Accuracy of Point-of-Care Rapid Antigen Tests for Diagnosis of COVID-19

Clinical Question

What is the diagnostic accuracy (i.e., sensitivity, specificity, positive predictive value, and negative predictive value) of point-of-care antigen testing for the SARS-CoV-2 virus?

Evidence-Based Answer

SARS-CoV-2 antigen tests have an average sensitivity of 69.3% (95% CI, 66.2% to 72.3%) and specificity of 99.3% (95% CI, 99.2% to 99.3%). Accuracy depends on the presence or absence of symptoms, time from symptom onset, and test brand.¹ (Strength of Recommendation: B, inconsistent or limited-quality patient-oriented evidence.)

Practice Pointers

The development of rapid diagnostic tests for the SARSCoV-2 virus has led to more access to testing. The goals of testing are earlier treatment and reduced transmission. The authors of the Cochrane review evaluated the accuracy of rapid diagnostic tests for SARS-CoV-2 infection.¹

The current review is the first update; the first review was released in 2020 and includes studies published through March 2021. The current review includes 155 cohorts from 166 studies of a single brand of rapid antigen test, and 152 evaluations of 49 different antigen assays were studied. There were 100,462 unique samples, and the presence of the SARSCoV-2 virus was confirmed by polymerase chain reaction testing in 16,822 cases. Most studies (66%) were conducted in Europe, and the remainder were in Asia (11%), North America (10%), South America (9%), and Africa (2%).¹

The average sensitivity and specificity of rapid antigen testing for SARS-CoV-2 infection were 69.3% (95% CI, 66.2% to 72.3%) and 99.3% (95% CI, 99.2% to 99.3%), respectively, based on 184 evaluations, 117,372 samples, and 21,017 confirmed SARS-CoV-2 cases. Average sensitivity was greater in those who had symptoms (73.0%; 95% CI, 69.3% to 76.4%; 109 evaluations; 50,574 samples; 11,662 confirmed SARSCoV-2 cases) vs. those without symptoms (54.7%; 95% CI, 47.7% to 61.6%; 50 evaluations; 40,956 samples; 2,641 confirmed SARS-CoV-2 cases). Average specificity was similar for patients who were symptomatic (99.1%) or asymptomatic (99.7%). Sensitivity was higher in patients who had symptoms for one week or less (80.9%; 95% CI, 76.9% to 84.4%; 30 evaluations; 2,408 cases) compared with those in their second week of symptoms (53.8%; 95% CI, 48.0% to 59.6%; 40 evaluations; 1,119 cases).¹

In patients who were asymptomatic at the time of testing, sensitivity was higher when epidemiologic exposure (i.e., contacts of patients with confirmed cases) to the SARSCoV-2 virus was suspected based on studies reporting specific criteria for testing or referral for testing in the absence of symptoms (64.3%; 95% CI, 54.6% to 73.0%; 16 evaluations; 7,677 samples; 703 cases). By contrast, when COVID-19 testing was reported to be widely available to any asymptomatic participant, sensitivity was lower (49.6%; 95% CI, 42.1% to 57.1%; 26 evaluations; 31,904 samples; 1,758 cases). Average specificity was similarly high in asymptomatic patients regardless of exposure to the SARS-CoV-2 virus (99.6% vs. 99.7%).¹

As the prevalence of COVID-19 rises, the positive predictive value of antigen tests improves. If the prevalence of COVID-19 in a community is 5%, an average antigen test would have a positive predictive value of 81%. If the prevalence is 10%, the average antigen test would have a positive predictive value of 90%. If the prevalence is 20%, antigen tests have a much better positive predictive value (95%). The negative predictive value in all three situations would remain above 95%.¹

Sensitivity varied widely among brands. Average sensitivities by brand ranged from 34.3% to 91.3% in symptomatic participants (20 assays with eligible data) and from 28.6% to 77.8% in asymptomatic participants (12 assays).¹

The authors noted that there was a lack of evidence for commercially produced tests because they were able to locate evaluations for only 49 of the 321 available antigen tests.¹ Only seven assays (AAZ, Abbott [BinaxNOW], BIONOTE, Denka Co, LumiraDx, Quidel, and Shenzhen Bioeasy) met the World Health Organization acceptable sensitivity standard of 80%; the 95% CIs of all but one of these tests (BIONOTE Nowcheck) crossed the 80% threshold.² A limitation of the review was the lack of inclusion of the SARS-CoV-2 variants, such as Delta and Omicron.

Although widely available, rapid antigen testing continues to perform more like confirmatory testing than screening, and family physicians should be prepared to counsel patients on the potential for false-negative results. The Infectious Diseases Society of America guidelines recommend the use of nucleic acid amplification tests over antigen-based tests, especially for symptomatic individuals or when the implications of missing the diagnosis of COVID-19 are significant (e.g., patients who are hospitalized or in long-term care facilities, when screening for asymptomatic infection before surgery).³ The Centers for Disease Control and Prevention guidelines recommend nucleic acid amplification testing following a negative antigen test result in symptomatic patients.⁴ Further study is needed to understand the evidence for repeat rapid antigen testing strategies in asymptomatic patients.

The practice recommendations in this activity are available at https://www.cochrane.org/CD013705.