Clinical Performance of SARS-CoV-2 Rapid Antigen Tests: A Systematic Review and Meta-Analysis

ObjectivesWe conducted a meta-analysis of RAT diagnostic accuracy for SARS-CoV-2 infections, and further evaluated test sensitivity versus the presence of symptoms, days post symptom onset (DPSO), sample viral load, and sample type (i.e. direct swabs versus specimens stored in transport media). MethodsThree databases were searched systematically for performance evaluations of the Roche-distributed SDB SARS-CoV-2 Rapid Antigen Test (Roche/SDB RAT) through March 2022. If the Roche/SDB RAT was compared with any of 9 commonly available antigen tests, data from these tests were also included. ResultsOverall sensitivity of RATs among different manufacturers and study cohorts varied between 36.0% (95% CI: 24.0-50.1) and 79.4% (95% CI: 64.8-89.0). Roche/SDB RATs demonstrated a competitive performance with a pooled (including off-label use) sensitivity of 70.0%, and nearly 100% specificity in included studies. The Roche/SDB RATs exhibited reliable sensitivity in patients with a relatively high viral load (96.6% [95% CI: 95.2-98.2] for Ct[≤]25). Roche/SDB RATs were more sensitive in symptomatic patients within the first 7 DPSO (85.5% [95% CI: 81.2-88.4]), and when used to test direct swabs (74.4% [95% CI: 69.7-80.3]). ConclusionRATs show reliable performance in clinical settings and should be considered when rapid diagnosis of SARS-CoV-2 infection is critical. HIGHLIGHTSO_LIMeta-analysis of 86 studies of SARS-CoV-2 rapid antigen test (RAT) performance C_LIO_LIRAT performance supports near-patient testing for early COVID-19 diagnosis C_LIO_LIRAT specificity is high and sensitivity is reliable in those with high viral load C_LIO_LIRAT sensitivity in symptomatic patients is higher than in asymptomatic individuals C_LIO_LIRAT sensitivity is higher for direct swabs compared to swabs in transport media C_LI