Mapping the respiratory microbiome in intubated children over time.

Abstract

Children who require airway intubation are almost always treated with empiric antibiotics, pending clinical microbiology results. Positive pathogen results are near inevitable, due to the frequent presence of potential opportunistic pathogens, whether they are causing disease or not. This pattern leads to potential antibiotic over-prescribing, to the detriment of both patients who do not require antibiotics and to antimicrobial stewardship goals. To assess our hypothesis that opportunistic pathogens are more common in children with prior lung disease, we prospectively profiled tracheal aspirate (TA) microbiome samples in children with and without prior lung disease. An IRB approval was obtained, and TAs were collected longitudinally. Samples were analyzed using 16S rDNA sequencing and conventional cultures. Patient demographics and clinical courses were obtained using electronic medical records. Thirteen subjects were included, producing 39 TA samples. Microbiome analysis identified 98 bacterial genera, dominated by Pseudomonas and Streptococcus. Patient identity was a significant determinant of TA sample variation, indicating a robust individual TA microbiome despite the presence of substantial antibiotic exposures. In contrast, clinical category and time since intubation were not significant predictors, in the context of substantial inter-patient variation. Our results reveal substantial inter-patient variation in TA microbiome structure, limiting our ability to test for significant impacts of clinical category on microbiome structure. Our results provide information for the design of larger-scale studies to evaluate the role of clinical history and specific taxa in governing the interplay between antibiotics and pathogen dynamics in critically ill children.IMPORTANCEClinicians often prescribe empirical antibiotics for critically ill, intubated children with suspected respiratory infections, contributing to antibiotic overuse and challenging antimicrobial stewardship. Our longitudinal tracheal aspirate analysis of cultures and 16S rDNA sequencing revealed significant inter-patient variability, regardless of the primary reason for intubation. We observed both concordance and discrepancies between clinical microbiology and sequencing results-gram-negative organisms aligned well between methods, whereas Streptococcus was detected in 34 of 39 samples by 16S rDNA but only once by culture. Our findings emphasize the value of longitudinal airway microbiome analysis in pediatric patients. Given the heterogeneous pathologies and diverse age groups in pediatric intensive care, future large-scale studies should account for antibiotic exposure, commensal bacterial interactions, and clinical conditions that influence microbiome dynamics. Expanding research in this area could improve our understanding of microbial shifts in critically ill children and inform more targeted treatment strategies.