Diagnostic Accuracy of Saliva for SARS-CoV-2 Detection in State-sponsored Quarantine in Thailand

The aim of this study was to assess the diagnostic accuracy of saliva for detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) genomes among people in state-sponsored quarantine in Thailand. A cohort of 233 Thais in state-sponsored quarantine in Bangkok was enrolled into the study. Baseline demographic characteristics, presence of underlying diseases, and symptoms related to COVID-19 were collected on day 1 of the quarantine. Saliva specimens and nasopharyngeal (NP) swabs collected on day 7 at the quarantine premises were tested for SARS-CoV-2 RNA by real-time reverse transcription polymerase chain reaction. Overall, the viral RNA was detected in 32 (13.7%) NP swab samples, but only in 12 (5.2%) of the saliva samples. No person had NP negative but saliva positive result. Among the SARS-CoV-2 infected cases, nearly 20% had COVID-19-like illness and around 80% were asymptomatic. Sensitivity and specificity of saliva specimen were found to be 37.5% (95% confidence interval (CI)=21.1-56.3%) and 100% (95% CI=98.2-100%), respectively compared to the NP swab specimens. The area under the receiver operating characteristic curve was found to be 0.7 (95% CI=0.6-0.8). Our findings indicate that despite no false-positives, a high false-negative rate can occur with saliva specimen due to its low sensitivity, which limits its application in ruling out SARS-CoV-2 infection in quarantine settings.