Rapid antigen testing using a nasal swab with the BinaxNOW test was able to detect the Omicron variant at an urban community testing site, whereas other swabbing routes did not perform as well, researchers found.
Sensitivity of the BinaxNOW assay for detecting PCR-positive cases within a cycle threshold (Ct) of <30, indicating a higher viral load, was 74% (95% CI 53.7-88.9%) using a throat swab versus 97.7% (95% CI 87.7-99.9%) using the nasal swab alone, reported John Schrom, MPH, of the University of California San Francisco, and colleagues.
“The throat swab significantly underperformed compared to the standard nasal swab,” they wrote in a pre-print on medRxiv. “Even with the prominent clinical feature of pharyngitis in persons with omicron, our data argue against replacing nasal swabs with throat swabs for diagnosis.”
Cheeks swabs were clearly the worst performing, though. Among those patients tested simultaneously via nasal and cheek swabs, the sensitivity of cheek swab specimens was 4.3%, and 9.0% for a Ct threshold <30, the authors said, adding that “our data are compelling that a simple oral cheek swab does not increase detection of SARS-CoV-2.”
Schrom’s group examined data from 4 days in January 2022 at a community testing site in downtown San Francisco. All ages were eligible. Lab assistants collected anterior nasal swabs for BinaxNOW antigen tests along with RT-PCR tests on the first 2 days of data collection.
Nasal swabs and cheek swabs were collected from a cohort of 75 randomly selected patients on the third day of data collection and analyzed via RT-PCR and antigen testing. Nasal and oral tonsillar swabs were collected from a cohort of 115 randomly selected patients on the fourth day.
Overall, RT-PCR and rapid antigen testing were performed on 731 samples during the first 2 days of testing. Over three-quarters of patients were Hispanic/Latinx (77%). About 53% of patients were men, and 15% were younger than 12. The most common motivation for testing was clearance for work (26%), followed by known or suspected exposure (23%) and school requirement (18%).
About 41% of these samples tested positive via RT-PCR. Out of 186 individuals who tested positive via RT-PCR and had a Ct of <30, 177 tested positive with BinaxNOW. Overall, the rapid antigen test was the most sensitive among samples with the lowest Ct values. Test sensitivity for Ct <30 was 95.2% (95% CI 92-98%), but Ct <35 fell to 82.1% (95% CI 77-87%), and overall test sensitivity was 65.2% (95% CI 60-71%). BinaxNOW detected none of the 61 RT-PCR positive cases with a Ct >35.
Unsurprisingly, BinaxNOW had higher sensitivity among symptomatic patients (97.6%, 95% CI 93-100%), though sensitivity among asymptomatic patients was still high (89.8%, 95% CI 79-96%).
Schrom disclosed no conflicts of interest.
One co-author disclosed support from Gilead Sciences and Abbott.