Over the past two years, diagnosing a coronavirus infection has often required probing the nose. Health care workers have inserted slender swabs deep into the recesses of Americans’ nasal passages, while at-home test kits have asked us to master the shallow double-nostril twirl.
“The traditional approach to diagnosing respiratory infections has been to go after the nose,” said Dr. Donald Milton, an expert on respiratory viruses at the University of Maryland.
But the rapid spread of the omicron variant, and questions about the sensitivity of at-home tests, have rekindled a debate over whether the best way to detect the virus is to sample a different site: the mouth.
“The virus shows up first in your mouth and throat,” Milton said. “That means that the approach we’re taking to testing has problems.”
Collecting samples of saliva, or swabbing the inside of the mouth, could help identify people who are infected with the virus days earlier than nasal swabs do, some research suggests.
The science is still evolving, and the data paint a complex picture, suggesting that saliva-based tests have limitations of their own. Many labs are not currently set up to process saliva, nor are the at-home antigen tests available in the United States authorized for it.
But even the saliva skeptics acknowledge that oral specimens have some unique advantages. And with omicron on the march, some experts say that testing companies, labs and federal officials should be working more urgently to determine the best sample sites and types for the virus.
“We need to be adaptable,” said Anne Wyllie, a microbiologist at the Yale School of Public Health, who is one of the developers of SalivaDirect, a noncommercial polymerase chain reaction (or PCR) testing protocol. “I see so many either labs or governments who are so fixated on a certain sample type or a certain test that even with changing data or test preferences, they don’t make the necessary adaptations to their testing programs.”
The Case for Saliva
Scientists began investigating saliva testing in the early months of the pandemic. They were eager to find a testing method that would be more comfortable than the deep nasopharyngeal swabs that were the standard at the time and that would not require trained health care workers or nasal swabs, both of which were in short supply. With saliva, people could simply spit into a tube and hand it over for processing.
Some laboratory professionals were skeptical that saliva testing would be a reliable way to detect infection.
“There were concerns initially that saliva was not the gold standard sample, that it wasn’t the most sensitive sample,” said Glen Hansen of the clinical microbiology and molecular diagnostics laboratory at Hennepin County Medical Center in Minnesota.
But by fall 2020, dozens of studies had suggested that saliva was a suitable sample for testing.
“There’s been a growing body of evidence that at the very least, saliva performs well — it’s as good as, if not better, when it’s collected properly, when it’s processed properly,” Wyllie said.
Evidence also emerged that the virus tended to appear in saliva before it built up in the nose, suggesting that saliva samples might be the best way to detect infections early.
Milton and his colleagues recently found that in the three days before symptoms appear and the two days after, saliva samples contained about three times as much virus as nasal samples and were 12 times as likely to produce a positive PCR result. After that, however, more virus began accumulating in the nose, according to the study, which has not yet been published in a scientific journal.
The Food and Drug Administration has now authorized numerous saliva-based PCR tests, which have proved popular for screening students in schools.
“Saliva really has turned out to be a valuable specimen type and one that has increasingly been advocated as a primary testing sample,” Hansen said.
Saliva’s advantages may be more pronounced with omicron, which appears to replicate more quickly in the upper respiratory tract and have a shorter incubation period than earlier variants. Any testing method that can reliably detect the virus earlier is particularly valuable, experts said.
“I think omicron has really changed the testing game because of how quickly the virus replicates and how quickly it spreads,” said Dr. Robby Sikka, who chairs the COVID-19 Sports and Society Working Group and who helped bring saliva testing to the NBA in 2020. (Both Sikka and Wyllie serve as unpaid board members for SalivaDirect.)
Some experts have also theorized that omicron may be better at replicating in the cells of the mouth and throat than other variants have been.
A team of South African researchers recently found that while nasal swabs performed better than saliva swabs when detecting the delta variant, the opposite was true for omicron. (The study, which used PCR tests, has not yet been reviewed by experts.)
More research is needed, and another small new study, conducted at a San Francisco testing site during an omicron surge, was less encouraging. Of the 22 people who tested positive on a rapid antigen test using standard nasal swabs, only two tested positive when their inner cheeks were swabbed. The scientists are currently studying whether throat swabs perform better.
Saliva also has trade-offs. While the virus appears to build up in saliva early, the nose may be a better place to detect it later in the course of infection.
Researchers at the California Institute of Technology found that while the virus often spiked first in saliva, it ultimately rose to higher levels in the nose. Their results suggest that highly sensitive tests, like PCR tests, may be able to pick up infections in saliva days earlier than they do in nasal swabs, but that less-sensitive tests, like antigen tests, might not.
The data on saliva are still mixed, some experts noted.
“There are these few studies that I have found really very interesting,” said Dr. Mary K. Hayden, an infectious disease doctor and clinical microbiologist at Rush University Medical Center in Chicago.
But Hayden said she was interpreting the new studies cautiously because “for years and years and years,” research has suggested that nasopharyngeal specimens are best for detecting respiratory viruses.
Some scientists also have practical concerns. The mouth is “a little more of an uncontrolled environment compared to the nasal passages,” said Joseph DeRisi, a biochemist at the University of California, San Francisco, who is a president of the Chan Zuckerberg Biohub and an author of the cheek swab paper. “Did you drink a Coke right before you took the test? The pH will be different. And those things matter.”
Saliva can be “viscous and difficult to work with,” especially when patients are sick and dehydrated, Dr. Marie-Louise Landry, director of the clinical virology laboratory at Yale New Haven Hospital, said in an email.
Ultimately, different approaches may be required in different circumstances. For people who have had symptoms for several days, nasal swabs might be a good choice, while saliva might be best suited for the large-scale surveillance screening of asymptomatic people, Hansen suggested. “We need to get the right test into the right places,” he said.
In Britain, some at-home tests require swabbing both the throat and the nose, an approach that may be worth pursuing, experts said.
“Sampling multiple sites is always going to give you an edge,” Hayden said.
But if test manufacturers want to add saliva samples or throat swabs, they will need to validate their tests with those samples and submit the data to regulators. At a Senate hearing Tuesday, Dr. Janet Woodcock, acting commissioner of the FDA, noted that manufacturers might also have to reconfigure their tests to accommodate the larger swabs that are designed for the throat.
It is not yet clear whether any of the major at-home testing companies have plans to do so. “We continue to monitor and evaluate,” said John M. Koval, a spokesperson for Abbott Laboratories, which makes rapid antigen tests. “Our test is currently indicated for nasal use only.”
Even scientists who were convinced of saliva’s potential were reluctant to recommend that people swab their mouths or throats with tests that are not authorized for that purpose. (The FDA has also warned against this.) The biochemistry of the mouth is different from that of the nose and may affect the test results, potentially yielding false positives, scientists said.
“It’s not as easy as just saying, ‘Hey, just use a rapid antigen for saliva,’ ” Hansen said.
But experts said they hoped that laboratories, test manufacturers and regulators would move swiftly to evaluate whether any currently available tests might perform better on other sample types.
Ultimately, the country will be well-served by having a wide variety of testing options and the ability to switch between them as circumstances warrant, scientists said.
“For future pandemics, and even maybe as omicron evolves,” Hayden said, “we need flexibility in our testing systems.”