NEW YORK – Microbiology labs are frequently asked to do more with less, and the SARS-CoV-2 pandemic has provided even greater challenges for directors as they have tried to keep up with the pace of test volume, supply shortages, and a lack of enough trained personnel to conduct molecular testing.
During an online session this week of the virtually held Association for Molecular Pathology annual meeting, five lab directors stepped forward to share their experiences during the pandemic. The wide-ranging conversation touched on topics such as test allocation, staffing and reagent shortages, and the ethics of reporting Ct values for non-quantitative tests, offering a sort of oral history of the perseverance to outlast nearly 10 months in crisis mode with seemingly no end in sight.
Esther Babady, director of the clinical microbiology service at Memorial Sloan Kettering Cancer Center in New York City and moderator of the session, said that when it was initially put together, she presumed that by November it would be a retrospective, in-person discussion of how the pandemic was handled by labs. "Little did we know that we would still be really in the middle of this," she said.
The session participants were selected to provide insights into the experience of different size labs serving various types of communities, Babady said. But their experiences have been similar, with the directors describing the past several months as "stressful," "insanity," and "a total blur."
Beth Marlowe, scientific director and head of R&D for Quest Diagnostics' infectious diseases division, spoke about bringing on testing across a national reference lab network currently managing a quarter of the US testing volume, while Anthony Tran spoke about the public health lab experience from his position early in the pandemic as director at the District of Columbia public health lab.
Representing somewhat smaller lab settings of community and academic hospital labs were Teresa Karre and Michael Bachman. Karre serves as medical director of microbiology and serology at Nebraska Methodist Hospital and Children's Hospital and Medical Center in Omaha, while Bachman is associate director of the clinical microbiology lab at the University of Michigan and Michigan Medicine in Ann Arbor.
As the pandemic has progressed, these directors brought on a variety of molecular diagnostic tests for SARS-CoV-2 in an attempt to accommodate shortages and meet demand.
But lack of coordination in the US national response to the pandemic led to regulations being developed simultaneously with tests and strategies, initially baffling many of those to whom the evolving rules applied.
Tran's public health lab in DC was among the first to be alerted to a potential need for testing, back in mid-January, and he and the others in the lab quickly began preparing.
Marlowe at Quest said the team met around that same time, too. "We all saw it coming. A reference lab like Quest has a lot of history in terms of getting an EUA for Zika, and H1N1, and being ready," she said, adding, "Early on we were talking to [the US Centers for Disease Control and Prevention], they had all the big reference labs on a call to say can you help us with capacity if needed."
Despite extensive experience bringing on tests for emerging infectious diseases, Marlowe also recalled telling her spouse in mid-January, "This feels different."
In general, it seemed directors in smaller labs expected a certain pace and sequence of events, based on past experience, Bachman suggested.
"We initially felt like this would be similar to what had happened for Zika, where the CDC would be doing the testing, there might be EUA assays that would come out from large manufacturers, that would kind of flow relatively smoothly, and it would become clear when we were allowed to do testing and there would be something available," he said.
Bachman cited as a turning point Nancy Messonnier's saying during a CDC call in February that the disruption to everyday life might be severe. "It rapidly became clear that we needed to be able to do testing, and it wasn't clear how we were going to be able to do that based on where we were" with the US Food and Drug Administration, he said.
The pandemic flashed over parts of the US quickly, as federal regulations evolved and tried to keep up. In late February, the FDA began allowing high-complexity labs to develop testing, and on March 13 the Roche Cobas SARS-CoV-2 test became the first commercial assay to get EUA.
Quest's test went live on March 9, Marlowe said, delayed by validation requirements. "If you recall the template from the FDA, you needed like 30 positives and 30 negatives – who had those?," she said. Quest ultimately had to negotiate with the FDA to use extracts from patient samples supplied by one of the lab's global diagnostic network sites in South Korea.
