Saliva as a testing sample for SARS-CoV-2 detection by RT-PCR in low prevalence community settings

Objectives The number of COVID-19 cases is increasing globally and there is an urgency for a simple non-invasive method for the detection of SARS-CoV-2. Our study aimed to demonstrate that saliva can be used as a specimen for SARS-CoV-2 detection notably for the screening of extensive population groups via pooling. Methods To demonstrate that saliva is an appropriate specimen for SARS-CoV-2 detection a field study including 3,660 participants was performed between September 29 and October 1, 2020. We collected paired nasopharyngeal/oropharyngeal swabs (NPS) and saliva specimens and processed them within 24 hours of collection. We performed 36 serial measurements of 8 SARS-CoV-2 positive saliva samples to confirm the stability of the specimen and completed 37 pools of saliva samples by adding one positive specimen per pool. Results Saliva specimens were stable for testing for up to 24 hours. Overall, 44 salival samples (1.2%) tested positive for SARS-CoV-2 during the field study. The results of saliva samples were consistent with those obtained from NPS from the same patient with 90% sensitivity (95% CI 68.3%-98.7%) and 100% specificity during the first two weeks after the onset of symptoms. Using pooling strategy 796 RT-PCR tests were performed. All pools showed 100% positivity in different pooling proportions. Conclusions Our findings demonstrate that saliva is an appropriate specimen for pooling and SARS-CoV-2 screening with accurate diagnostic performance. Patient-performed simple specimen collection allows testing an extensive number of people rapidly, obtaining results of the spread of SARS-CoV-2 and allowing authorities to take timely measures.