Repeat testing with antigen tests reduces false-negative results, leading FDA to update its testing recommendations
By now, most people have used an at-home test to find out if they have COVID-19. Rapid antigen tests are a fast and convenient way for people to test themselves for SARS-CoV-2, the virus that causes COVID-19. They’re a popular alternative to the polymerase chain reaction (PCR) tests that require a laboratory to process the results. With rapid antigen tests, you know the results in 15 minutes compared to up to a few days with a lab.
Despite these advantages, PCR tests are better (more sensitive) at detecting the SARS-CoV-2 virus than antigen tests—they generally produce a positive result 95% of the time when someone is infected compared to 80% of the time for antigen tests, according to the Food and Drug Administration (FDA).
Some people using antigen tests have experienced a negative test result only to find out that they have the virus. These false-negative results can occur early in the infection and among people without symptoms. A large clinical study, published in the Annals of Internal Medicine, looked at whether repeat testing every other day in this population would increase the antigen tests’ detection of the SARS-CoV-2 virus.
The Test Us at Home study was a collaboration with the FDA, NIBIB’s Rapid Acceleration of Diagnostics (RADx®)Tech program, and researchers from the University of Massachusetts Chan School of Medicine.
“The results clearly showed that serial testing with rapid antigen tests in a population that was screening for COVID-19 can perform as well as PCR tests,” said William Heetderks, M.D., Ph.D., a study co-author and an NIBIB consultant.
“Since the pandemic started, more than a billion antigen tests have been distributed. This study has a key public health impact because it showed how effective antigen tests are when used serially,” he said.
Participants in the study tested themselves using both antigen and PCR tests every 48 hours for 15 days. First, the researchers evaluated how well the antigen tests could detect the virus the day the infection began, which was the first day the participants tested positive with PCR. An antigen test sensitivity was higher among those with symptoms (60%) compared to those without symptoms (9%).
The performance of the antigen tests improved following serial testing: after 48 hours, the sensitivity increased among both symptomatic participants (92%) and asymptomatic participants (39%). The sensitivity continued to improve after an additional 48 hours (94% among symptomatic participants and 56% among asymptomatic participants).
Next, the researchers wanted to assess how antigen tests might perform in a real-world scenario—that is, when a person starts testing after the infection begins. To do this, they evaluated the performance of antigen tests on specific days after a participant first tested positive with PCR testing. Then, they combined the antigen test data spanning from the day the infection began through the following six days.
Among those with symptoms, the antigen tests’ sensitivity increased from 83% on the first test to 93% on the second test 48 hours later, according to the research paper.
The antigen tests’ sensitivity in people who tested positive without symptoms was lower, but it increased with repeat testing every other day, especially when researchers excluded participants who only tested positive with PCR once over the 15-day period. The sensitivity increased from 39% on the first test to 63% on the second test, and 79% on the third test in people who never developed symptoms, according to the research.
Rapid Antigen Testing Performance
48 Hours Apart
| Testing Once | Testing Twice | Testing Three Times | |
|---|---|---|---|
| Participants with Symptoms | 83% | 93% | n/a |
| Participants without Symptoms | 39% | 63% | 79% |
Based on the findings, the FDA updated its guidance last year for authorized uses of rapid antigen tests to recommend repeat testing every other day when users receive a negative result. Users should test at least twice if they have symptoms and three times if they don’t have symptoms.
The FDA also allowed antigen test developers to accept the new guidance and update their product instructions instead of conducting repeat testing themselves.
Innovative Virtual Study Design
A unique feature of the clinical study highlighted above was that it was conducted entirely virtually. The antigen and PCR test kits were shipped to participants’ homes and the PCR samples were picked up for lab analysis, the researchers reported in a related paper published in the Journal of Clinical and Translational Science. Participants used smartphones and a study app to enroll, give their consent, record their symptoms, and receive testing reminders.
The clinical study was also the largest to date to pair rapid antigen testing with PCR testing during the first week of the participants’ infection. Of the 7361 individuals enrolled in the study between late 2021 and early 2022, 492 tested positive for SARS-CoV-2, including 154 who had tested negative and were asymptomatic when the study started. These 154 individuals were included in the final analyses.
“This was a challenging study to do because we needed to catch the SARS-CoV-2 infection at the onset, which was part of the FDA requirement to have confidence in the results,” said Apurv Soni, M.D., Ph.D., principal investigator of the Test Us at Home longitudinal study and co-director of the digital medicine program at the University of Massachusetts Chan School of Medicine.
“We did that by leveraging digital health technology to recruit and enroll a large number of participants from across the country. The technology also gave us the flexibility to follow the pandemic and shift enrollment to SARS-CoV-2 ‘hot spots,’” said Soni.
Researchers were also able to enroll more people from rural areas, different races and ethnicities, and a wider age range (2 to 94) than when they conducted in-person enrollment, said Soni.
Soni said the findings from the Test Us at Home study also support the data from a different clinical study published in the COVID journal that found that participants were not biased by a previous testing history.
He mentioned a few limitations of the longitudinal study: A smartphone was needed to download the study app to enroll, which ruled out people who lacked the technology. Additionally, the study was only available to people who understand English or Spanish.
Soni added, “However, these limitations should be weighed against data suggesting that nearly 90% of Americans own smartphones, which is a three-fold increase from ten years ago, and more than 90% of Americans can understand English or Spanish.”
Also, the study was conducted when the omicron and delta variants were dominant during the pandemic so the findings may be different with newer variants.
Study References:
1. Soni A, Herbert C, Lin H, et al. Performance of Rapid Antigen Tests to Detect Symptomatic and Asymptomatic SARS-CoV-2 Infection : A Prospective Cohort Study. Ann Intern Med. (2023) https://doi.org/10.7326/M23-0385
2. Soni A, Herbert C, Pretz C, et al. Design and implementation of a digital site-less clinical study of serial rapid antigen testing to identify asymptomatic SARS-CoV-2 infection. Journal of Clinical and Translational Science. (2023) https://doi.org/10.1017/cts.2023.540
3. Lai E. Prior Knowledge of Diagnostic Results Does Not Bias A Subject’s Interpretation of At-Home COVID-19 Antigen Tests. COVID (2023). https://doi.org/10.3390/CoVid3040045
The clinical trials included in this study were supported by the NIBIB ITAP to speed the authorization and commercialization of at home COVID-19 antigen tests (Contracts #75N92022C00027 for the author and #75N92022D00010 for the clinical trials)
The Test Us at Home study was also funded by the NIH RADx® Tech program administered by NIBIB: Grant #U54HL143541-02S2 and #UL1TR001453.