Tran said his lab in DC started testing on March 3 and got positives right away. "It popped really quickly," he said, adding that diagnostic companies then began approaching the lab for access to those positive samples. "I knew we needed more companies to actually have a test, to not solely rely on CDC's test," he said.
However, the scale of the demand quickly became overwhelming for testing labs.
Karre's molecular test volume has now more than doubled, she said, requiring shifting of staff from the core lab to the molecular, while Bachman's lab saw a fifteenfold increase of testing on its Abbott m2000 platform alone, while it also runs testing on four other platforms. Tran described a peak increase in molecular diagnostic testing volume of nearly 30 times the lab's previous maximum test volumes during peak flu surveillance.
At Quest, the team scaled from 2,000 tests per day when it went live at its first site to 200,000 per day across its entire US network.
The lab directors all agreed that maintaining such high volumes while meeting rapid turn-around times, and still remaining accurate, was an ongoing challenge.
Shortages of extraction reagents and nasopharyngeal swabs cropped up almost immediately, and shortages in pipette tips and other plastic consumables and in viral transport media appear to have persisted despite claims from manufacturers that they would be short-lived.
Tran said that during the initial peak, lab directors were using listserves to facilitate bartering with each other, trading primers and probes for enzymes, for example. "It was pretty crazy," he said.
Labs also faced a staffing shortfall, since molecular diagnostic tests require skill and expertise to run well. "We are in the process of hiring, but it is a challenge finding qualified medical technologists to do molecular testing," Bachman said.
If that all wasn't enough, they sometimes found they lacked test systems to meet the demand, or, if they had the instruments they sometimes couldn't get the test kits or cartridges.
Babady highlighted that one word that has now grown to an outsized representation in the lab director's vocabulary: allocation. "I knew that word existed, but I never paid any attention to it, and then it became the word that I would hear every day," she said.
When Washington, DC said it would set up drive-through and walk-up clinics, Tran realized the lab needed an automated system. His lab was fortunate to be collaborating with Hologic as a clinical trial site for another assay prior to the pandemic, so it had staff trained to run a newly installed Panther Fusion system. When Hologic's EUA came through, "We were the first knocking on the door to say, 'Hey, we really need this,'" Tran said.
In Nebraska, Karre's lab was a bit later getting testing up and running, and had already been hearing about shortages for some time. The lab's all-hands-on-deck approach meant highly complex tests were not a good fit, and the team corralled its point-of-care instruments in the central lab.
Allocations were a problem for Karre as well. "When it really hit me was when we were told our first allocation of Cepheid cartridges was going to be only about 10 percent of what we needed for our total testing needs," she said. The lab then validated a lab-developed test on the Luminex Ares platform, and collected Roche Liat platforms from decentralized sites as soon as that test got EUA. But it still needs to send out more than half of its testing to a reference lab, said Karre.
Interestingly, the Nebraska lab submitted an EUA to FDA for its Luminex test. The pending test got caught in the recent kerfuffle over EUAs for lab-developed testing, although as of today, it appears the EUA pathway is now available to labs again.
Quest began running an LDT on March 9 at its facility in San Juan Capistrano, California and received an EUA for an assay on March 19. But not all of the labs in its system had the capabilities to run that test, so Marlowe said it also validated testing on systems from Roche and Hologic to build capacity.
"Every day is a day of, 'Which hole do I plug, and how do I make things work?,'" Marlowe said. For example, early on, even though its EUA test was an LDT, "We couldn't get extraction kits for the [Roche] MagNaPure," she said, adding that Quest has since found smaller reagent suppliers that it may not have considered before to be perfectly fine.
In the university health system setting, Bachman said his team is now using five RT-qPCR tests on different systems. He did not specify which ones during the session, but said they include lab-based and point of care molecular tests, as well as antigen tests. Importantly, the directors noted that running multiple tests with different performance characteristics also requires more nuanced results reporting.
Scaling the workflow processes for molecular diagnostic testing has also proved daunting.
