How can innovation and regulatory policy accomplish robust COVID-19 testing?

By Lisa Larrimore Ouellette, Nicholson Price, Rachel Sachs, and Jacob Sherkow

It’s now clear that expansive, population-wide testing is part-and-parcel of every successful COVID-19 containment strategy. But US testing efforts, from the beginning of the pandemic until now, have been widely criticized as lacking. Perhaps as a direct consequence of this failure, the US now leads the world in COVID-19 cases and deaths. What are these tests and what’s our capacity to test; why is it important to test; how have the FDA and other administrative agencies addressed the issue; and what can we do about it?

What is the status of US testing capacity?

It is important to distinguish between two types of COVID-19 tests: reverse transcription polymerase chain reaction (RT-PCR) tests for SARS-CoV-2, the virus that causes COVID-19; and serological tests for the body’s immune response to SARS-CoV-2. The tests are not interchangeable: RT-PCR tests detect the presence of the virus’s genome, itself, and thus determine whether someone is currently infected. Someone who was once infected and has since recovered will return a negative result. A serological test, by contrast, detects whether the body has produced antibodies to the virus; that’s useful to determine whether someone has been infected for long enough to mount an immune response.

To date, virtually all of the testing has been of the RT-PCR type, useful for answering the question: Is the patient infected now? Testing centers in the US are currently running approximately 135,000 tests a day—far fewer per capita than in other countries. The US’s maximum, overall testing capacity is unclear and is, in any event, a moving target given that new tests are now being cleared by the FDA with some frequency. But it’s widely acknowledged that testing is not at the level that it needs to be to accurately assess the number of people infected with SARS-CoV-2.

There are myriad reasons for this deficit in testing: an initially slow ramp-up of tests approved by the FDA; difficulties in speeding manufacturing of kits used to conduct the tests; a shortage of reagents to conduct the tests, including solutions, primers, and even the swabs used to collect samples from patients; the capacity of clinical laboratories to run tests and return results; and less technical hang-ups like patients’ difficulties in finding or physical getting to testing sites and questions concerning who will pay for such testing.

Why is it important to have robust testing for COVID-19?

The importance of robust RT-PCR testing for COVID-19 is more—much more—than simply returning a diagnosis to an individual patient. First, it’s useful in assessing whether and how to provide care to tested individuals: young patients with little medical history and mild symptoms—even if they are able to obtain a test and test positive—are being instructed to essentially let the virus work its course and self-isolate at home. This is in stark contrast to older, medically fraught patients who test positive, many of whom are admitted to the hospital for continuous observation. Second, identifying positive hotspots is critical in assessing where important medical resources should be deployed. With limited supplies of medical staff and equipment, properly directing those to areas of contagion in real-time is critical. Third, robust testing can be used in containing existing outbreaks and ascertaining where new outbreaks are likely to spread. Groups of positive patients in a tight cluster—such as nursing homes or tightly knit religious communities—are much more likely to ignite a forest fire of the illness than individuals who are more isolated from their peers.

Beyond RT-PCR tests, robust serological tests will eventually also be important in containing the virus’s spread. Understanding the number of individuals who have recovered from infection will further an understanding of the population’s broader immunity to the disease and may shed light on why some patients are totally asymptomatic while others succumb to the illness. Calculating the remaining infectious capacity of a population can also help us begin to assess when things can “return to normal”—a hot-button political issue that should be informed by serological data. Using serological tests to assess the virus’s spread will also be important for future epidemiological studies of coronaviruses.

How has the FDA impacted US testing capacity?

New in vitro diagnostic tests—like those for COVID-19—are regulated by the FDA under the broad category of “medical devices,” with the level of scrutiny depending on the riskiness of the test. As Erika Lietzan has helpfully summarized, to circumvent the lengthy FDA review process in a public health emergency like this pandemic, the agency can use Emergency Use Authorizations (EUA) on a product-by-product basis. The aim of EUAs is to make sure that the tests actually work, but to accomplish that review as quickly as possible. EUAs have been used successfully for past disease outbreaks like Zika and MERS-CoV. The FDA has issued EUAs for 33 in vitro diagnostic tests for COVID-19. Almost all are for the PCR tests to identify viral genetic material, but one serological test for antibodies has received an EUA.

Health and Human Services Secretary Alex Azar declared a public health emergency on January 31 and declared that “circumstances exist justifying” EUAs on February 4. Although intended to ease regulatory hurdles for bringing COVID-19 testing online, the FDA’s EUA guidance ended up slowing the process down. In general, the FDA doesn’t enforce its regulatory requirements for so-called “laboratory developed tests” (LDTs) that are developed and run in a single facility (such as a hospital that develops, makes, and then uses its own diagnostic test). However, when the FDA moved to an EUA regime for COVID-19, the agency announced that LDTs would not be permitted for COVID-19 tests; the only way to get approved was to request an EUA. At the end of February, the FDA modified its stance, allowing some laboratories to keep testing using their LDTs while they waited for their EUA to be approved. But as the New York Times recently detailed, this led to a “lost month” when “new tests sat unused”—including at Stanford, which had already used WHO protocols to deliver more than 250,000 tests around the world but was unable to use them in its own backyard.

