Rapid COVID-19 tests done at home are reliable, study finds

In a study involving nearly 1,000 patients seen at a Baltimore field hospital during a five-month period in 2022, researchers at Johns Hopkins Medicine, the University of Maryland School of Medicine, and five other collaborators report that a rapid antigen test for detecting SARS-CoV-2, the virus that causes COVID-19, can be used at home with accuracy comparable to the same test being administered by a health care professional.

The study was first posted online on Feb. 13 in the American Society for Microbiology journal Microbiology Spectrum.

The researchers say their findings, based on patients seen at the Baltimore Convention Center field hospital, show that self-administered rapid tests rivaled the clinician-administered tests in both:

• Sensitivity, or the ability, according to the National Institutes of Health (NIH), to provide a positive result for a person infected with SARS-CoV-2; and
• Specificity or the ability, also according to NIH, to yield a negative result for a person not infected with the virus.

They added that both the self-administered and clinician-administered rapid tests evaluated in the study were comparable to the sensitivity and specificity of the established standard test for COVID-19—the polymerase chain reaction, or PCR, assay.

"We found the results between the self- and clinician-administered rapid tests were statistically similar in sensitivity—83.9% to 88.2%, respectively—and specificity—99.8% to 99.6%, respectively," says study senior author Zishan Siddiqui, assistant professor of medicine at the Johns Hopkins University School of Medicine. "We believe this is an important finding because it suggests physicians can feel confident prescribing treatment based on patient-reported, self-administered tests with positive results."

The first step for both the PCR and rapid antigen tests is obtaining a sample from a patient, either by nasal swab or collected saliva. The difference lies in how the sample is processed and analyzed. A PCR test takes a tiny bit of SARS-CoV-2 genetic material from a sample and reproduces it thousands of times so it can be more easily detected.

A rapid antigen test uses laboratory-produced antibodies to seek out and latch onto proteins on the surface of SARS-CoV-2 particles in the sample. The PCR test requires a skilled laboratory technician, special equipment, and up to an hour or more to process. Testing on a massive scale can only be conducted at a large, centralized testing facility, such as a hospital laboratory.

On the other hand, rapid antigen testing uses a premade kit with a reagent that contains antibodies specific for SARS-CoV-2. The test can be conducted by anyone, administered anywhere, and provides results in approximately 15 minutes.

For their study, the researchers enrolled 953 adult patients being seen at the Baltimore Convention Center between Feb. 12 and July 15, 2022. The study group was 60.6% female, 58.6% white, and 98.2% English-speaking. The median age was 34.

Participants reporting at least one COVID-19 symptom were categorized as symptomatic, while those reporting no symptoms were marked asymptomatic. Attending staff conducted both rapid antigen and PCR tests on the participants, and then had them independently perform a rapid test as well. The patients were asked to report their own results, which the researchers re-read and assessed.

In an earlier study at the Convention Center, the researchers showed that rapid tests could be broadly, quickly, and effectively applied to a large community such as Baltimore.

"What we determined from our previous work was that while the PCR test may be a better test from a clinical perspective—as it's basically 100% accurate at detecting SARS-CoV-2—the rapid antigen test provided significant advantages from a public health standpoint because of its ease of use, and the fact that it proved to have sufficient accuracy, specificity, and reliability for detecting the coronavirus in a high-volume setting," says study lead author Mary Jane Vaeth, a clinical assistant at the field hospital, which is now closed.

Added Siddiqui: "The first study suggested that a health care system can provide an equitable response to COVID-19, with [rapid antigen tests] making it possible to test all socioeconomic levels of a large population quickly and repeatedl. The new study bolsters that belief by showing comparable accuracy between self- and physician-performed [rapid antigen tests] and, therefore, helping build clinician confidence in patient-reported results from at-home tests."

The rapid antigen test used in this study is the BinaxNOW COVID-19 test manufactured by Abbott and supplied at no cost by the Maryland Department of Health.

"Despite the robust findings of our latest study, there were some limitations," Vaeth says. "The main one, of course, was the fact that we evaluated the self- and clinician-administered accuracy of only one brand of RAT, and therefore, future studies should look at an array of brands across a broader demographic spectrum so that our results can hopefully be confirmed and amplified."

Along with Siddiqui and Vaeth, the members of the study team from Johns Hopkins Medicine are Ishaan Gupta, Melinda Kantsiper, Charles Locke, Yukari Manabe, Henry Mitchtalik, Sophia Purekal, Resham Rahat,and Matthew Robinson.

Study co-authors from other institutions are Charles Callahan (BCCFH co-director), James Campbell, Minahil Cheema, E. Adrianne Hammershaimb and J. Kristie Johnson from the University of Maryland School of Medicine; Maisha Foyez, Asia Mitchell and Sarah Omer from the BCCFH; Maryam Elhabashy and Lucy Wilson from the University of Maryland Baltimore County; Aamna Cheema from the University of Maryland College Park; Kristie Sun from the Case Western Reserve University School of Medicine; Binish Javed from the Atal Bihari Vajpayee Institute of Medical Sciences and Dr. Ram Manohar Lohia Hospital; and the CONQUER COVID Consortium.

No funding support was received for this study.

Manabe receives salary support from National Institutes of Health grants U54EB007958, A1272201400007C and 3U54HL143541. The other study authors do not have financial or conflict of interest disclosures.