- Last week, researchers at New York University’s health system publish a report showing accuracy problems with Abbott’s rapid coronavirus test.
- They said it missed up to 50% of positive samples, depending on the collection method.
- The report instantly made headlines and prompted a warning from the US Food and Drug Administration.
- Abbott has major issues with the way the NYU researchers evaluated the machine.
- One executive called the study “not good science.”
- Visit Business Insider’s homepage for more stories.
Healthcare giant Abbott Laboratories is pushing back on a study that claims its rapid coronavirus test is about as accurate as a coin toss.
Abbott says its machine, called the ID Now COVID-19, is very reliable when used properly, and the study by researchers at New York University has several flaws.
The study was “totally inadequate,” said Philip Ginsburg, a senior medical director for Abbott’s rapid diagnostics wing, in an interview with Business Insider. It used “a skewed sample and not good science,” he said.
The team of researchers at the NYU health system found that Abbott’s test was prone to falsely telling people that they don’t have coronavirus, when they actually do. The rate of false negatives depends on how samples are handled beforehand, according to findings published on May 12.
The findings haven’t yet been peer-reviewed or published in a scientific journal, but they cast doubt on the accuracy of a coronavirus test that’s in use at the White House and across the country. They also helped prompt a warning from the US Food and Drug Administration, which said it was working with Abbott to further evaluate the reported accuracy issues.
Related News: ‘The Trip to Greece’ Review: Isles, Miles, Smiles
A lab technician dips a sample into the Abbott Laboratories ID Now testing machine at the Detroit Health Center in Detroit. Federal health officials are warning about potential accuracy problems with the rapid test for COVID-19 used at thousands of hospitals, clinics and testing sites across the U.S., including the White House. (AP Photo/Carlos Osorio, File)
The next day, FDA Commissioner Stephen Hahn said that many users have contacted the FDA saying they haven’t had any problems, but suggested sick people should get their results double checked nonetheless.
If a “doctor suspects a person has the disease, it might be worth it, if the test is negative, getting a second confirmatory test,” Hahn told CBS This Morning. “That’s what our guidance is about.”
US President Donald Trump, who uses the test himself, also spoke in support of it.
“It’s a great test. It’s a very quick test. And it can always be very rapidly double-checked,” Trump said, according to a transcript provided by the White House. “If you’re testing positive or negative, it can always be double-checked.”
Abbott has shipped more than 1.7 million test kits, which generate results in a matter of minutes, to all 50 states, according to the company. In data submitted to the FDA, it correctly diagnosed 100% of both negative and positive coronavirus samples in a study of 60 people. FDA said Abbott is now conducting further studies on the test.
NYU’s report stood in stark contrast to Abbott’s initial findings. When using “wet” nasal swabs in vials, Abbott’s test failed to identify a third of positive coronavirus samples in that study. When using dry nasal swabs, it missed nearly half.
Abbott’s Ginsburg had several criticisms of the methodology. According to the most recent instructions from Abbott, the testing instrument should be fed a fresh swab directly after it’s collected — without being stored in tubes and liquid called “viral transport media” beforehand.
When that’s not possible, Abbott says the swabs can be held in their original package at room temperature for up to two hours prior to testing. After that, they can be refrigerated for up to 24 hours from the time of collection.
NYU researchers said testing at the point of care wasn’t practical, since they had to transfer specimens from the hospital to the laboratory. Instead, they stored half the samples in wet tubes, and the other half in dry tubes.
The samples, taken from 101 patients at the Tisch Hospital Emergency Department, were transported to the lab within 2 hours, but the report doesn’t specify whether they were also tested within the same window.
NYU Langone Health “fully stands behind the findings in the study,” according to a spokesperson. The spokesperson declined to comment beyond the researchers’ published findings.
When ID Now won emergency authorization from the FDA on March 27, Abbott’s instructions said that the samples could be stored in a solution. The company worked to change that after clinicians reported accuracy problems, according to Abbott and a letter obtained by STAT News last month.
Ginsburg told Business Insider using tubes and liquid was allowed initially because it was an “urgently launched test” that the company wanted users to be able to validate independently.
“ID Now is designed to be used at point of care, near the patient, with a direct swab. And if the studies aren’t done in that manner and in a controlled study with sufficient numbers, then you get this total outlier,” Ginsburg said.
Storing the samples in this fashion can dilute the sample past the test’s “limits of detection,” according to Ginsburg. When the test receives a sample that’s under the limit, it says the sample is negative, Ginsburg said.
“The authors acknowledge the limitations of the study, which include a relatively small sample size and testing of the nasal swabs in the laboratory rather than at the point of care,” the NYU spokesperson said in a statement.
Ginsburg said his team has attempted to contact researchers involved in the study with a number of questions, like what happened to the samples once they arrived at the lab, but hasn’t been able to get many answers.
The report from NYU isn’t the only study to indicate issues with the Abbott test.
The Cleveland Clinic found that it generated a false negative reading 14.8% of the time, according to Dr. Francis Collins, director of the National Institutes of Health.
In research published in April in the Journal of Clinical Microbiology, Abbott’s rapid test missed about 25% of positive cases compared to the company’s slower coronavirus test that uses a large machine.
In these studies, as well as NYU’s, one big issue appears to be whether samples are tested immediately after they’re collected from patients. Both research teams stored them in tubes first, whereas more recent studies indicating better performance did not.
In a study from Detroit’s health department, for example, Abbott’s test correctly identified 48 out of 49 coronavirus-positive samples when swabs were inserted into the machine right after collection.
Abbott is working on further studies that will include 150 coronavirus-positive patients “in a variety of clinical settings,” according to the FDA.
Something is loading.