METHODS: We recruited 81 travelers and 15 non-travelers (including ten controls) prospectively within a mean of 3·22 days of RT-PCR confirmed COVID-19. Each study participant provided 2 mls of early morning fresh drooled whole saliva separately into a sterile plastic container and GeneFiX™ saliva collection kit. The saliva specimens were processed within 4 h and tested for SARS-CoV-2 genes (E, RdRP, and N2) and the results compared to paired NPS RT-PCR for diagnostic accuracy.

RESULTS: Majority of travellers were asymptomatic (75·0%) with a mean age of 34·26 years. 77 travelers were RT-PCR positive at the time of hospitalization whilst three travelers had positive contacts. In this group, the detection rate for SARS-CoV-2 with NPS, whole saliva, and GeneFiX™ were comparable (89·3%, 50/56; 87·8%, 43/49; 89·6%, 43/48). Both saliva collection methods were in good agreement (Kappa = 0·69). There was no statistical difference between the detection rates of saliva and NPS (p > 0·05). Detection was highest for the N2 gene whilst the E gene provided the highest viral load (mean = 27·96 to 30·10, SD = 3·14 to 3·85). Saliva specimens have high sensitivity (80·4%) and specificity (90·0%) with a high positive predictive value of 91·8% for SARS-CoV-2 diagnosis.