Read our COVID-19 research and news.

A team from Seattle's Harborview Medical Center carry supplies in preparation for home visits to collect samples from people potentially exposed to the new coronavirus.

David Ryder/Reuters

‘We're behind the curve’: U.S. hospitals confront the challenges of large-scale coronavirus testing

China and South Korea continue to report drops in their cases of the respiratory disease COVID-19, and there’s an increasing realization that, as the World Health Organization (WHO) has insisted for several weeks, countries that aggressively confront their outbreaks of the coronavirus causing it can make powerful inroads to slow the spread. Again and again, the key to success has been widescale testing for the virus’ RNA in people and then isolation—either by choice or mandate—of the infected and their contacts.

In the United States, in contrast, COVID-19 cases are climbing at an alarming rate, and people who have received tests for the virus, SARS-CoV-2, remain few and far between. (No official number exists, but fewer than 5000 people in the United States had received a test as of 9 March, according to an Atlantic article that attempted to tally the total.) The White House coronavirus task force has insisted that early problems created by a faulty test distributed by the Centers for Disease Control and Prevention (CDC) have been overcome and that the U.S. government has increased production of functioning diagnostic kits. There has also been a push to move testing away from CDC itself and state labs to hospitals and commercial companies. The U.S. Food and Drug Administration (FDA) on 29 February changed its regulations to allow diagnostic labs that previously have met federal quality standards to modify the CDC protocol or design their own kits.

At a White House press conference on Tuesday afternoon, Vice President Mike Pence, who heads the country’s coronavirus task force, said that by the end of the week, there will be more than 5 million tests available and that industry is ramping up production of even more. Pence stressed that the government has also removed criteria that initially restricted testing to people who had traveled to China, come in contact with a confirmed case, or had severe symptoms. “As the president said, anyone who on a doctor's order wants to be tested, can, at a doctor's indication, be tested now,” Pence said.

Not so, says physician and epidemiologist Michael Mina, who helps run the diagnostic lab at Brigham and Women’s Hospital in Boston and has been critical of the Trump administration in blistering tweets. “The reality is most people will not be able to get a test this week, and most people will not be able to get a test next week,” Mina says. He also anticipates that there will be—or may already be—a shortage of reagents needed to run the test kits. He’s not alone. “We are deeply concerned that as the number of tests increases dramatically over the coming weeks, clinical labs will be unable to deploy them without these critical components,” the American Society for Microbiology said in a statement it issued yesterday. CDC Director Robert Redfield echoed that concern yesterday in an interview with Politico, and the agency on 9 March revised its COVID-19 guidelines to require testing only one patient sample, rather than two, which will cut the reagent needs in half.

Linoj Samuel, who heads the clinical microbiology division that oversees lab testing for the Henry Ford Health System in Detroit, spoke with Science about the situation he’s facing. Henry Ford has six hospitals and many medical centers that cover most of southeast Michigan. The state has yet to have a confirmed COVID-19 case. This interview has been edited for clarity and brevity.

Q: What’s the situation now in your system?

A: We're currently using the state health department, which is in Lansing, about an hour and a half from here. They went live with testing last week. Initially, there were very restricted criteria based on travel and other issues to decide who can get tested. So we didn't really test that many patients. But since CDC relaxed their criteria, that has changed, and so in the last day or two, we've been testing more patients. We are in the process of acquiring reagents for our own in-house testing. We've ordered them from the CDC-designated vendor and we should be getting them in the next day or two. But that doesn't mean we can just turn the switch on and start testing. We'll still need to do a fairly extensive validation based on the FDA requirements for emergency use authorization.

Q: Isn’t that simply sending five positive and five negative samples to the state lab for verification that your test works?

