Last reviewed: September 6, 2022
The following is a curated review of key information and literature about this topic. It is not comprehensive of all data related to this subject.
Overview
The use of rapid antigen tests for diagnosis of SARS-CoV-2 infection is a key element of COVID-19 prevention outside of health care settings.
These tests differ from molecular tests such as RT-PCR because they are designed to detect the presence of viral proteins using simple immunochromatography methods, commonly referred to as lateral flow assays. Rapid antigen tests target the SARS-CoV-2 nucleocapsid protein, the most abundant protein expressed by the virus. They directly assess for the presence of viral proteins, making them different from serology tests, which look for antibodies produced by the host in response to the infection. Thus, rapid antigen tests assess for acute infection only, not prior infection or response to vaccination.
Rapid antigen tests offer a number of advantages over molecular assays. Testing can be deployed outside of hospital laboratories and is designed to be performed by members of the general public. They are relatively inexpensive relative to RT-PCR and simple to interpret. They have a turnaround time as fast as 15 minutes, allowing for testing prior to entrance into congregate care facilities or other public settings. They are compatible with samples taken from the anterior nares rather than the nasopharynx, making them more comfortable for serial testing.
There are now more than 45 antigen tests approved by FDA for use in health care facilities, including 17 tests that are cleared for use at home. FDA maintains an up-to-date list of rapid COVID-19 tests available in the U.S.; performance reviews have been published in the New England Journal of Medicine (Drain, January 2022 [PDF]).
Comparison to RT-PCR
A major difference between rapid antigen tests and RT-PCR is the difference in the analytic sensitivity of the assay. Typically, the sensitivity of a single antigen test is 30% to 40% lower than for RT-PCR, depending on whether tested subjects were symptomatic or asymptomatic (Cochrane Library, March 2021).
However, lower sensitivity has both disadvantages and benefits. The primary disadvantage is a risk of falsely negative results in people with low viral loads who may be early in their infection and who go on to spread it to others in subsequent days. In practice, this risk can be mitigated through serial testing algorithms. A major advantage of these assays is the lower likelihood of detecting residual viral nucleic acid left over from a remote infection in recovered individuals (Mina, November 2020). This reduces the chance of unnecessary initiation of isolation and quarantine precautions and subsequent rounds of testing. Furthermore, when limiting comparisons to people with high viral loads, the sensitivity of antigen testing increases to above 90% in certain studies (Pilarowski, April 2021; Ford, September 2021).
There is a slightly elevated rate of false positives relative to molecular tests, though the rate is dependent on the prevalence of disease and the proportion of people who are symptomatic. For many commonly used rapid antigen tests, the negative predictive value (e.g., the likelihood someone with a negative test is truly negative for infection) is greater than 98% (Boum, August 2021).
Rapid antigen tests are subject to the same considerations as molecular tests with respect to factors that affect clinical sensitivity. These include the quality of sampling and the timing of testing relative to the onset of infection. They have been noted to have diminished performance in the asymptomatic population, which may be related to the lower levels of virus in this group relative to those with symptoms, rather than characteristics of the tests themselves (Pray, January 2021).
Use Cases for Rapid Antigen Testing
Rapid antigen tests for SARS-CoV-2 are in widespread use, but best practices for their deployment continue to evolve. In general, rapid antigen tests have lower sensitivities and specificities than RT-PCR based tests when performed at a single time point, which is why many testing programs have deployed confirmatory RT-PCR for positive antigen tests in asymptomatic (low pretest probability) individuals and for negative antigen tests in symptomatic individuals and close contacts of positive cases (high pretest probability) settings.
However, serial testing increases the sensitivity of rapid antigen tests, is much more practical than serial RT-PCR testing and can therefore play an important role in identifying infections during the critical early transmission period. For example, in a prospective cohort study of rapid antigen tests conducted during the first Omicron surge (October 2021 to February 2022), performance of rapid antigen tests during the first week of infection was best when asymptomatic participants tested three times at 48-hour intervals (sensitivity increased from 38% to 79%) and when symptomatic participants tested two times separated by 48 hours (sensitivity of two serial tests was 93%) (Soni, August 2022 - preprint, not peer-reviewed). These and similar data were the basis of revised August 2022 FDA guidance on serial at-home rapid antigen testing.
