s “kind of a slippery fiber” requiring adjustments to the machines making the swabs. Then there was the matter of length.

At first, the government wanted it to make six-inch swabs, according to Nims. But its machines weren’t set up for that. The company experimented and worked with the FDA and the Gates Foundation to arrive at a workable alternative. It settled on a 73-millimeter, or three-inch, swab, less invasive than the longer versions. It also fits neatly into the tube after the sample has been taken, and it doesn’t have to be broken off like the longer, scored swabs.

U.S. Cotton already produces seven days a week, 24 hours a day. Because the materials for the test swabs are more difficult to work with, it has had to run more machines to do the work, but it has been going full tilt.

As of a couple of weeks ago, it was shipping 2 million a week directly to FEMA, and another 1 million, on the way to 1.5 million, a week to packagers to put individual swabs in individual packages, ultimately to get sent to FEMA as well.

“There’s a lot of interest from labs, and states, and universities for us to produce swabs, have them packaged, and then shipped to them,” Nims says. “I don’t know how long this will continue. We are ramping up to be able to do a lot of swabs, but we’ll see how many are needed. We’re prepared to go as long as there’s a desire for them. And when that desire goes away, we’ll just move back to our cotton-swab business.”

The Manufacturers Step UpThen, there were the diagnostic manufacturers.

At the beginning, Scott Whitaker says, the Trump team was “very aggressive in their outreach to every company, engaging through the FDA, primarily, with each of these companies to determine whether or not they could make a test, manufacture a test, and get it scaled up quickly enough to begin testing on a tremendous level.”

Giroir says the administration has continued the close relationship with each of the diagnostic companies. “We really have a relationship manager that talks to that company three or four times a week,” he explains, “so we understand where they are” and what “any holdups with their projections of the future are.”

The FDA pushed through a steady diet of speedy emergency-use authorizations. “A CEO of one of our companies,” Whitaker recalls, “said to me after they got their first test approved, ‘That usually takes us a year. It took us a week and a half.’”

More than 60 molecular tests have been authorized for use.

The experience of Hologic, a major player in this space, shows how we’ve got to a different place on testing. The company is focused on women’s-health diagnostic and medical-device products. Along with Roche, it is the leader in molecular diagnostics. Its state-of-the-art machines are called the Panther and the Panther Fusion, which attaches to the Panther.

According to Hologic’s CEO, Stephen MacMillan, early on a couple of its scientists focused on the outbreak in China, including one with family in Wuhan, and encouraged the development of a test even when the CDC said it was going to be taking the lead.  “So we, in a skunkworks way,” says MacMillan, “started working on developing, effectively, our first test, which goes in our Panther Fusion. It leverages a lot of the flu-testing technology that we use, and there was a clear pathway for approval.”

That got approved in late March. “At that point,” he says, “we were getting inundated with calls from governors, customers, everybody, and including a whole bunch of world leaders saying, Can we help them?”

Hologic could make about 600,000 of the test per month, which was much more than the number of its flu test, at about 50,000 a month. But it realized that the demand was going to be bigger.

It began work on a higher-volume test using a form of nucleic-acid testing. It got approved in early May, and Hologic is making about a million of the tests a week. It runs on its high-throughput Panther platform. There are 1,000 Panthers installed in the United States, and each of them can run 1,000 tests a day. Results are available within a couple of hours.

MacMillan says that the first shipment of this second, high-volume test went to hundreds of labs and hospitals around the country in late April and that it was coordinated with the coronavirus task force to put a special emphasis on hotspots. “That’s where things really kicked up in May,” he says of the overall testing number. “I think it’s as we started to get really major volume on high-throughput instruments around the country.”  And, of course, as he notes, “other companies were ramping up capacity as well.”

The innovation around a new test is complex — using the reagents to identify a strand of DNA, developing the software to program the Panther machines to run the tests, making sure it all works and is accurate.

“We’ll do it in the lab, where we’ll inject the tube with the actual virus,” MacMillan says. “We’ll then run our tests through it, and we’ll make sure, Are we getting it when we know there’s the virus in it? Also, are we not ruling it out when it’s not?”

“Literally, our R&D team, they were working through the weekends. We had R&D people coming in at three in the morning on Fridays, Saturdays, and Sundays,” he continues. “They put the tests going in before they leave work and it would be running, and they’d come back in as soon as it was coming off.”

“We’ve committed that we’ll provide at least a million tests a week, from here going forward. We are scaling additional capacity, to be able to provide even more than that. We haven’t stated it publicly yet, but e by September we’ll be delivering well over a million a week, should the country need it.”

How Many Tests? How much testing we’ll need is, like everything with COVID, contested. A HHS report in late May was criticized for saying we needed only 300,000 tests a day. Regardless, we are well above that now, and some days approaching twice that number.

The more extravagant estimates, like that the Edmond J. Safra Center at Harvard, say we’ll need millions a day.

The administration is skeptical of those kind of numbers. Birx says that the questions she asks regarding the level of testing in the states are

Do they have enough tests to identify outbreaks, stop outbreaks? Do they have enough tests to protect the vulnerable individuals in their communities where they’re decreasing hospitalizations, decreasing deaths? Do they have enough tests to ensure that they can diagnose every single person with symptoms and contact-trace? To me, that’s the indicator. But I can’t find anybody who says, “That’s where I came up with the 3 million a day.”

What shouldn’t be a matter of dispute is that after a stumbling out of the gate, we are in a much better place on testing — thanks, in large part, to a Trump-administration strategy that the press scorned, or said didn’t exist.