SARS-CoV-2 antigen tests have significant advantages over real-time reverse transcription PCR (RT-PCR) and other nucleic acid amplification tests. Rapid SARS-CoV-2 antigen tests are comparatively inexpensive, can be performed at the point of care, and can provide a result within 30 minutes. Providing a rapid result to identify SARS-CoV-2-infected individuals can help to prevent and truncate the spread of disease, as quarantine and contract-tracing measures can be initiated sooner than in individuals who do not receive such a test (5–7). However, many of the available rapid SARS-CoV-2 antigen tests have been shown to lack sensitivity (8–11). False negative results can have significant public health consequences. In the context of a false negative result, many - individuals with mild disease may believe they are not infected and may never return for another test, and thus may contribute to the further transmission of disease throughout their disease course. In this study we show that the LumiraDx SARS-CoV-2 Ag Test has a low false negativity rate of 3.8% compared to PCR when used in a community-based healthcare setting.
The product package inserts for other leading rapid SARS-CoV-2 antigen tests report false negative rates similar or even better than those reported for the LumiraDx SARS-CoV-2 Ag Test (14–16). For example, the BinaxNOW COVID-19 Ag Card (Abbott Diagnostics Scarborough, Inc.; Scarborough, ME, USA) and the Veritor System for Rapid Detection of SARS-CoV-2 (Becton, Dickinson and Company; Sparks, MD, USA) report false negative rates of 5.1% and 2.5% respectively, while the Sofia 2 Flu + SARS Antigen FIA (Quidel Corporation; San Diego, CA, USA) reports a false negative rate of only 0.6% (14–16). However, the false negativity rate is influenced by disease prevalence and hence false negativity rates may not be directly comparable, especially when disease prevalence varies significantly. In fact, the false negativity rates for these products were much higher when deployed and determined in an actual community setting (11, 17–19). Furthermore, there are significant differences in the studies represented in the product package inserts compared with the study reported here. The studies represented in the product package inserts are all smaller than the study of 800 patients reported here, with the BinaxNOW COVID-19 Ag Card, Veritor System for Rapid Detection of SARS-CoV-2, and Sofia 2 Flu + SARS Antigen FIA studies consisting of 460, 226, and 164 patients respectively (14–16). Furthermore, the study of the Veritor System for Rapid Detection of SARS-CoV-2 only included symptomatic patients within 5 days of symptom onset (15). The studies of the BinaxNOW COVID-19 Ag Card and Sofia 2 Flu + SARS Antigen FIA included symptomatic patients within 7 days of symptom onset, but the majority of patients included in the Sofia 2 Flu + SARS Antigen FIA study were within 4 days after symptom onset (14, 16). The study reported here included symptomatic patients within up to 12 days of symptom onset.
A false negative test result could be rendered when SARS-CoV-2 viral loads are low, such as very early (20, 21) and very late (22) in the disease course of COVID-19. Studies have demonstrated that rapid SARS-CoV-2 antigen tests have the greatest sensitivities when viral loads are high (23, 24). This suggests that the false negativity rate of the BinaxNOW COVID-19 Ag Card, the Veritor System for Rapid Detection of SARS-CoV-2, and the Sofia 2 Flu + SARS Antigen FIA could have been higher if the respective studies had included patients who were tested more than 7 days of symptom onset, when viral load levels would have been declining. In support of this hypothesis, studies have demonstrated that the limit of detection for the Veritor System for Rapid Detection of SARS-CoV-2 and the Sofia 2 Flu + SARS Antigen FIA has a sensitivity equivalent to a molecular CT of 27–28 (25). In contrast, the LumiraDx SARS-CoV-2 Ag Test is more sensitive, with a sensitivity equivalent to a molecular CT of < 33 (13). The increased sensitivity of the LumiraDx SARS-CoV-2 Ag Test versus other rapid antigen tests is likely the consequence of the testing methodology. The LumiraDx SARS-CoV-2 Ag Test is a rapid microfluidic immunofluorescence assay that utilizes active control of the reaction time, volume, and temperature. The assay utilizes a wash step and a spectrophotometric read. In contrast, the BinaxNOW COVID-19 Ag Card, Veritor System for Rapid Detection of SARS-CoV-2, and Sofia 2 Flu + SARS Antigen FIA all utilize conventional lateral flow testing methodology. Lateral flow test methods entail only passive control of the specimen, with no wash steps.
The BinaxNOW COVID-19 Ag Card and the Sofia 2 Flu + SARS Antigen FIA both have FDA EUA claims for utilization in patients within up to 7 days of symptom onset, while the Veritor System for Rapid Detection of SARS-CoV-2 has an FDA EUA claim for utilization in patients within up to 5 days of symptom onset. Studies have demonstrated that SARS-CoV-2 can be successfully cultured in symptomatic patients 8–15 days following symptom onset (26–28). If the presence of culturable virus provides at least some measure of infectivity, then it would be important that a rapid SARS-CoV-2 antigen test be able to identify infected patients more than 7 days after symptom onset. The LumiraDx SARS-CoV-2 Ag Test has an FDA EUA claim for utilization in patients up to 12 days after symptom onset, and this study demonstrated that the LumiraDx SARS-CoV-2 Ag Test has a low false negative rate of 3.8% compared to PCR, even when patients presenting up to 12 days after symptom onset were included in the analysis.
This study also demonstrated that 11/17 (64.7%) of the false negative results rendered by the LumiraDx SARS-CoV-2 Ag Test had CTs > 30. This suggests that many of the specimens rendering a false negative result with the LumiraDx SARS-CoV-2 Ag Test had low viral loads and were potentially non-infectious (13). A previous study demonstrated that the LumiraDx SARS-CoV-2 Ag Test had 100% concordance with SARS-CoV-2 PCR-positive specimens with CTs < 33 (13). The difference between the results of this previous study and those reported here may be reflective of the differences in the PCR assays used in each study. Drain et al. utilized the Roche SARS-CoV-2 PCR assay, while the study presented here primarily utilized an LDT SARS-CoV-2 PCR assay (13). CTs achieved across different PCR-based assays are not directly comparable.