Abstract
Metagenomic next-generation sequencing (mNGS) is widely used to diagnose complex infections in hospitalized patients, particularly those associated with COVID-19 which has garnered significant concern over the past five years. To investigate the molecular epidemic of the viral variant and the potential co-infection pathogens, we conducted retrospective mNGS analysis of 254 SARS-CoV-2-positive specimens collected from 200 hospitalized patients between March and September 2023. Phylogenetic analysis of the identified Omicron subvariants showed minimal evolutionary divergence, with no association between sub-lineages and pneumonia severity. Notably, mNGS demonstrated enhanced detection of polymicrobial coinfections, identifying bacterial, fungal, and viral co-pathogens in 92.5% (185/200) of cases. Pneumonia severity was associated with advanced age (proportion of elderly patients: 61.1 vs 78.3%; p = 0.032) and comorbid conditions, particularly diabetes mellitus (OR 2.03, 95% CI 1.03-4.02, p = 0.041), but showed no correlation with SARS-CoV-2 sub-lineages or coinfecting pathogens. While mNGS enhances coinfection diagnosis, COVID-19 outcomes are predominantly driven by host factors rather than Omicron subvariant evolution. Prioritized monitoring of elderly and comorbid individuals remained critical for severe pneumonia management.