NEW YORK – A newly published multisite clinical trial has shown that the Roche Cobas Liat SARS-CoV-2 assay has accuracy comparable to laboratory-based RT-PCR tests. The 20-minute point-of-care test had nearly 99 percent agreement with testing in the lab on the Roche 6800 and 8800 instruments.
The study published recently in the Journal of Clinical Microbiology included clinical evaluations at five sites — Hennepin County Medical Center in Minneapolis; Weill Cornell Medicine in New York City; Thomas Jefferson University in Philadelphia; the University of Chicago; and the University of California at Davis. Researchers at Roche Molecular Systems in Pleasanton, California also participated.
The researchers found that the positive percent agreement for SARS-CoV-2 between the Liat and 6800/8800 systems was 100 percent, negative percent agreement was approximately 97 percent, and overall agreement was nearly 99 percent.
Uptake of rapid point-of-care molecular diagnostics has historically been impacted by issues of cost and throughput, tensions between ease of use and the possibility for contamination when deployed by non-experts, and a potential tradeoff between speed and accuracy compared to laboratory testing.
But the diagnostics needs during the global COVID-19 pandemic may have tweaked the balance somewhat.
Glen Hansen, medical director of the microbiology and molecular diagnostics laboratory at Hennepin County Medical Center and first author on the JCM study, acknowledged in an email that the high accuracy of the point-of-care test may come as a surprise to some people.
Nevertheless, five different labs from across the county were able to show that the Liat SARS-CoV-2 test "matches performance of what centralized labs can offer," he said, but on a platform "that can move closer to where it's needed."
Hansen had already been aware of the high accuracy of another Liat test. He and his team had previously shown that the Cobas Liat influenza A + B assay used in an emergency department (ED) setting could match the performance of centralized lab testing, with sensitivity and specificity of approximately 99 percent compared to the lab's standard of care, the GenMark E-sensor Respiratory Viral Panel.
That evaluation also demonstrated a rapid flu diagnosis changed patient management in just over 60 percent of cases, saving about $200 per ED visit.
In the case of the Roche Cobas Liat SARS-CoV-2 assay, a test that can be done at the point of care with results provided in 20 minutes could be particularly useful in the ED for more rapid and accurate patient triage to avoid mistakenly putting uninfected people into a COVID ward, or placing infected patients with the general population.
Other PCR-based rapid molecular tests for the virus have also shown high accuracy.
An evaluation of the Cepheid GeneXpert SARS-CoV-2 test at seven international sites showed near perfect alignment with standard of care lab-based testing. In this case, the comparator assays included lab-developed and reference lab testing, commercial kits from Seegene, Altona Diagnostics, and molecular assays on systems from EliTech, Abbott, and DiaSorin. The study authors noted that the best use of the technology is likely to be in acute-care hospitals in high-prevalence settings, "where rapid triage decisions need to be made regarding patient disposition and isolation."
Hennepin's Hansen said he has become a big proponent of moving diagnostic testing closer to patients.
In a Clinical Chemistry review in January, he argued that point-of-care molecular tests — with oversight from the clinical microbiology lab — could improve patient outcomes, and he defined the parameters that these outcomes studies should address.
The high accuracy of the Liat SARS-CoV-2 test meets one of the goals of point-of-care testing for infectious diseases, Hansen said, which is "to set the bar high and advocate for assays that match the performance of centralized lab testing."
Minimizing concerns over the performance of a point-of-care test allows labs to focus, instead, on the delivery of care, and "Highly sensitive assays provide us this opportunity," he said.
Hansen and his team have been running the Roche Liat assay for SARS-CoV-2 and influenza in the ED for about a month now, he said. "We are happy to report that our publication isn't just academic."
The Henry Ford Health System in Detroit, Michigan, has also brought on the Liat SARS-CoV-2 test across its sites. When Linoj Samuel, division head of clinical microbiology, and his team at HFHS initially brought on Liat flu testing a few years ago, they had to overcame some initial challenges. As with the molecular flu assay, the lab continues to limit the Liat SARS-CoV-2 testing so that the systems are operated only by medical technologists in "stat labs" placed throughout the health system, he said.
Still, Samuel said the health system is running Liat SARS-CoV-2 tests on hundreds of samples each day, focused on the ED primarily. "We have used the Liat for years and it has performed well for us," he said, adding, "We expect the same with the SARS-COV2 assay based on our initial studies."
Additionally, Samuel said that he anticipates the Liat combination flu/COVID test will be "quite helpful," although there isn't yet flu in the community and HFHS is not routinely testing for flu yet.
He said that the performance in the JCM study is "definitely a massive improvement" over point-of-care antigen tests and isothermal molecular assays, such as Abbott's ID Now test.
"We feel that this assay offers a much more reliable answer than the antigen test and allows providers to make patient care decisions with confidence, as compared to antigen assays where you do not have complete confidence in either positive or negative test results," Samuel said.
Nevertheless, high accuracy has not necessarily led to immediate gains in point-of-care tests in other settings. An evaluation earlier this year showed that molecular influenza testing in pediatrics had high sensitivity and specificity, but a number of lab directors noted that they were still wary of the potential for contamination in point-of-care molecular assays, and were put off by the capital investment and consumable costs.
Overall, the pandemic "has accelerated the landscape" of point-of-care testing, Hansen said, and post-COVID-19, the space will be likely be changed as labs are now being forced to look at such tests differently.