Evaluation of the specificity and accuracy of SARS-CoV-2 rapid antigen self-tests compared to RT-PCR from 1015 asymptomatic volunteers

ObjectiveEvaluation of the specificity and accuracy of four CE-approved SARS-CoV-2 antigen rapid self-tests (AG-ST) Anbio, Clungene, Hotgene and Mexacare. Method1015 asymptomatic volunteers were screened for SARS-CoV-2 by means of an oropharyngeal swab taken by qualified personnel and subsequent RT-PCR testing. Each participant additionally performed nasal self-swabs for two of the four rapid antigen tests at the same day according to the manufacturers instructions. Study participants transmitted a photo and own interpretation of their test results to the study center. The results of the two self-tests provided by the participants were correlated with the results of the SARS-CoV-2 RT-PCR and independently assessed and evaluated by the study center. ResultsNone of the volunteers tested positive upon RT-PCR, whereas 13 AG-ST showed a false positive test result (0.7 %). The highest false positivity rate was found for the Clungene test (2.1 % compared to 0.2 % for the other tests), while the highest test failure rate (invalid) was found for the Mexacare test (3.7%). The Anbio and Hotgene tests produced the fewest false positive results when evaluated by the participants and also showed the best agreement among themselves. ConclusionSARS-CoV-2 Antigen rapid self-tests with higher false positive test rates, such as the Clungene test, or with high rates of invalid test results, such as the Mexacare test, are less suitable for screening purposes of asymptomatic study participants especially in low-prevalence settings. False positive or inadequate test results increase the burden on certified test laboratories due to verification PCR tests and cause a substantial economic loss due to unnecessary quarantine measurements and cause psychological stress in the affected study participants. In addition to earlier defined requirements for sensitivity for SARS-CoV-2 detection, a lower acceptance boundary for the false positivity rate of < 0.3% should be demanded.