ABSTRACT
CONTEXT.: Diagnostic testing for SARS-CoV-2 in symptomatic and asymptomatic children remains integral to care, particularly for supporting return to and attendance in schools. The concordance of SARS-CoV-2 detection in children, using various specimen types, has not been widely studied. OBJECTIVE.: To compare 3 sample types for SARS-CoV-2 polymerase chain reaction (PCR) testing in children, collected and tested at a single facility. DESIGN.: We prospectively recruited 142 symptomatic and asymptomatic children/young adults into a sample comparison study performed in a single health care system. Each child provided self-collected saliva, and a trained health care provider collected a mid-turbinate nasal swab and nasopharyngeal (NP) swab. Specimens were assayed within 24 hours of collection by using reverse transcription-polymerase chain reaction (RT-PCR) to detect SARS-CoV-2 on a single testing platform. RESULTS.: Concurrently collected saliva and mid-turbinate swabs had greater than 95% positive agreement with NP swabs when obtained within 10 days of symptom onset. Positive agreement of saliva and mid-turbinate samples collected from children with symptom onset >10 days prior, or without symptoms, was 82% compared to NP swab samples. Cycle threshold (Ct) values for mid-turbinate nasal samples more closely correlated with Ct values from NP samples than from saliva samples. CONCLUSIONS.: These findings suggest that all 3 sample types from children are useful for SARS-CoV-2 diagnostic testing by RT-PCR, and that concordance is greatest when the child has had symptoms of COVID-19 within the past 10 days. This study provides scientific justification for using sample types other than the NP swab for SARS-CoV-2 testing in pediatric populations.