A: No, that’s only if you follow the CDC protocol to the letter. The CDC protocol requires specific instrumentation that not every hospital has. We don't happen to use that specific instrumentation. So in our case, FDA requires that you test at least 30 specimens spiked with the virus and you have to establish a limit of detection, and then you have to show that there’s no cross-reactivity [to genetic material unrelated to SARS-Cov2]. That creates a larger set of hurdles to clear. This may take a week, if not more.

Q: Do you know how many samples you sent to the state lab prior to them relaxing the criteria?

A: Not many. It was probably in the single digits.

Q: Do you know how many you've sent since they've relaxed it?

A: That just happened in the last 2 days. So we've probably sent maybe four or five samples.

Q: You’re a major healthcare system, and you haven't been able to do widespread testing?

A: No, and I don't believe the state has either. They are moving in that direction.

Q: Is it your hunch that there are no cases in the state or that there are cases and that they have not been detected?

A: It's likely there are cases in the community and we're just starting to test widely enough to pick them up. As of the end of last week, I wasn't aware of any hospital in Michigan that had the test up and running. The only place that was actually testing was the health department.

Q: So what's your reaction when the president and the secretary of Health and Human Services and CDC repeatedly say that anyone who wants to test can get a test?

A: It’s clear that they have been shipping out a lot of test kits. We expect to get them tomorrow. I don't think that's really the issue anymore. I think the challenge for us is doing all the other stuff that's required as part of this process. We could have been working on this earlier, if FDA had relaxed the requirements earlier. But in the past, the process was very rigorous, and really too much of an undertaking for your average clinical laboratory. So we've stayed away from that kind of a process. And even now, it's no cakewalk—it's a fairly rigorous process. It's just not simply plug and play. We can’t just turn them on and start testing.

Q: Couldn’t you also go to a commercial lab to test your samples?

A: When you go to the reference laboratories, you're still looking at a 3 to 4 or maybe even a 5 day turnaround time, when all is said and done. And that's just too long of a timeline. We can typically get results in about 24 to 48 hours.

Q: What's the turnaround time when you have a sample on hand? How long does it take to extract RNA from the sample, put it into a polymerase chain reaction machine, etc.?

A: About 2 to 3 hours to do the testing, but because the process requires us to batch samples, you can't really just test the sample as they come in. In order to conserve reagents and labor, you' wait until you have a certain bolus of samples and then test them together. So you do testing once or twice a day.

Q: Do you use 96-plate wells? [These are plates that have 96 miniatures wells and liquid from a patient’s sample can go in each one.]

A: Yeah, but each patient will require multiple wells on that plate. So it's not like you can test 96 patients.

Q: Do you have a concern about the supply of reagents?

A: The advantage of using CDC recommended testing kits and protocol is that you don't have to submit an FDA application. So, there may be a shortage of reagents for those specific platforms. As it stands, however, we'll be using an alternative platform. I don't think there'll be a shortage with that one.

I am concerned that if we had a large scale outbreak whether your average clinical lab will be able to handle that kind of testing volume. In the current influenza season, for example, we could easily test at the peak of the season 3000 samples a week But we're not going to be able to turn out those kind of numbers with this [COVID-19] test because it's lab developed batch testing with lower throughput.

Q: Is the United States going to miss out on this window of opportunity to stop the virus before it becomes widespread?

A: We're behind the curve. I don't think we've missed a window of opportunity yet. While it may be in the community here, it likely isn't there in extremely high numbers. It's unfortunate that we're in the situation we are, and part of that is because we didn't test early on because of restrictions placed on our ability to test. But I think that there's still a chance if we move aggressively to catch up.

Q: What are you hearing from colleagues in testing labs in different states?

A: It's the same level of frustration. We need to be moving quickly, and we understand the need for oversight. But the last thing we needed was a restrictive process, and there was really a lack of clarity. We've validated lab tests for decades, and time and time again we've made them work. This is not new to us. We need to be allowed to do this without this extra regulation. Moving forward, we need to have a more robust strategy that can move quickly in these settings. If this has been addressed earlier, we could have had a head start.