The FDA has published data outlining the limits of 58 different molecular coronavirus tests—illustrating the ability of each authorized diagnostic to detect smaller and smaller amounts of the virus that causes COVID-19.
The data can be used to compare which manufacturer’s test is more sensitive than another’s, with each being evaluated under the same conditions using a single set of controlled reference samples distributed by the agency earlier this year.
The FDA listed each diagnostic’s limit of detection, a statistic describing the lowest amount of virus particles present within a sample that could still provide a correct test result, measured in the number of individual units of RNA detectable by nucleic acid amplification, per milliliter.
A test’s limit of detection differs from what is known as its analytical sensitivity—or the diagnostic’s ability to spot truly positive samples—which is in turn used to determine its overall rate of false-negative results.
RELATED: FDA publishes validation data from 5 more COVID-19 antibody tests, both authorized and not
According to the FDA, the limit of detection alone does not paint the full picture of how sensitive a particular test can be, and is not used by itself to determine whether a test receives an Emergency Use Authorization or a regulatory clearance.
“Instead, the data gives laboratories, healthcare providers and patients a new resource on the relative performance of available tests to better inform which tests they choose to use,” the agency said in its announcement.
RELATED: Trump administration revokes FDA authority over lab-developed tests, including some COVID-19 diagnostics
Topping the list, with a limit of detection of 180, was PerkinElmer’s coronavirus detection kit, using nasopharyngeal swabs in viral transport media.
Close behind—by about a factor of three, in the 540 to 600 range—were tests from ScienCell Research Laboratories, BioCore, DiaCarta, Seasun Biomaterials and Hologic.
PCR tests developed by BD, Quest Diagnostics, Helix and Roche all posted detection limits of 1,800—while Abbott’s RealTime assay and Cepheid’s Xpert Xpress rapid point-of-care test were set at 5,400.
The original diagnostic panel supplied by the Centers for Disease Control and Prevention, during the earliest stages of the pandemic, was listed at 18,000. Tests by Diatherix Eurofins and Luminex—as well as Qiagen’s respiratory disease panel, which checks for COVID-19 plus a number of influenza strains, rhinovirus and respiratory syncytial virus—rounded out the list at 180,000.
Rapid, point-of-care screening tests using dry collection swabs and no viral transport media were listed separately by the agency. They include Abbott’s ID NOW test, with a limit of detection of 300,000, and Quidel’s Lyra Direct assay at 540,000. The FDA also posted data from Fluidigm’s Advanta saliva-based test, which hit 54,000.
The FDA began supplying its reference panel of testing samples to developers in May, and withheld the amount of viral material in each to provide a blinded evaluation. The agency said it has delivered the panels to more than 150 test makers.
“Reference panels can be used in many ways to support test development and authorization, but most importantly, they are a powerful tool in monitoring test performance and ensuring that Americans have access to diagnostics they can trust,” said the agency’s device center director, Jeff Shuren.
The full list is available here, and will be updated as the FDA receives additional results.