What we have done over this past year to develop the world’s largest testing system is not just unprecedented, but a uniquely American achievement—something that we could do because we believed in enlisting every creative corner of society and bringing the public and private sectors together.
Thank you, Avik, for that introduction, and thank you everyone for joining us. Today, I want to focus on a topic that closely intersects with much of the work FREOPP does: how the government and the private sector worked together on our national COVID-19 testing strategy over the past year, and the lessons this has for future healthcare and public health efforts.
Today, it is indisputable that the United States has built the most extensive testing system and strategy of any major country: While most European nations require a patient to present with symptoms in order to receive a test provided by the public sector, both the federal government and every single state have supported broad availability of testing, through both private and public channels, for people with or without symptoms and with or without known exposure. For testing people with symptoms and other key categories, the United States has invested in, developed, and has now provided broader access to rapid antigen tests than any other large country on the planet.
As I reflect on our experiences and learning over the past year, I believe there are three key lessons we can take from our testing program: first, how the federal government got out of the way of test development in safe and sensible ways; second, how the federal government has played a strategic role in boosting the development, manufacturing, and distribution of tests; and third, some limitations we've learned about how our healthcare system can adapt to test for a novel virus.
To explain how we got here, I want to take everyone back to the beginning, to understand the context in which tests were first developed—and what it actually takes to make a test for a novel virus.
Back on January 10, China finally shared the viral sequence for the novel coronavirus. The day before that, CDC and FDA had agreed to begin work on a test for this virus—and that test, developed by CDC, was used to confirm cases arriving from China in the United States during January and early February.
To understand what we needed to do to begin detecting the virus, it's worth remembering what you need to develop a test of the type used to confirm viruses like this: an RT-PCR test, short for reverse-transcription polymerase chain reaction.
Much of the COVID-19 testing now done in the United States is done through rapid tests, some of which are literally just a card with chemicals on it, like a pregnancy test, or involve depositing a sample into a machine that produces a result within 15 minutes or so.
But that's not the backbone of how we test for a novel virus, certainly not in the early days. Obviously, that kind of tailor-made point-of-care product is too complicated to develop in days or weeks. That's why we use a PCR test, which is more like a standard protocol or recipe: a combination of preexisting components coupled with a specific nucleic acid sequence to target and amplify specific genes of the novel virus.
Here are the basics of a PCR test: First, you need to design primers for the virus. A primer is a small strand of genetic material that matches a unique sequence of the virus's DNA. Most PCR tests target multiple different sequences of the virus genome. Primers are not difficult to pick or develop—they are something that most commercial labs and all academic medical centers and universities can generally do. Once the primer has been mixed with your biological sample that may contain the virus, you combine it with reagents that, with the addition of free DNA nucleotides, can replicate any DNA strands to which the primer has bound itself.
You heat and cool the sample to replicate the DNA repeatedly until, even if there were barely any cells with that targeted DNA strand in your sample, you end up with a significant amount of the DNA you wanted to detect. Then, you look for a signal from a probe, which contains your primers and a marker that will display a visual indication.
Of course, errors can occur: These tests can easily go wrong if the wrong primers are selected, if there is contamination of the sample, of manufactured components, or of the control against which the sample is being judged, or if something else goes wrong.
But there also isn't any complex manufacturing involved: The reagents and free nucleotides I described are stocked by most diagnostic labs, of which there are thousands across America. The primer can be synthesized using technologies in the individual laboratories or ordered from commercial vendors.
It is not an overstatement to say that a relatively sophisticated high school student in their high school lab could develop an RT-PCR test. It is important to remember that, in these early stages, it is really more of a menu or recipe than the sort of kit that non-clinicians might think of. Most people without experience in this area probably think of a "test" the way we think of a strep or flu test, a sample is taken and put in a machine and a result is read, or we put a sample on a piece of paper and it gives us a colored indication of positive or negative, like a pregnancy test. But when dealing with a new virus, these PCR tests really are often quite manual.
The test that the CDC developed was this kind of test. They developed a primer—three, actually, two for SARS-CoV-2 and another for any type of coronavirus, in order to detect any broad shift in this novel pathogen. They decided on the mix of reagents that made sense for spurring the chain reaction that replicates enough RNA to yield a result.
That's what they sent to public health labs around the country: primers, controls, and some of the reagents necessary, with instructions on how to mix and use them.
