NEW YORK – The US Food and Drug Administration on Wednesday provided new guidance for the validation of molecular and antigen tests to be used outside of laboratories for the diagnosis of SARS-CoV-2 infection.
The guidance follows updates on pooled, multi-analyte, and point-of-care testing Emergency Use Authorization submissions published by the agency earlier this week.
The new guidance was offered in documents and updates on FDA's website, and was reviewed in the 19th of the agency's weekly virtual town hall calls for developers during the ongoing COVID-19 pandemic.
In last week's town hall, Timothy Stenzel, director of the Office of In Vitro Diagnostics and Radiological Health at FDA's Center for Devices and Radiological Health, had noted that the agency intended to describe at-home testing as "non-laboratory" testing going forward, and to change the name of any then-in-development templates accordingly.
The non-laboratory template posted today notes that the settings covered include the home as well as non-traditional diagnostic testing sites, like offices, sporting events, airports, and schools.
The agency is "supportive of testing for COVID-19 that can be performed at home or in other non-laboratory settings," according to its FAQ page, "provided there is data and science to support consumer safety and test accuracy."
The new document – called the Template for Manufacturers of Molecular and Antigen Diagnostic COVID-19 Tests for Non-Laboratory Use – includes performance and safety recommendations, strategies for validation of assays and any associated smartphone apps, and suggestions for shipping and shelf-life evaluations. Developers must also explain how test results will be reported in accordance with local, state, and federal requirements.
The FDA recommends that the comparator sample for non-lab test evaluations be a healthcare provider-collected nasopharyngeal swab sample collected "within a reasonable time frame" from when the non-lab test sample was obtained, or a home-collected nasal or mid-turbinate swab. These must be assessed with a robust molecular SARS-CoV-2 test that has received EUA.
To make sure the non-lab test is usable by laypeople, the FDA suggests non-prescription over-the-counter tests should be run by at least 100 participants, while prescription-only tests can be run by only 30 participants.
For OTC tests that are to be used at non-laboratory sites, the usability study can be split such that 50 participants test themselves and 50 participants test another person, depending on the intended use population. For prescription-only tests for use on children, 15 of the 30 usability study participants should be parents or legal guardians performing the test on their children, the agency guides.
The FDA also suggests that participants in the usability studies represent different education levels and ages, and that a portion provide instructions in Spanish.
A separate clinical evaluation, meanwhile, should enroll symptomatic and non-symptomatic people, and should be continued until 30 positives are obtained, with an overall study size of no less than 150 individuals and the goal of at least 10 positive tests detected among asymptomatic people.
Non-prescription OTC tests for use in non-laboratory settings should have a positive percent agreement of 90 percent for asymptomatic and symptomatic people, and a negative percent agreement of 99 percent. Prescription tests are meant for use in symptomatic people, and so have a lower PPA bar of 80 percent. The FDA noted in the template this is because the inclusion of symptoms as a requirement for testing increases the pre-test probability of a positive result and therefore increases the positive predictive value of the test.
Other updates
In the town hall on Wednesday the FDA also noted it has updated its molecular diagnostic templates for commercial developers and for labs.
These documents now include additional information on recommendations for validation of pooling strategies and multi-analyte respiratory panels. The molecular diagnostic template for manufacturers also includes new recommendations for validation of point-of-care testing.
For pooling, the guidance now includes recommendations for validating pools made up of samples from aliquots of transport media and also pooled patient swabs.
Stenzel said that, overall, "nothing really has changed in what our recommendations are for validation," but noted the regulatory pathway for pooling has been clarified somewhat.
For pooling, "We have seen highly variable results, even on the same platforms in different labs," Stenzel said, adding, "We believe that the science here is still evolving, and we would like to be involved."
That said, the agency guides for pooling to be validated and the FDA to be notified, followed by submission of an EUA within 15 business days.
"During that whole time, you can use the test, as long as you don't experience any issues," Stenzel said. "This is a bit on the honor system, as we say, that the labs will do a good job validating this, and manufacturers as well, and we look forward to working with you to expand testing in this way."
The agency guides for a PPA of 85 percent between pooled tests and individual test results and said that a plan for monitoring of local positivity rates needs to be included.
For multi-analyte testing, the agency noted that respiratory panels that can detect SARS-CoV-2 and other infections, like influenza, are already authorized. The FDA clarified that in determining whether to authorize a test it considers whether the inclusion of other targets will aid in differential diagnosis of SARS-CoV-2, whether the other targets are already authorized, and how the panel fits into current public health recommendations, among other things.
For point-of-care testing, the agency said an EUA request should include data to demonstrate that non-laboratory personnel can perform the test accurately in the intended use environment, as well as data to demonstrate the robustness of the test device for near patient testing. Tests on a system that is already CLIA waived generally only require assay performance data, FDA said. The PPA and NPA should be 95 percent, although a PPA of 80 percent could be appropriate for some intended uses.
Finally, in the town hall call the FDA highlighted two amendments recently granted to tests that already have EUAs. An assay from Laboratory Corporation of America is now authorized for pooling and asymptomatic testing, and an assay from Quest Diagnostics is now authorized at certain sites for use with a high-throughput extraction method.
Regarding the Quest EUA amendment, Stenzel said, "Normally we don't make these updates that big of a deal — although they are all important to developers — but the new update allows Quest to test 3,600 more tests per day, and if you add pooling on top of that … they can substantially increase the throughput." He also said that FDA welcomes working with all kit developers and labs in this way, "to help provide more testing."