RESUMO
This study aimed to evaluate diagnostic accuracy of SARS-CoV-2 RNA detection in saliva samples treated with a guanidine-based or guanidine-free inactivator, using nasopharyngeal swab samples (NPS) as referents. Based on the NPS reverse transcription-polymerase chain reaction (RT-PCR) results, participants were classified as with or without COVID-19. Fifty sets of samples comprising NPS, self-collected raw saliva, and saliva with a guanidine-based, and guanidine-free inactivator were collected from each group. In patients with COVID-19, the sensitivity of direct RT-PCR using raw saliva and saliva treated with a guanidine-based and guanidine-free inactivator was 100.0%, 65.9%, and 82.9%, respectively, with corresponding concordance rates of 94.3% (κ=88.5), 82.8% (κ=64.8), and 92.0% (κ=83.7). Among patients with a PCR Ct value of <30 in the NPS sample, the positive predictive value for the three samples was 100.0%, 80.0%, and 96.0%, respectively. The sensitivity of SARS-CoV-2 RNA detection was lower in inactivated saliva than in raw saliva and lower in samples treated with a guanidine-based than with a guanidine-free inactivator. However, in individuals contributing to infection spread, inactivated saliva showed adequate accuracy regardless of the inactivator used. Inactivators can be added to saliva samples collected for RT-PCR to reduce viral transmission risk while maintaining adequate diagnostic accuracy.