Why Is COVID Testing so Hard to Scale?
Two full years into the pandemic, it’s almost impossible to believe how badly the U.S. has botched its approach to testing. Rapid antigen tests, while never widely available, have become nearly impossible to find—and if you do score one, you can’t be sure the results are accurate. For gold-standard PCR tests, people wait in hours-long lines, and then wait much longer for results—often a week or more.
As the Omicron variant rips across our country, outrage about testing has followed closely in its wake. America’s clinical laboratories are among the best in the world. So how have we failed this testing challenge so spectacularly?
The heart of the answer is policy: neither the Trump nor Biden administrations prioritized testing, and neither invested enough funding to build the massive infrastructure that would have been needed to absorb such enormous swings in demand. To make a real difference, significant funds would have had to flow, not only to clinical labs, but also to instrument and test manufacturers and developers of the chemicals, tubes, and other elements needed to run the tests.
But even if testing had been given greater priority in the last two years, certain realities of how clinical laboratories operate would still have made it very hard to scale quickly enough to handle a situation like the one we’re in now with Omicron.
There’s a reason clinical labs handle routine testing so well: it’s predictable. Lab managers know how many test kits they need for any given period, and their instrumentation is set up to handle the right volume. Testing is always more difficult for labs to manage when demand ebbs and flows.
Flu testing is a great example. Many clinical lab managers need to place their orders each year for test kits, chemicals, and other consumables by late summer, in order to prepare for the winter surge. During the 2017-2018 flu season, when flu rates in some areas were three times higher than usual, labs were brought to their knees. They needed more tests and more consumables than anyone predicted. But getting these scarce supplies, which must be produced in highly regulated manufacturing facilities, pitted labs against each other.
Sound familiar? A three-fold surge in demand crippled labs then; in the past six months or so of COVID testing, we’ve gone from a pre-Delta daily low of fewer than 200,000 tests to a recent Omicron one-day high of 4.2 million. That’s a 20-fold difference, and it may get worse. Expecting labs to handle this with existing infrastructure is like expecting a two-lane country road to handle the volume of rush-hour traffic in LA.
While test kits and consumables are part of the problem, the main factor that has delayed big labs’ ability to respond comes from their need to add new instrumentation. The only way to significantly increase a lab’s capacity is to add new equipment for sample processing and PCR testing. During normal times, these sophisticated scientific machines can take weeks or months to be ordered and delivered, and the installation process can be slow as well. During the pandemic, with labs around the world vying for the same instruments at the same time, the delays got even worse.
Even when needed equipment arrives, it doesn’t immediately increase a lab’s capacity. “Plug and play” is a concept that does not exist in clinical labs. Before a new instrument or test can be used to generate results for patients, it must go through a rigorous calibration and validation process to ensure that it is consistently producing accurate data. Once the instrument or test is validated, in-depth training is required for each member of the lab who will have to use it. Together, these processes can take weeks. Validation for new instruments and new tests is required by the strict laws that govern how clinical labs in the U.S. can operate.
With the crush of the pandemic upon us, it’s tempting to think we could just skip these validation and training requirements in the interest of ramping up COVID testing capacity faster. But that would be a huge mistake. Getting accurate test results can literally be a matter of life or death. This is precisely why such rigorous validation practices were put in place, and laboratories must continue to comply with them.
Accuracy vs. Speed
As almost everyone knows by now, there are two common types of diagnostic tests for COVID: the so-called “rapid antigen” tests you can take at home, and polymerase chain reaction (PCR) tests that must be performed in a lab.
Unfortunately, the tests we need are not the tests we have. For virtually any application, rapid tests are less accurate than PCR tests. The rapid tests developed for COVID were not bad; their sensitivity and specificity actually turned out better than expected. But they were designed and evaluated before vaccinations became widely available. The thresholds set then to determine positive or negative results were not able to take into account the lower levels of virus now often seen in COVID-positive patients who have been vaccinated. Antigen tests require higher levels of virus to report a positive result. Even in unvaccinated people, this means that in the early and later stages of infection when viral loads are low, the tests can come back negative and still be functioning as intended.
PCR tests, meanwhile, are highly accurate. They’re based on a process hijacked from nature that allows scientists to make many, many copies of the genetic material in a sample. Even if the original sample contains extremely low levels of virus, this amplification technique makes it possible to return a positive result.
But PCR is a sophisticated science that can’t be rushed. The process involves a number of heating and cooling cycles that are often performed dozens of times over a few hours. Some labs use self-contained systems that handle the entire PCR method, but many opt for less automated approaches that require technicians to move samples from one instrument to another, in a more tedious workflow. (For a look at a major genomics lab that transformed itself into a high-capacity COVID testing facility, check out this short video.)
Whatever the method used, capacity is tough to change: systems may run four samples or 384 samples at a time, but the time to results cannot be accelerated. In times of extremely high demand for testing—such as during the holidays or when kids are returning to school after a break—PCR test labs take unacceptably long times to return results because there’s simply no other way to get through the backlog of samples submitted.
Unfortunately, there is no great solution to this problem. We will continue to see peaks and valleys in COVID testing demand. If we had built enough capacity to return results for 5 million tests in less than 24 hours, most of that infrastructure would have been sitting idle for most of the pandemic, and there would probably have been an outcry that the funding could have been better used. What we really need is a leaps-and-bounds improvement that would merge PCR accuracy and rapid test speed, delivered in an at-home test made with broadly accessible materials. We’re not there yet, but scientists around the world are working hard to make that kind of breakthrough happen.