Other tests around the world were similar. Just remember what we're not talking about: We're not talking about a pre-configured, easy to use cartridge that one puts in a machine with the primer and reagents already packaged together. It isn't a lateral flow test, like a pregnancy test. Any other test around the world at this time was exactly what the CDC was doing: selecting a sequence, manufacturing a primer, and using that primer with a selection of reagents to amplify the RNA.
When you hear people ask why the U.S. didn't use "the WHO test," it's worth knowing that there was no such thing. The WHO had not developed or cleared a test. Rather, it had contracted with a German test maker to produce an unapproved set of primers and a mix of reagents, which the WHO then sent to dozens of low-income countries that either didn't have the lab capacity to make the primers, didn't have easy access to the reagents they'd need, or both.
It was, in essence, the same as the CDC test or any other test being used at the time around the world. It lacked any form of regulatory approval, as the WHO did not put any products through its process until April, and European countries do not hold such diagnostics to the same regulatory standards we did at the time.
The same goes for South Korea's commendable scale-up of testing. It was largely one South Korean private sector manufacturer that developed specific primers and the right mix of reagents, and started performing tests just like the CDC was—but on a larger scale, without the unfortunate mistakes that occurred in CDC's manufacturing of the testing kits they eventually distributed to other public health labs.
It just so happened that South Korea had a fairly concentrated outbreak. Unlike the United States, where cases were arriving from China and Europe and seeding outbreaks across the country, South Korea's outbreak began with a super-spreader event around a megachurch, which accounted for more than half of their early cases. They were able to focus their testing capacity on that outbreak, with highly effective results. They used strict quarantines and investigative tools that we would not find acceptable in the United States.
The only—repeat, only—difference between South Korea and the United States testing protocol was that we had a lab in Atlanta doing the RT-PCR testing and South Korea had private-sector manufacturers making these primers and reagents and labs in South Korea were using them at a scale appropriate to their localized outbreak. CDC was doing the exact same thing, just in Atlanta.
Now, that is where we stood around the end of January and beginning of February. As we had made clear, we could not hermetically seal the United States off from the virus, and we would need more than just the CDC's lab testing capacity to detect it across the country. Labs around the U.S.—the more than 100 public health labs associated with state and local health departments—were initially expected to rely on the primers and reagents being shipped by CDC and CDC's subcontractor.
The solution to this bottleneck might seem fairly clear: Let these labs develop their own tests, with their own materials, and start churning out tests themselves. We could have had a hundred or more labs doing what that one lab in Seoul was doing.
But we didn't, because federal rules stood in the way. Incredibly, these were federal rules that actually imposed higher burdens on these labs than they would have faced in ordinary times.
To understand how this was possible, you have to understand how these lab-developed tests are regulated. The labs allowed to perform them are regulated by CMS under a statute known as CLIA.
But, on top of that, FDA has suggested at times that it believes lab-developed tests can and should be regulated as devices under the Food, Drug, and Cosmetic Act, which requires device-makers to submit an application to FDA for premarket review before the device can be sold. In 2014, FDA issued two draft guidances, never finalized, that put forth a framework for requiring pre-market review of LDTs as devices. This idea had been floating around in the Bush administration, too, and as both General Counsel and Deputy Secretary, I had strongly resisted it. There appeared to be few safety and quality issues that would justify this sweeping assertion of federal authority, and many also believed there was no statutory authority for this regulatory grab.
And yet, the FDA had made this sweeping assertion in a post on its website, without any opportunity for public comment. Prior to COVID-19, FDA had decided to impose pre-market review and authorization requirements for lab-developed tests when there was a public health emergency that permitted the agency to issue emergency use authorizations, or EUAs.
The goal of allowing the FDA to issue EUAs is to speed products to market. But in this case, arguing that stricter scrutiny was needed for tests during a pandemic, the agency was actually using EUAs to slow the process down.
Worse, they were doing so in a non-transparent, non-accountable way. The FDA repeatedly assured HHS leadership that they were not asserting pre-market approval jurisdiction over LDTs, even as we later learned they were sending warning letters to academic medical centers and other labs who had developed their own LDTs in January and February. Indeed, we learned of the FDA's blog post asserting pre-market approval jurisdiction over LDTs only when a reporter pointed us to it in March, after FDA had repeatedly claimed to HHS leaders that it was not asserting this jurisdiction.
Achieving the right balance of speed and safety in these regulatory decisions is always a difficult task. But FDA was not transparent that they had actually raised the standards on these labs. Labs complained to us throughout February that they found the EUA criteria perplexing and challenging, which made a great deal more sense if it had been clear that FDA was imposing a wholly new and unprecedented requirement.
