Validation of Saliva and Self-Administered Nasal Swabs for COVID-19 Testing

BackgroundActive cases of COVID-19 has primarily been diagnosed via RT-PCR of nasopharyngeal (NP) swabs. Saliva and self-administered nasal (SN) swabs can be collected safely without trained staff. We aimed to test the sensitivity of "naso-oropharyngeal" saliva and SN swabs compared to NP swabs in a large cohort of migrant workers in Singapore. MethodsWe recruited 200 male adult subjects: 45 with acute respiratory infection, 104 asymptomatic close contacts, and 51 confirmed COVID-19 cases. Each subject underwent NP swab, SN swab and saliva collection for RT-PCR testing at 1 to 3 timepoints. We additionally used a direct-from-sample amplicon-based next-generation sequencing (NGS) workflow to establish phylogeny. ResultsOf 200 subjects, 91 and 46 completed second and third rounds of testing, respectively. Of 337 sets of tests, there were 150 (44.5%) positive NP swabs, 127 (37.7%) positive SN swabs, and 209 (62.0%) positive saliva. Test concordance between different sample sites was good, with a kappa statistic of 0.616 for NP and SN swabs, and 0.537 for NP and saliva. In confirmed symptomatic COVID-19 subjects, the likelihood of a positive test from any sample fell beyond 14 days of symptom onset. NGS was conducted on 18 SN and saliva samples, with phylogenetic analyses demonstrating lineages for all samples tested were Clade O (GISAID nomenclature) and lineage B.6 (PANGOLIN nomenclature). ConclusionThis study supports saliva as a sensitive and less intrusive sample for COVID-19 diagnosis and further delineates the role of oropharyngeal secretions in increasing the sensitivity of testing. However, SN swabs were inferior as an alternate sample type. Our study also provides evidence that a straightforward next-generation sequencing workflow can provide direct-from-sample phylogenetic analysis for public health decision-making.