Sept. 14, 2022 – The Omicron variant of the coronavirus not only spreads faster and lessens the ability of vaccines to protect us, it also is more likely to fool common home tests, new research shows.
Researchers compared three rapid antigen tests designed to be used at home: Flowflex (Acon Laboratories); MPBio (MP Biomedicals); and Clinitest (Siemens-Healthineers) after Omicron emerged.
They found that the tests’ ability to correctly identify a positive sample dropped during the emergence of Omicron, from 87% to 81% for Flowflex; 80% to 73% for MPBio; and 83% to 70% for Clinitest. However, the decline was only considered statistically significant for Clinitest.
But throat samples done together with common nose samples improved the sensitivity of MPBio to 83% and Clinitest to 77%. The combination was not done with Flowflex.
Only One Test Met WHO Bar
But even with both nose and throat samples, only one test – MPBio -- met the World Health Organization’s standards of at least 80% sensitivity and at the ability correctly identify a negative sample at least 97% of the time among people with symptoms.
The study, published today in the journal The BMJ suggests, included 6,497 people with COVID-19 symptoms ages 16 and older who came in for testing at three public health service test sites in the Netherlands between Dec. 21 and Feb. 10.
Trained staff gave all study participants a PCR test and participants were then asked to take a rapid antigen test at home as soon as possible, but within 3 hours of their test site visit. They then filled out a questionnaire.
Nasal self-sampling alone was used during early Omicron and when Omicron accounted for more than 90% of infections (phase 1 of the study). When Omicron accounted for 99% of infections, the study moved to phase 2 and combined throat and nasal sampling.
The authors say their findings show that people with symptoms can rely on a positive rapid antigen test result regardless of the variant they have, “while individuals with a negative self-test result should adhere to general preventive measures because a false negative result cannot be ruled out.”
The authors write that, test manufacturers should consider updating their instructions to include both throat and nose samples.
Others: Adding Throat Testing Not the Answer
However, Timothy Feeney, research editor at The BMJ, and Charles Poole, a doctor of science and associate professor in the Department of Epidemiology at University of North Carolina, Chapel Hill, write in a linked editorial that adding throat sampling to nasal testing is not the answer.
They point out that sensitivity increased only slightly to 77.3% and 83% when throat sampling was added.
They also point out that the tests were inconsistent: MPBio had much better sensitivity when nose and throat were tested, whereas Clinitest’s improvement was less dramatic.
“No test, however, reached anywhere near the level of performance advertised by the manufacturers,” they write.
It’s also unclear, they say, how many at-home tests could be expanded to include nasal and throat sampling.
Feeney and Poole highlight the policy implications.
In the U.K. and the U.S., they write, policies around using tests to return to normal activities are “confusing, poorly explained, and frequently change.” In the U.S., for example, a single negative test result allows you to return to work or school in many situations. In the U.K., however, government guidance suggests that a negative result ”means it’s likely you are not infectious.”
Poole told the BMJ, "Given the less than ideal performance of antigen tests, updates to guidance in the public and private sector should take this into account when suggesting action based on test results,” according to a press release.
Additionally, people should consider their circumstances in deciding next steps after test results, the editorialists say, including whether they have COVID-like symptoms, have been exposed to an infected person, have been in a high-risk indoor environment or whether enough time has passed with the exposure to build up a high viral load.