These labs were slowed down in part because they weren't accustomed to actually submitting EUAs or any kind of applications to FDA. Remember that the Obama Administration had backed down from its assertion of jurisdiction over LDTs, so these labs had never learned how to deal with FDA's device center and its requirements. They were told reams of data would be required, and even directed, in at least one case, that it be submitted in hard copy—at a time when every hour of every day counted.
The standards were hardly pro forma, too. The FDA, for instance, required five validations be run of a test, requiring five viral samples—at a time when, in early February, barely five cases of the virus had been confirmed in the United States.
When you combine FDA's lack of clarity about what they were asking of labs, and the independence traditionally accorded to FDA regulators, it took time to undo this morass. A partial resolution came on February 29: We agreed with FDA leadership that they would allow labs to start using tests they had developed, as long as they also submitted a completed EUA application within 15 days.
This was still a significant bureaucratic barrier to LDTs, but it was an improvement. It was not until mid-March that FDA finally stated that it would exercise enforcement discretion and not require EUA submission at all for COVID-19 LDTs that were developed under state authority.
I want to make one final point about the role of testing in the very early response. There is a myth out there that, if only we'd had a superior testing system, we simply could have caught any cases and isolated them. But given the retrospective analyses we've now seen, that would have been like finding a needle in a haystack. We did not have an outbreak like South Korea's.
At the time, CDC was using its systems that detect respiratory illnesses, which was the appropriate way to look for the virus given what we knew about it, and that data did not show widespread presence of the virus. The same goes for CMS claims data—no evidence of a significant rise in respiratory illnesses. No country had a specialized syndromic surveillance system back in January and February that could have more accurately picked up cases based on symptoms, and given that we now know that up to 50 percent of people are asymptomatic, any such symptomatic screening system would have missed half of the cases.
Although there is reason to believe the virus potentially arrived here in December and January, retrospective analysis from CDC has now confirmed that there was not large-scale community spread in the United States throughout February. A study out of Seattle retrospectively analyzed specimens from patients with respiratory disease from January 1 through February 20, and did not find one positive result in 5,270 samples.
With the premarket requirement from FDA out of the way, hundreds of labs could provide far more testing capacity as we learned more about the virus and the appropriate testing criteria, but they would not provide anything like the kind of capacity we would need for a nationwide pandemic.
That is why, back in January, our Office of the Assistant Secretary for Preparedness and Response had begun reaching out to diagnostics companies to encourage them to get into COVID-19 testing. Beginning in early February—when the number of confirmed cases in the U.S. was in single digits—we began allocating funding to support test companies in developing products. This kind of early investment was essential in particular because testing companies were gun-shy after the experience of some previous novel viruses, like SARS, MERS, and Zika, when some invested in testing development only to see the viruses not become the crises some worried they might be. These early investments have proved crucial, later on, to the authorization not just of commercial PCR tests but also of rapid and at-home tests. We would not have the capabilities we do today if ASPR and BARDA had not moved so quickly during those first two months.
In mid-March, a number of large-scale commercial testing companies received EUAs for their tests, which could be purchased by hospitals, pharmacies, and other entities that purchase hundreds of millions of tests from these companies each year. Once these tests started hitting the market, this next stage brought new challenges, and a role for the federal government in coordinating solutions. Unlike LDTs, these tests often come prepackaged, as a kit or even a cartridge to insert in a machine, and, as all of you know, there are many different such machines in the U.S.
Thus, when one hears that we don't have testing supplies, what they're often referring to is that the testing platform that they invested in lacks supplies. Providers are often reliant on very easy-to-use lab equipment that requires pre-made cartridges from the manufacturer and low skill in operating. If their brand isn't available, this means, in their mind, tests aren't available—even if millions of other types of tests and platforms are available down the street or through commercial vendors. In their mind, there is a testing shortage—but that's only because we have a segmented ecosystem.
On top of this segmented ecosystem, our largest testing companies were not even significant players in testing for respiratory viruses. They mostly did blood work, rather than work with nasopharyngeal swabs. You couldn't just walk into a LabCorp location to get a COVID-19 test like you might some routine bloodwork because they simply don't take swabs themselves. These companies could do incredibly high volume testing of blood samples, but not the swabs you needed for a respiratory virus. In the event of a pandemic flu, this might have been much less of a challenge because we could have very rapidly produced antigen tests, which doctors use at the point of care to test for the seasonal flu today.