Rapid antigen tests have also been evaluated for use in certain patient populations and settings given their important role in transmission mitigation efforts. A study performed at two universities in Wisconsin using the Quidel Sofia 2 SARS-CoV-2 antigen assay with mid-turbinate swabs reported a sensitivity of 79% in symptomatic people and 44% in asymptomatic people relative to RT-PCR (Wu, January 2022; Ford, April 2021; Pray, January 2021). The majority of false negative results occurred in people with low amounts of the virus, inferred by high cycle threshold values on RT-PCR. Data for pediatric populations suggests lower sensitivity of antigen testing relative to tests performed in adults, with results ranging between 30% to 60% lower depending on the study (Ford, December 2021; Villaverde, May 2021). This is an important consideration for school-based testing programs. Antigen tests may also play a crucial role in diagnosis for lower-resource settings that may not have ready access to PCR testing with actionable turnaround times (Hodges, August 2021; Prince-Guerra, January 2021). The extent to which true PCR-positive/antigen-negative individuals carry live, transmissible virus is not known.
Scant data exist on changes in performance for antigen tests in vaccinated individuals, but there is a theoretical risk that receipt of a COVID-19 vaccine could result in lower viral loads, which would reduce the assay’s sensitivity. However, a number of studies have now shown that peak viral loads are similar between vaccinated and unvaccinated individuals during recent SARS-CoV-2 variant waves (Kissler, December 2021; Brown, August 2021).
Antigen Tests and Infectivity
While the risk of transmission of SARS-CoV-2 is directly proportional to the viral load in the index case, there are a number of other factors that are also at play. These include:
- The degree of host immunity, which may lead to virus particles being coated in neutralizing antibodies;
- The duration of exposure to other people;
- Ventilation in the exposure space;
- The use, quality and fit of masks;
- The inherent transmissibility characteristics of the virus variant itself.
For these reasons, it is challenging to say a positive rapid antigen test means a person is contagious and a negative test means they are not contagious. To answer this question, it is necessary to study the correlation between antigen test positivity and contacts of exposed individuals and carefully document the modulators of transmission noted above.
Guidelines
IDSA guidelines make no recommendations for or against using rapid tests (i.e., result time ≤1 hour) versus standard nucleic acid amplification testing (e.g., RT-PCR) in symptomatic individuals suspected of having COVID-19, citing knowledge gaps.
Key Literature
General performance
Rapid Diagnostic Testing for SARS-CoV-2 (Drain, January 2022).
This is a comprehensive review of rapid diagnostic testing for SARS-CoV-2, which mostly focuses on rapid antigen tests. The article covers the relationship between viral load and clinical diagnostics, as well as different algorithms for test interpretation depending on pretest probability for infection.
Serial Antigen Rapid Testing in Staff of a Large Acute Hospital (Wu, January 2022).
This study, which took place in July 2021 during the Delta wave, screened asymptomatic hospital staff at a 120-bed tertiary care hospital in Singapore twice weekly with antigen testing or daily if they had a known exposure. This approach identified 20 true positives and 11 false positives. There was no comparator group, but standard of practice was PCR screening every 14 days.
Epidemiologic Characteristics Associated With SARS-CoV-2 Antigen-Based Test Results, rRT-PCR Cycle Threshold Values, Subgenomic RNA and Viral Culture Results From University Testing (Ford, September 2021).
This study evaluated the test performance of the Quidel Sofia SARS Antigen Fluorescence Immunoassay versus RT-PCR and viral culture in a population of university students and staff in Wisconsin. The authors found the antigen test to have a sensitivity of 79% and 44% in symptomatic and asymptomatic people, respectively. Of nine asymptomatic people with false negative antigen tests, five went on to develop symptoms. All 15 false positive antigen tests came from a single lot and may be attributable to manufacturing artifacts. Virus was culturable from 46% of positive cases, including 12% of false negative antigen tests.
Use of Rapid Antigen Testing for SARS-CoV-2 in Remote Communities — Yukon-Kuskokwim Delta Region, Alaska, September 15, 2020–March 1, 2021 (Hodges, August 2021).
