Abstract
BACKGROUND: Less invasive alternatives for nasopharyngeal swab samples are needed for the detection of respiratory pathogens in children. METHODS: Prospective diagnostic study comparing the accuracy and convenience of saliva and mouthwash samples with nasopharyngeal swab samples in detecting respiratory pathogens using a multiplex polymerase chain reaction panel for 19 respiratory viruses or viral subtypes and 4 respiratory bacteria. Acutely ill children aged 2-16 years with a suspected respiratory tract infection provided saliva and mouthwash samples in addition to a nasopharyngeal swab sample. RESULTS: Altogether, 850 samples were obtained from 302 children, including 300 nasopharyngeal swab, 277 saliva, and 273 mouthwash samples. Among 302 participants (mean age, 9.7 years [SD, 3.0], 151 males [50%]), at least 1 respiratory pathogen was detected in 219 (73%). The most common pathogens in nasopharyngeal swab samples were rhinovirus or enterovirus (86 [29%]), Mycoplasma pneumoniae (46 [15%]), influenza viruses A and B (32 [11%]), adenovirus (15 [5.0%]), and respiratory syncytial virus (14 [4.7%]). The sensitivity of saliva samples was 71% (95% CI, 60%-81%) for rhinovirus or enterovirus, 93% (95% CI, 82%-99%) for M. pneumoniae, 86% (95% CI, 64%-97%) for influenza A virus, and 88% (95% CI, 47%-100%) for influenza B virus. The sensitivity of mouthwash samples was 48% (95% CI, 36-60%) for rhinovirus or enterovirus, 90% (95% CI, 77%-97%) for M. pneumoniae, 73% (95% CI, 50%-89%) for influenza A virus, and 67% (95% CI, 30%-93%) for influenza B virus. In total, 99% (257/260) of the children preferred either saliva or mouthwash samples to nasopharyngeal swab samples. CONCLUSIONS: The sensitivity of saliva and mouthwash samples varied across respiratory pathogens in children. Less invasive saliva samples may serve as an alternative to nasopharyngeal swab samples for detecting certain respiratory pathogens in children.