Sensitivity and Specificity of Rapid SARS-CoV-2 Antigen Detection Using Different Sampling Methods: A Clinical Unicentral Study

Abstract

Rapid antigen detection of SARS-CoV-2 has been widely used. However, there is no con-sensus on the best sampling method. This study aimed to determine the level of agreement between SARS-CoV-2 fluorescent detection and a real-time reverse-transcriptase polymerase chain reaction (rRT-PCR), using different swab methods. Fifty COVID-19 and twenty-six healthy patients were confirmed via rRT-PCR, and each patient was sampled via four swab methods: oropharyngeal (O), nasal (N), spit saliva (S), and combined O/N/S swabs. Each swab was analyzed using an immuno-fluorescent Quidel system. The combined O/N/S swab provided the highest sensitivity (86%; Kappa = 0.8), followed by nasal (76%; Kappa = 0.68), whereas the saliva revealed the lowest sensitivity (66%; kappa = 0.57). Further, when considering positive detection in any of the O, N, and S samples, excellent agreements with rRT-PCR were achieved (Kappa = 0.91 and 0.97, respectively). Finally, among multiple factors, only patient age revealed a significant negative association with antigenic detection in the saliva. It is concluded that immunofluorescent detection of SARS-CoV-2 antigen is a reliable method for rapid diagnosis under circumstances where at least two swabs, one nasal and one oropharyngeal, are analyzed. Alternatively, a single combined O/N/S swab would improve the sensitivity in contrast to each site swabbed alone.