Over-the-Counter Home Rapid Antigen SARS-CoV-2 Test

Of about 500 evaluable subjects, the test showed a PPA of about 82%, an NPA of 100% and an accuracy of about 98% against the RT-PCR comparator. There were no false positives and about ten false negatives; SARS-CoV-2 was not detected in a few of the false-negative specimens by a different high-sensitivity EUA RT-PCR method.

By age, PPA was 100% for ages 2-13, about 72% for ages 14-24, about 84% for ages 25-64 and about 80% for ages 65 and above, with NPA of 100% in every age group. By days since symptom onset, PPA ranged from about 70% on day 3 to 100% on days 0, 5, 6 and 7, where counts of positives were very small.

FDA issued the Emergency Use Authorization for non-prescription home use. Symptomatic individuals within the first 7 days of symptom onset are to test at least twice over three days with at least 48 hours between tests. Individuals without symptoms are to test at least three times over five days with at least 48 hours between tests. FDA later granted revisions to the authorization. FDA has announced that, following an HHS determination, COVID-19 IVD EUAs will no longer be in effect after a transition period.

Against FDA's published home-use template, the 100% NPA exceeds the 98% expectation, and the roughly 82% PPA meets the 80% point estimate recommended for OTC symptomatic use. The PPA lower bound sits just above the 70% lower bound that the supplemental serial-testing template ties to serial-testing claims. With only about 60 positives, the interval is wide: a few more false negatives would have pushed the lower bound below 70%. With a prevalence of roughly 11%, an all-comer study needs several hundred subjects to accumulate enough positives.

The age and day-of-onset strata are descriptive, not powered comparisons; the 14-24 age group, for example, rests on only a handful of positives and has a very wide interval. The discrepant RT-PCR results are given as a footnote and not used to adjust the primary 2x2 table, in line with the template. Because the PPA is below 95%, negative results are presumptive, and the serial-testing instructions in the authorized use are the risk mitigation.

The intended use was non-prescription home testing by lay users with no laboratory training. Individuals aged 14 years or older would self-collect an anterior nasal swab, and adults would collect from and test individuals aged 2 years or older. The clinical evidence therefore had to show performance with the specimen collected and the test run and read by intended users following only the quick reference guide in the kit.

The comparator was a separate nasal swab collected by a site healthcare professional and tested with a highly sensitive FDA EUA RT-PCR assay. The study had to enroll symptomatic subjects within seven days of symptom onset across pediatric and adult age groups, so that performance could be reported overall, by age group and by days since symptom onset. Because the authorization would extend to asymptomatic serial testing without an asymptomatic clinical cohort, the symptomatic PPA and its lower confidence bound carried particular weight.

During the COVID-19 public health emergency, home antigen tests reached the market through Emergency Use Authorization under section 564 of the Federal Food, Drug, and Cosmetic Act, not through 510(k) or De Novo review. FDA published EUA templates describing the data it expected. The Molecular and Antigen Home Use Test Template recommended a prospective clinical evaluation in the intended use population, with fresh samples collected, tested and interpreted by lay users and compared against paired samples run on an authorized high-sensitivity molecular test. It also recommended at least two U.S. testing sites, all-comer enrollment until 30 positives and 30 negatives were obtained, and an observed usability study before the clinical study.

The template set performance expectations by claim: NPA of at least 98%, and, for over-the-counter testing in symptomatic individuals, PPA of at least 80%. Tests with PPA below 95% were to be limited to presumptive negative results. The supplemental serial-testing template added a further route. A test could receive a serial-testing claim on the basis of a symptomatic-only study if it showed PPA of at least 80% with a two-sided 95% CI lower bound of at least 70%.

In 2022 FDA tightened how all antigen tests are used, drawing on a decentralized NIH RADx study in which participants serially tested with authorized OTC antigen tests. FDA revised every antigen EUA. Tests previously authorized for symptomatic individuals became authorized for use at least twice over three days, with at least 48 hours between tests. Tests authorized for asymptomatic individuals became authorized for use at least three times over five days, again with at least 48 hours between tests. Antigen tests authorized after that revision, including this one, carry the serial-testing indications from the outset.

• Having the lay user, or a parent for young children, collect and run the test with only the kit's quick reference guide meant the reported PPA reflected real home use, which is what FDA's home-use template asks for.
• Collecting the comparator swab by a healthcare professional and testing it with a high-sensitivity EUA RT-PCR assay followed the template's recommendation for an authorized highly sensitive molecular comparator. It also kept reference quality independent of the lay user.
• All-comer enrollment, not enrichment, avoided selecting for high viral loads, which the template warns would overstate sensitivity.
• Enrolling within seven days of symptom onset and stratifying by day produced the per-day data FDA asks antigen developers to report, and it supported the first-7-days indication.
• Including children aged 2-13 tested by a parent, and adults 65 and older, covered the full age range claimed in the intended use.