Abstract
BACKGROUND: Metagenomic next-generation sequencing (mNGS) offers unbiased pathogen detection. However, its integrative value in simultaneously revealing resistance, virulence, and host-response interplay in Intensive Care Unit(ICU)-infected patients remains underexplored. METHODS: In this retrospective cohort study of 156 ICU-infected patients, we compared the diagnostic performance of mNGS against conventional microbiological testing (CMT). We analyzed mNGS-derived antibiotic resistance genes (ARGs) and virulence factors (VFs) and correlated them with host immune-inflammatory markers and clinical outcomes. RESULTS: mNGS demonstrated a significantly higher positive detection rate (89.7% vs. 67.3%, P < 0.001) and clinical concordance (75.6% vs. 35.9%, P < 0.001) than CMT. It revealed a high mixed-infection rate (72.1%). ARGs were detected in 49.0% of bacterial infections, predominantly β-lactamase genes, showing 72.0% concordance with phenotypic susceptibility. Key VFs (e.g., rmpA in K. pneumoniae) were identified. Based on mNGS results, 47.4% of patients had their antimicrobial therapy adjusted. CONCLUSION: mNGS provides a comprehensive diagnostic tool by integrating pathogen identification, resistance and virulence profiling, and host-response context, enabling more precise and timely management of ICU-infected patients.