How did other federal agencies play a part in the testing rollout?

The FDA is not the only federal agency to be involved in the regulation of testing and the delayed rollout of COVID-19 tests. Extensive media reports have focused on the role of the Centers for Disease Control (CDC), which initially provided test kits to state labs around the country. However, the CDC soon discovered a flaw in its tests and told those labs to stop testing. The CDC was unable to correct its mistake for weeks.

Fewer news accounts have illustrated the role of the Centers for Medicare and Medicaid Services (CMS), which is responsible for regulating all laboratory testing in the United States through its oversight of the Clinical Laboratory Improvement Amendments (CLIA). CLIA imposes a wide range of requirements on laboratories, ranging from the laboratory level (including the physical facilities available and the credentials of the laboratory employees) to the level of each diagnostic test (requiring each type of test run by a laboratory to meet substantive proficiency testing standards). Many research laboratories that had the technical ability to run COVID-19 tests were not permitted to do so because they lacked CLIA certificates, and many have found it challenging to either obtain these certificates or partner with other labs to boost capacity.

CLIA’s existence may help explain why New York has raced ahead of other states in terms of its testing capacity. New York has long had its own extensive clinical laboratory certification program. As a result, New York is one of just two states to be CLIA exempt, meaning that its testing requirements are equal to or more stringent than CLIA’s own requirements. New York state labs therefore only need approval from New York, not from CMS, in order to begin COVID-19 testing. (The labs did need FDA approval after the EUA issued, but the FDA decided to exercise enforcement discretion as to such labs.)

In some ways, the rollout of COVID-19 testing was therefore a classic interagency coordination problem. Each of these agencies—FDA, CDC, and CMS—had a role to play in standing up a robust nationwide system of testing. But they failed to coordinate their responsibilities and cost the US weeks in potential response time. In many ways, the agencies’ priorities were even in open conflict, such as in February and March, when the FDA required the CDC to spend time retesting all positive tests from public health labs, further using up scarce testing resources.

This coordination failure cannot be laid entirely at the feet of these agencies, although they undoubtedly have many questions to answer (particularly the CDC). Importantly, these three agencies are all within the control of the Department of Health and Human Services. It was HHS Secretary Alex Azar’s role to mediate disputes between them and identify where the agencies’ policies might have conflicted and prevented laboratories from developing their own tests. Media reports suggest that he largely failed to do that, perhaps in light of his political incentives and constraints. There are also other indications that the White House may not have understood the relevant legal levers that would be involved here: FDA Commissioner Stephen Hahn and CMS Administrator Seema Verma were not added to the COVID-19 task force until March, well after the testing failures were well known, even though the task force was formed in January.

What steps could be taken to increase the supply of available tests?

Emerging safely from this crisis will require a dramatic expansion in testing. To increase testing rates, policymakers should reduce the kinds of regulatory barriers described above so that new facilities can be brought online as quickly as possible while maintaining public health standards. And they also need to create much stronger incentives to improve existing tests and expand testing capacity.

The diagnostic testing incentive problem is different from that for COVID-19 vaccines and treatments. The cost of developing a vaccine or drug is enormous—including expensive clinical trials and manufacturing facilities—and private firms routinely drop candidates from their pipelines that lack sufficient patent protection. Promising sufficient profits to speed effective vaccines and drugs to market should thus be a first-order policy priority. In contrast, diagnostics are comparatively easy to develop and cheap to bring to market, as illustrated by how quickly COVID-19 tests became available.

Diagnostics are difficult to patent in the US, and some commentators have argued that the lack of patent protection has chilled entry into the COVID-19 testing market. We agree that entering this market needs to be more profitable—but that doesn’t mean that the tests need to be patentable. Instead, policymakers can turn to other tools in the innovation policy toolkit. In addition to directly funding research on test improvements and issues like the connection between antibodies and immunity, governments could offer market-based “challenge prizes” for cheaper, faster, and better tests. The simplest way to implement such a policy in the US is probably through increased government purchasing: CMS should pay enough per test to make it worth companies’ while to scale up their testing.

We do not mean to understate the scale of this challenge. Vox’s Umair Irfan argues that scaling up testing sufficiently “would require a national mobilization on the scale of a world war,” including command-and-control coordination of the private sector, universities, and different levels of government and use of the Defense Production Act to compel firms to manufacture necessary equipment. Although the federal government’s repeated missteps in managing the pandemic so far make it difficult to imagine this level of coordination, the emergence of strong regional alliances of states, including states with strong scientific and innovation infrastructure, suggests that such planning may be forthcoming.