Diagnostic accuracy of rapid antigen test for SARS-CoV-2: A systematic review and meta‐analysis of 166,943 suspected COVID-19 patients

Abstract

To assess the diagnostic accuracy of the rapid antigen test (RAT) compared with RT-PCR (reference standard) for SARS-CoV-2, we searched MEDLINE/PubMed and Web of Science for relevant records. The QUADAS-2 tool was used to assess study quality, and quantitative synthesis was conducted using a bivariate random-effects model. The meta-analysis included 135 studies (166,943 samples). The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, and diagnostic odds ratio were 0.76 (95%CI: 0.73–0.79), 1.00 (95%CI: 1.00–1.00), 276.1 (95% CI, 184.1–414.1), 0.24 (95% CI, 0.21–0.27), and 1171 (95% CI, 782–1755), respectively. Compared to other sample types, nasal samples had the best RAT sensitivity [0.79 (95%CI: 0.71–0.85)]. The sensitivities of the different RAT kits ranged from 0.41 (95%CI: 0.23–0.61) to 0.90 (95%CI: 0.70–0.97). Sensitivity was markedly better in samples with lower Ct, and RAT achieved excellent pooled sensitivity at 1.00 (95%CI: 0.70–1.00) among samples with Ct < 20. Testing within 10 days of symptom onset resulted in a high sensitivity. For ≤ 3, ≤ 7, and ≤ 10 days, the sensitivities were 0.91 (95%CI: 0.83–0.96), 0.89 (95%CI: 0.84–0.93), and 0.88 (95%CI: 0.83–0.92), respectively. RAT kits show high sensitivity and specificity in early infection, especially when the viral load is high. Moreover, using nasal samples for antigen testing, which are moderately sensitive and patient-friendly, is a reliable alternative to nasopharyngeal sampling. RAT might be effective for fighting the COVID-19 pandemic; however, it must be complemented by the careful handling of negative test results.