This study evaluated the impact on case incidence of the introduction of Abbott BinaxNOW antigen tests to an existing RT-PCR testing program in rural Alaska. Due to infrastructure challenges, the mean turnaround time for test results went from 4 to 12 days during the surge of winter 2020. The addition of antigen testing reduced test turnaround times back down to 2 days, and with it, case incidence also steeply dropped. Antigen tests were reserved for testing symptomatic people and close contacts of index cases, while PCR testing was utilized for screening. Through close coordination with local health departments, the use of rapid antigen testing in this setting was able to change the trajectory of SARS-CoV-2 outbreaks in this rural community.
Rapid, Point‐of‐Care Antigen and Molecular‐Based Tests for Diagnosis of SARS-CoV-2 Infection (Dinnes, March 2021).
This Cochrane systematic review looked at 78 studies encompassing 24,087 samples tested across 16 antigen assays in Europe and North America. Nearly all studies were considered biased due to participant selection, lack of an appropriate reference or timing of testing. The pooled estimate of sensitivity was 72% in symptomatic people and 58% in asymptomatic people. Sensitivity was 78% in the first week after symptoms and dropped to 51% thereafter. There were significant differences by brand. Assuming a prevalence of 5%, the best performing tests, which included the Abbott PanBio, would result in one in six to one in 10 results being false positive and one in four to one in six being falsely negative. The Abbott Panbio is the equivalent of the Abbott BinaxNOW marketed in the U.S.
Surveillance Testing for SARS-CoV-2 Infection in an Asymptomatic Athlete Population: A Prospective Cohort Study With 123,362 Tests and 23,463 Paired RT-PCR/Antigen Samples (Harmon, June 2021).
This study evaluated the performance of the Quidel Sofia SARS Antigen Fluorescence Immunoassay versus RT-PCR in college athletes in the Pac-12 college sports conference. Screening antigen testing was performed 6 days per week alongside once weekly RT-PCR, with additional testing occurring as part of contact tracing and symptomatic testing. Eighty-nine of 172 positive study participants were detected by antigen testing on days they would not have been screened for PCR, which resulted in 234 days of potential infectiousness that was avoided. There were also 98 false positive tests and seven instances of false negatives. PCR tests turned positive on average a day before antigen tests. Two outbreaks occurred during the study period despite the aggressive testing program.
Pediatric populations
Antigen Test Performance Among Children and Adults at a SARS-CoV-2 Community Testing Site (Ford, December 2021).
This study examined the analytic sensitivity of the Abbott BinaxNOW rapid antigen test in 225 children <18 years of age tested at a community testing site. The antigen test was positive in 73% (27/37) of specimens that were identified by PCR compared to 81% (240/297) in adults. Among children with a reported exposure, antigen test sensitivity was 86%. Live virus was culturable from 70% of PCR+/Ag+ tests and from 0% of PCR+/Ag- tests.
Diagnostic Accuracy of the Panbio SARS-CoV-2 Antigen Rapid Test Compared With RT-PCR Testing of Nasopharyngeal Samples in the Pediatric Population (Villaverde, May 2021).
This is a retrospective study of symptomatic children tested for SARS-CoV-2 across seven medical centers in Spain. The authors calculated the sensitivity and specificity of the Abbott PanBio SARS-CoV-2 rapid antigen test to be 45.4% and 99.8%, respectively, to RT-PCR. At a prevalence of 4.8%, the positive predictive value was 92.5%, and the negative predictive value was 97.3%. This study utilized nasopharyngeal swabs only.
Omicron
Comparison of SARS-CoV-2 Reverse Transcriptase Polymerase Chain Reaction and BinaxNOW Rapid Antigen Tests at a Community Site During an Omicron Surge (Schrom, March 2022).
This study looked at 731 people tested at a community testing site in San Francisco during the Omicron wave in January 2022. Study participants were tested using anterior nares swabs on the Abbott BinaxNOW assay versus RT-PCR. Overall prevalence was 40.5%, and antigen test sensitivity was 65% overall. Sensitivity rose to 95% in those with a cycle threshold value of <30 and 82% in those with a Ct value of <35. The test detected 98% of symptomatic individuals with a Ct value <30. Compared to RT-PCR from nasal swabs, cheek swabs had a sensitivity of just 4%, while throat swabs were moderately better with a sensitivity of 55% at Ct values <35 and 74% at Ct values <30.