At-home direct rapid antigen tests may catch all new COVID-19 cases if used twice a week, new data suggest.

Researchers had 257 employees use the tests twice a week for six months and also undergo regular qRT-PCR tests. The home tests caught all 15 cases, 11 on day one or two, according to the report published in JAMA Network Open.

“We were faced with an issue to solve,” said the study’s senior author, Laura Holberger, vice president of strategic partnerships at BioInnovation Labs, LLC, a privately held network of co-working labs with locations in biotech hub cities across the country.

“Employees of biotech companies can’t work from home,” Holberger said. “We had to come up with a solution to help advance the research of the member companies.”

“The main take home is that with this protocol we were able to detect all 15 infections and prvent all cases of community spread during the pandemic,” Holberger said.

The new study included workers at three co-working laboratories in the Massachusetts cities of Cambridge and Boston. At the time, the prevalence of COVID-19 in this area ranged from less than 1 percent to 8 percent.

Participants self-collected nasal swab specimens twice weekly at home for the Direct Antigen Rapid Test (DART) from E25Bio Inc., and the findings were compared with laboratory qRT-PCR tests.

“The process of administering the antigen test involves swabbing both nostrils, suspending the sample in a small volume of buffer and using a dropper to apply the sample to the test cassette,” Holberger explained. “The sample flows over the test window which reads out the result with a positive control.”

Information on symptoms was collected contemporaneously. Self-reported race and ethnicity were also collected in accordance with Department of Health and Human Services and Food and Drug Administration reporting guidelines for non-laboratory-based tests.

The median participant age was 35, 120 (46.7 percent) were women, 161 (62.6 percent) were white, 49 (19.1 percent) were Asian, 29 (11.3 percent) were Hispanic, and 8 (3.1 percent) were Black. A total of 2,951 pairs of nasal swabs were self-collected by participants and tested by qRT-PCR and DART.

The sensitivity of DART within days 0 to 12 of symptom onset was 78.9 percent (60 of 76 swabs), and the specificity of DART was 97.1 percent (2,791 of 2,875 swabs).

The duration of SARS-CoV-2 nucleocapsid and RNA detection for individual infections ranged from one to 12 days, with peak levels observed between 2 and 6 days of symptoms (median, three days).

The sensitivity of DART was calculated for each day. DART sensitivity was 96.3 percent (26 of 27) within days zero to three of symptoms. Eleven participants tested positive on day one or two. One was presymptomatic the day of their initial DART positive result. One infection was detected by qRT-PCR one day before DART. For one positive participant, DART detected infection one day before qRT-PCR did.

There are a number of issues with the report, said Dr. Otto Yang, a professor of medicine in the division of infectious diseases and of microbiology, immunology and molecular genetics at the David Geffen School of Medicine at the University of California, Los Angeles.

First, Dr. Yang said, “they apparently use a home-brewed RT-PCR kit, and not an FDA-approved commercial kit. Is their test as good as the commercial tests? What was their limit of detection? It’s true that RT-PCR is the gold standard, but not all RT-PCR tests are equal.”

Beyond that, “the overall number of patients is only 15,” Dr. Yang said in an email. “By any standard, that’s pretty small. They inflate the number by mentioning the number of swabs since there were multiple swabs per patient, but the bottom line is that 15 is quite few.”

With so few participants, “the sensitivity and specificity numbers are hard to trust; they are doing these per swab rather than per patient,” Dr. Yang said. “It would be easy to get better specificity numbers by running more samples on negatives.”

Moreover, “many people are asymptomatic–especially if vaccinated–or shed virus for a couple of days before showing symptoms,” Dr. Yang said. “How does the test compare to PCR in those situations? From the standpoint of public health, this is critical. Confirming a diagnosis in someone who is already symptomatic is useful, but not as useful as finding infection in people without symptoms.”

The paper is in line with the current knowledge about antigen testing for COVID-19, said Dr. Heba Mostafa, an assistant professor of pathology at the JHU School of Medicine.

“Rapid antigen testing has lower sensitivity (can cause false negatives) compared to the gold standard nucleic acid amplification methods (for example, PCR), yet highly specific, Dr. Mostafa said in an email. “The sensitivity of antigen testing is largely highest when viral loads are high and within the first few days after symptoms onset.”

“Even though the gold standard for diagnosis so far remains nucleic acid amplification, as the authors mentioned, the cost could be a limitation as well as turnaround time in certain locations where molecular methods are not readily available,” Dr. Mostafa said. “In these situations, antigen testing could be helpful, especially if used frequently for diagnosis and quarantine. The impact of frequent antigen testing on reducing the transmission of SARS-CoV-2 has been proposed but to the best of my knowledge was not experimentally shown yet.”