Further, these lab companies rely on the same global supply chains that every other country's testing system relies on. There was soon a worldwide shortage of swabs, reagents, and other materials needed to collect samples and perform tests, and, like PPE, almost all of these were manufactured outside of the United States. There were also supply constraints: Production of many medical products in China had been slowed by the outbreak there, and the largest manufacturer of swabs in the world was in northern Italy—which was being hit incredibly hard by the virus. We also needed to ensure that there was access to these testing systems in vulnerable communities, which have been hit especially hard by COVID-19.
On each of these challenges, the federal government stepped in. In mid-March, I appointed Admiral Brett Giroir lead for the federal role in testing, giving him plenary authority over the FDA and CDC with regard to testing. Starting in April, the federal government directly purchased supplies from companies all over the world and then worked with states to fairly allocate swabs, collection media, and other crucial materials. In total, the federal government has distributed 182 million swabs and 225 million tubes of viral transport media.
We supported the validation of multiple different types of material and methods—like nasal swabbing instead of nasopharyngeal swabbing, thank heavens!—to dramatically increase the availability of supplies and spare desperately needed PPE. FDA continually provided more regulatory flexibility around things like storage media for samples, determining that, for instance, simple saline solution could be used to store samples. We worked with academic medical centers and private industry to validate pooled sampling, a method used around the world that could dramatically increase labs' throughput capacity, so that FDA would permit it.
To address the issue I mentioned earlier about mismatches between the tests providers were used to or capable of using and the tight supplies for these products, we held literally hundreds of hours of coordinating calls among private sector players, to help them identify testing capacity that wasn't in use and get them in contact with manufacturers that had the correct products.
We brought qualified, validated supplies in from wherever in the world they could be found, at several points using military or chartered aircraft to fly swabs from the one factory in northern Italy that made most of the world's supply. As of today, we have performed 41 Air Bridge flights—think of that, 41 cargo jets—just to bring in testing supplies like swabs and pipette tips. At the end of April, under the Defense Production Act and other authorities, we awarded several contracts to a company in Maine, Puritan, to double their swab production from 20 million to 40 million swabs a month. They are now increasing that to 100 million swabs per month. In total, we've used the DPA 13 separate times to speed production and allocation of supplies necessary for testing production, and invested literally billions of dollars into domestic infrastructure, research, and development.
Then there was the challenge of ensuring these tests could get to the people who needed them. In March, we pioneered a community-based testing site model that initially relied on federal personnel, including members of the U.S. Public Health Service, to perform sample collection and ensure tests could be performed in hotspots and vulnerable communities. Of course, there is no supply of federal medical personnel that could ever meet the testing needs of the whole country. That is why the community-based testing sites were a model that we scaled up with the assistance of the private sector. Today, we have more than 1,800 retail sites across the country in operation under this model, supported through a federal contract that pays for sample collection.
CVS, Walgreens, and other pharmacies have opened more than 5,000 testing locations made possible by this federal model and regulatory flexibilities we provided. We've also set up more than 600 surge testing sites over the past six months, bringing extra testing capacity into especially hard-hit areas, with 118 currently active in 12 states.
The federal government has also taken the lead on investing in and allocating the next generation of diagnostics: molecular and antigen tests that can provide rapid results. As soon as the first rapid test was authorized by FDA, the Abbott ID Now product, the federal government bought up 40 percent of the world's supply, and allocated it to those who needed it most and could use it best: state public health departments, nursing homes, ultra-critical infrastructure, the Indian Health Service, and other partners. It cannot be overstated what a triumph it is to have rapid test options this quickly: These tests typically take years and years to develop, and they really only exist for common viruses like the flu.
We have been allocating these rapid tests strategically since April, and continue to do so as more enter the market. With the next wave of rapid tests, which use instrument readers to automatically provide a result, we used the DPA to prioritize federal orders and sent them to every nursing home in America by mid-September. Finally, this fall, we took another leap forward with the distribution of 150 million BinaxNOW antigen tests, to states, long-term-care facilities, HBCUs, and Tribes. Again, we could make that strategic allocation because the federal government bought about the first five months of Abbott's manufacturing.
As I mentioned at the start, all of this now means that Americans now have broader access to rapid tests than any other large country on the planet. We have broader access to testing than most European countries.
Throughout this summer, when we were performing approximately 1 million tests per day, we had the capacity to do about 3 million per day—running at about one-third capacity. We now are performing nearly 2 million lab tests a day—plus tens or hundreds of thousands of point-of-care rests that are only reported in certain circumstances. This past month, for December, we had the clear capacity to perform over 180 million tests, not even counting pooling.
But, as much of a success as this is, there have been plenty of lessons learned—lessons that need to inform the next phases of our pandemic response and our preparation for future pandemics.
