Abstract
BACKGROUND: Lower respiratory tract infections are the leading cause of morbidity and mortality in children worldwide. Timely and accurate pathogen detection is crucial for proper clinical diagnosis and therapeutic strategies. The low detection efficiency of conventional methods and low specificity using respiratory samples seriously hindered the accurate detection of pathogens. METHODS: In this study, we retrospectively enrolled 1,032 children to evaluate the performance of metagenomics next-generation sequencing (mNGS) using bronchoalveolar lavage fluid (BALF) sample and protected bronchial brushing (BB) sample in diagnosing pneumonia in children. In addition, conventional tests (CTs) were also performed. RESULTS: The specificity of BB mNGS [67.3% (95% CI 58.6%-75.9%)] was significantly higher than that of BALF mNGS [38.5% (95% CI 12.0%-64.9%)]. The total coincidence rate of BB mNGS [77.6% (95% CI 74.8%-80.5%)] was slightly higher than that of BALF mNGS [76.5% (95% CI 68.8%-84.1%)] and CTs [38.5% (95% CI 35.2%-41.9%)]. During the epidemics of Mycoplasma pneumoniae, the detection rate of M. pneumoniae in the >6-year group (81.8%) was higher than that in the 3-6-year (78.9%) and <3-year groups (21.5%). The highest detection rates of bacteria, fungi, and viruses were found in the <3-year, >6-year, and 3-6-year groups, respectively. mNGS detection should be performed at the duration of 5-7 days after the start of continuous anti-microbial therapy or at the duration of 6-9 days from onset to mNGS test. CONCLUSIONS: This is the first report to evaluate the performance of BB mNGS in diagnosing pulmonary infections in children on a large scale. Based on our findings, extensive application of BB mNGS could be expected.