For example, the need to accession hundreds, sometimes thousands, of NP swabs in tubes each day pushed Tran's lab to move from a paper-based system with faxed results to a web portal-based one mid-pandemic. Bachman said his lab's specimen processing is currently handled by volunteers pitching in from other departments.
Responding to a call for increasing test availability through the use of pooling strategies caused collateral strain on adjacent groups, like information technology and lab information management systems support teams. "It made our LIMS person's head almost explode," Tran said, adding, "There is a lot of work that has to go into that in order for it to be reported out appropriately."
In terms of pooling, Marlowe noted that Quest has an EUA claim for pools of four and the lab recently published its data. Tran said the DC team is standing by to use Hologic's pooling EUA, and Bachman is also moving toward pooling with his lab piloting pools of five sample per test on a high-throughput system.
But Karre noted that in her lab's case, allocation has prevented them from validating pooling. "Also, our positivity rate right now is over 30 percent, so I don't think we would gain any efficiencies if we were to do [pooling] at this point, unfortunately," she said.
And the surprises keep coming for lab directors. "If you had told me in March some of the things that we are doing now, I wouldn't have believed you," Babady said.
Although they may be used to improvising lab workflows, none could have imagined using homemade phosphate buffered saline as viral transport media, for example.
In addition, alternative sample types that were once shunned are now being embraced by some. Tran noted that self-collected nasal swabs are outside of standard pre-pandemic practice, but Quest now has a claim for these. Saliva-based molecular testing for a respiratory virus is unprecedented, although Babady noted her team found in a surveillance study that saliva samples were pretty stable, even at room temperature.
Tran's lab also grappled with non-standard swab options at one point after receiving a federal shipment labeled "US cotton." Cotton inhibits PCR, but the supply was received by multiple directors along with a letter that assured them that although the packaging said the swabs were made with cotton, they were actually synthetic.
The potential for flu season to soon impact already skyrocketing testing volumes is also daunting, but the lab directors are finding ways to manage.
"The allocation is the big challenge," Bachman said. "We sat down and said, 'What did COVID look like per week in the Spring, and what was our worst flu season — you add those together and those numbers look pretty staggering," he said. "We are still trying to figure out how to make that work with the allocations we've got."
The lab directors also noted that their work is in the spotlight in ways it has never been before, and this is not always helpful.
For example, although it may be exciting and fascinating that the average American now knows the difference between a PCR test and an antibody test, patients and clinicians do not have a deep understanding how prevalence relates to positive predictive values, and explainers in the popular press have inadvertently sewed confusion.
Perhaps because of the public's crash course in PCR, the lab directors reported facing more demands to report cycle threshold data.
Some of the directors indicated that in rare cases they might consider providing that number to the infection prevention specialists, but without any interpretation — for example, a hypothetical patient about to go into surgery with a SARS-CoV-2 test Ct value of 42, or an elderly patient who had previously tested positive but needed a negative test to return to a skilled care setting, and the retest had a very high Ct values.
"I'm hesitant to provide that information routinely, and I'm certainly not willing to report it as part of the results, because these aren't intended to be quantitative tests, we haven't validated them in that way, and the instructions for use don't say that you can do that," Karre said.
Marlowe, also declines to provide Ct values. "People don't understand real-time PCR," she said. "What's amazing about this pandemic is suddenly everyone is an expert in infectious disease molecular diagnostics, including Elon Musk on Twitter," she said.
That said, the directors all seemed surprised find themselves open to the concept of a quantitative nasopharyngeal swab test for a respiratory pathogen, provided it would have a good internal control. Quantitative molecular diagnostic tests are typically more common for blood samples, in part because it is easier to standardize phlebotomy sample collection, but with a validated quantitative NP test, they could report Ct values.
As testing volumes continue to rise, these lab directors will continue to evaluate their responses. "We are moving so fast," Marlowe said. "If we had a little time to think through of some of this, we could maybe be a little smarter about how we are doing it, but we are just moving so fast to either make up for the allocations or try to get through the volumes, it's just really challenging every day," she said.
Babady concurred. "It's like whack-a-mole," she said.