First, there is a clear role for the federal government—providing financial and scientific support, and coordinating products that are scarce on a national level. That's what we've done. There is a role for states and the private sector: setting more granular priorities about where testing is needed and what facilities and personnel exist to deliver it. The United States needs a federal testing strategy, which we have; it does not have and would never build a federal government testing system to serve the whole country.
There may, however, be a more prescriptive federal role about how testing is used. There are really three types of testing needed to deal with a virus outbreak: diagnostic testing, based on symptoms or other risk factors, to identify cases; screening, which can be used to test at risk groups, like nursing home staff to protect residents; and then surveillance testing, which is intended to be done methodologically, to track the presence of the virus within a population and determine whether mitigation measures are working.
We have often confused these concepts. While it is a natural aspiration for the American healthcare system that everyone who wants a test for any reason can get one—and that is essentially the case today—that is not the key priority for a public health response.
Essentially, high volumes of testing for people who don't have known exposure to the virus or known risk factors acts as a form of surveillance, but not an especially useful one.
Surveillance testing must be done in a deliberate way, and it can benefit greatly from rapid testing of asymptomatic people—an effort that has been hindered by FDA's refusal to authorize many testing options for asymptomatic people.
We have also encouraged state and local governments and other institutions, like universities and long-term-care facilities, to use wastewater testing, which provides an extremely low cost way to detect the virus in a population—and then bring in a coordinated testing regimen to address an outbreak found. In the coming months and in future pandemics, there may be a larger role for the federal government to prescribe these kinds of protocols, even before they have been fully demonstrated, and to support them actively.
At the same time, because of technological advances fueled by the federal government, we are now nearing the point where Americans will have access to over-the-counter tests that can provide a rapid result for them at home.
That was never going to be possible in the early days of the pandemic, because of the technological development that had to happen, but thanks to federal leadership and the remarkable work of private sector innovators, it is now around the corner.
Even as we began this work back in January and February of last year, we also learned that engaging the private sector in the extremely early days of the public health response may have been helpful as well.
It would have been better if we'd had many alternatives to the CDC test developed by labs around the country, but it also would have been better if CDC had not attempted to shoulder as much of the burden of designing and manufacturing its own test as it did.
Another lesson has been the mismatch between America's existing private sector testing system and the testing needs for a novel virus. I like to compare the way our testing system works with other countries as a Burger King vs. McDonald's situation, which will slightly date me. Burger King was once famous for "have it your way"; you could customize your burger exactly as you wanted. McDonald's, you just got your cheeseburger with the standard number of pickle slices, squirt of ketchup, and slivers of onion.
America's testing system is paradigmatically Burger King. Hospitals and other providers are used to certain types of tests and certain testing companies. When I was visiting a hospital in Iowa this summer, for instance, they mentioned that they were unable to get COVID-19 tests. I asked them where they'd been trying to buy them from: They named the one testing company they typically used. If you're only used to using one testing company, a crisis like this that causes major supply and demand dislocations will be an even greater challenge than it needs to be.
A final major lesson from testing is one we now know very well from Operation Warp Speed: the federal government can create a massive market by making large pre-purchase orders, and those can be very wise investments indeed. A crucial piece of preparing for the next pandemic will be ensuring that there is funding to drive the private sector to make these kinds of investments in testing early on.
Even with these challenges, it is remarkable what has been accomplished this year, by American scientists, innovators, and public health officials, to advance our testing regime. More good news is on the horizon: In the past month, after significant investment by NIH and technical assistance by our HHS teams, FDA authorized the first fully at-home tests, which can generate results within minutes. Through NIH's RADx initiative, we have invested nearly half a billion dollars in next-generation testing options—likely more public investment than any other country has put into development of advanced COVID-19 tests.
Even as we roll out safe and effective vaccines through Operation Warp Speed, continued growth of testing options and testing capacity will be essential. And preparing properly for the next pandemic will require much more deliberate coordination between public and private sectors about what might be necessary to scale up testing for a novel virus.
What we have done over this past year to develop the world's largest testing system is not just unprecedented, but a uniquely American achievement—something that we could do because we believed in enlisting every creative corner of society and bringing the public and private sectors together.
Because of the heroic efforts of Americans from every walk of life—from lab technicians who have been working overtime to manufacturing line workers and researchers putting in long hours to ordinary people making sacrifices to protect loved ones—we will get through this crisis. We could not get through it without all of you.
Thank you for what every one of you has done to support our response so far, and know that, because of your efforts, victory is in sight. Thank you.
