The accuracy of saliva versus nasopharyngeal and/or oropharyngeal samples for the detection of SARS-CoV-2 in children. A rapid systematic review and meta-analysis

BackgroundThe comparative performance of saliva and nasopharyngeal samples for the detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection by reverse transcriptase polymerase chain reaction (RT-PCR) in children remains unclear. As schools reopen around the world, there is an interest in the use of saliva samples for the detection of SARS-CoV-2 in children to circumvent barriers with nasopharyngeal sampling. We systematically reviewed the literature to understand the performance of saliva sampling using RT-PCR on naso- and/or oropharyngeal swabs as the reference standard. MethodsArticles from PubMed/MEDLINE and Living Evidence were accessed until 28th April 2021. A search method without restriction to children population was applied and during the review phase, if a study included patients <18 years old, authors were contacted to provide additional information on the subset of children. Studies were eligible if they reported on matched saliva and naso- and/or oropharyngeal samples, taken from the same patient on the same day. Studies using other respiratory samples such as sputum samples were excluded. Each paired patient sample had to be tested on the same RT-PCR platform. ResultsTen studies were included, comprising 1486 matched saliva and on naso- and/or oropharyngeal pairs from children aged 0 to 18 years old. The pooled absolute sensitivity and specificity of saliva sampling using RT-PCR on nasopharyngeal samples as the reference standard was 84.5% (95% CI; 78.0%-90.3%) and 99.5% (95% CI; 98.2%-100.0%). Comparable performance of saliva to nasopharyngeal samples was shown in both symptomatic and asymptomatic children. Stratified analyses of various covariates showed no significant differences. DiscussionOur pooled accuracy estimates of RT-PCR SARS-CoV-2 testing on saliva in children did not seem to be different from meta-analyses of studies that enrolled mainly adults. Saliva could potentially be considered an alternative sampling method for screening in children and to pick up those with high viral load.