Abstract
This retrospective study compared metagenomic next-generation sequencing (mNGS) and traditional culture for pathogen detection in 43 patients with lower respiratory tract infections (LRTI), including 34 COVID-19 cases (14 critical, 20 non-critical) and 9 non-COVID controls. mNGS demonstrated superior sensitivity (95.35% vs. 81.08%) and broader pathogen coverage, identifying 36.36% of bacteria and 74.07% of fungi detected by cultures. Concordance between methods was observed in 63% of cases. Severe COVID-19 patients exhibited reduced respiratory microbiota abundance, potentially linked to viral dominance or therapeutic interventions. Clinical outcomes correlated positively with inflammatory markers (procalcitonin/PCT, N-terminal pro-B-type natriuretic peptide/N-proBNP, neutrophils, lactate dehydrogenase/LDH, neutrophil-to-lymphocyte ratio/NLR) and negatively with lymphocytes, highlighting systemic inflammation's role in disease progression. While mNGS offers rapid, high-sensitivity pathogen profiling, limitations include small sample sizes, unresolved specificity concerns and unmeasured confounders. The study underscores mNGS as a promising tool for LRTI diagnosis in COVID-19, though larger prospective cohorts and standardized outcome metrics are needed to validate clinical utility, optimize interpretation, and address cost-effectiveness compared to conventional methods.