Skip to Main Content

The Food and Drug Administration said late Thursday that the Abbott ID Now rapid test for Covid-19 — touted by the White House and widely used nationally — may return false negative results, meaning it can incorrectly indicate that patients who are infected with the novel coronavirus are not.

“We are still evaluating the information about inaccurate results and are in direct communications with Abbott about this important issue,” Tim Stenzel, director of the FDA’s Office of In Vitro Diagnostics and Radiological Health, said in a statement. “We will continue to study the data available and are working with the company to create additional mechanisms for studying the test.”

advertisement

Stengel said that while the test can still be used and correctly identify positive test results in minutes, negative results may need to be confirmed with another test. 

STAT first reported on accuracy problems with the Abbott rapid test a month ago.

Support STAT: If you value our coronavirus coverage, please consider making a one-time contribution to support our journalism.

In a statement, Abbott said that it is seeing studies that use the ID Now test “in ways it was not designed to be used.” The company pointed to recent studies that have had false negative rates of 20%, 10%, and one as low as 6%. In previous statements to news outlets, Abbott has said that the false positive rate of the test was only 0.02%. More than 1.8 million Abbott ID Now tests for the novel coronavirus, SARS-CoV-2, have been distributed, and the test is reportedly in use at the White House.

advertisement

The normal process of testing new diagnostic tests has not been possible during the Covid-19 pandemic. But Anirban Maitra, a professor of pathology at M.D. Anderson Cancer Center, said that all the samples were simulated by putting known quantities of viral genetic material on swabs, and without comparing the results to other platforms. There are enough real world Covid-19 cases, he said, to do a head-to-head comparison with more traditional tests.

Abbott will now conduct studies that will include at least 150 Covid-19-positive patients “in a variety of clinical settings” and the FDA will continue to review data from the company on an ongoing basis.

Abbott had previously said that the reason that the test did not perform well in some areas was that nasal swabs were placed in “viral transport media,” a standard practice that it said diluted samples too much for the ID Now.



But on Wednesday, researchers at NYU Langone Medical Center released a paper that showed that even with dry nasal swabs, the Abbott ID Now test missed 48% of infections that were detected with more traditional platforms, which use a technology called the polymerase chain reaction, or PCR.

The NYU study has not yet been peer-reviewed by other researchers or published in a medical journal. In a tweet, Michael Mina, a Harvard epidemiologist, wrote that the paper is “flawed” and said that the way the samples were taken could explain the difference. He said the experiment compared two different kinds of swabs, swabbed in different parts of the nose, and argued that could explain the difference, not the machines that were used.

Abbott now says that if negative results are inconsistent with patients’ symptoms, patients should be tested with an alternative diagnostic test. The company said it is also “reinforcing proper sample collection and handling instructions.”

In a technical brief sent to diagnostic laboratories and obtained by STAT, Abbott lists a number of potential causes of false negatives, including improper collection, transport, or handling of samples and low levels of virus in a patient. The company also said it would be adding new instructions to the test, including one saying that swabs should be circulated in liquid that is provided for at least 10 seconds to help remove the sample from the swab.

Ed Silverman contributed reporting.

STAT encourages you to share your voice. We welcome your commentary, criticism, and expertise on our subscriber-only platform, STAT+ Connect

To submit a correction request, please visit our Contact Us page.