The speed of Abbott's molecular point-of-care Covid-19 test ID NOW is a major advantage over many available molecular assays, but payers will now have to weigh this up against the potential need to double-check its results.
This is because the US FDA yesterday suggested that ID NOW could return inaccurate results – false negatives in particular. Abbott has come out fighting, saying it would optimise the test, and stressed that negative results with ID NOW should indeed be presumed negative. But it conceded that some patients might still need to be re-tested with another viral RNA assay.
ID NOW is faster than any other marketed test that detects the novel coronavirus’s RNA, taking around five minutes to return a result. But Abbott now says that that if a negative result from this test is inconsistent with the patient’s clinical signs and symptoms, or a definitive answer is necessary for patient management, the patient should be tested again with a different molecular assay.
The FDA was actually quite conciliatory, stressing that the ID NOW test could "still be used and can correctly identify many positive cases in minutes”, and saying that it was still evaluating information about inaccurate results and was in communication with Abbott. But the agency has received 15 adverse event reports about the device that suggest some users are getting inaccurate negative results.
The agency’s statement followed the publication, ahead of peer review, of research conducted at New York University School of Medicine comparing the test with Cepheid’s Xpert Xpress molecular assay. The Xpert Xpress gained FDA emergency authorisation for Covid-19 on March 20, a week earlier than the ID NOW, and takes 45 minutes to yield results.
The NYU researchers found that ID NOW missed a third of the samples that had been determined to be positive by Cepheid’s test when using nasopharyngeal swabs in viral transport medium, and over 48% when using dry nasal swabs.
Insisting that the results of the NYU study were “not consistent” with other trials of the ID NOW, Abbott said there had been a few trials with sensitivity percentages in the 80s and other with sensitivity at or above 90%, “and one as high as 94%”.
Sensitivity is more of a problem for viral RNA than antibody tests; for the latter, used in a screening setting, specificity is key (Covid-19 antibody tests face a very specific problem, April 22, 2020). Molecular tests are used in people who are already very ill with Covid-like symptoms, so sensitivity and specificity alike are important.
Abbott is now to conduct postmarket studies for ID NOW that will each include at least 150 Covid-19-positive patients in various clinical settings.
The company will of course be very keen to demonstrate high accuracy in these trials to reassure potential customers who might be spooked by the NYU data. But in any case it has also gained US emergency use authorisation for no fewer than three other Covid-19 tests – the most of any manufacturer.
One of these, an antibody test for prior Covid-19 infections, has just been approved by Public Health England, and Abbott says it will ship five million tests to the UK each month, starting immediately.
The other two are high-throughput viral RNA tests; these are much slower than ID NOW but run on Abbott’s laboratory analysers, which can run large numbers of tests at once – the Alinity m system, for example, can run up to 1,080 tests in 24 hours.
With no questions over the accuracy of these tests, payers might be keener on the slower but steadier options.