Pulmonary microbiome analysis and screening of prognostic markers in severe viral pneumonia

To investigate the association between the pulmonary microbiome and 90-day mortality in patients with severe viral pneumonia using metagenomic next-generation sequencing (mNGS). Cox proportional hazards regression and LASSO regression were used to identify variables associated with prognosis. An external cohort of 56 patients with severe pneumonia who underwent mNGS testing between June 2023 and February 2025 was used to assess the reproducibility of the association between pulmonary microbial diversity and clinical outcome. We retrospectively enrolled 87 patients with severe viral pneumonia admitted to the intensive care unit (ICU) between June 2022 and September 2023. The primary outcome was 90-day all-cause mortality. Differences in pulmonary microbial composition and diversity were compared between Survival group ( n = 23) and Death group ( n = 64). Compared with Dead group, Survival group exhibited significantly higher pulmonary microbial alpha-diversity, including the Shannon index ( P = 0.032) and Gini-Simpson index ( P = 0.012). Underlying diseases were not significantly associated with pulmonary microbial diversity. A prognostic model incorporating the Gini-Simpson index demonstrated a higher discriminative ability than the model based solely on conventional clinical variables (AUC: 0.863 vs. 0.799). Similar associations between pulmonary microbial diversity and clinical outcome were observed in the external cohort. Survival patients with severe viral pneumonia exhibited greater pulmonary microbial α-diversity and higher abundance of oral commensal bacteria, whereas Enterobacteriaceae were more frequently detected in Death group. Pulmonary microbial diversity may provide additional prognostic information for assessing 90-day mortality in patients with severe viral pneumonia.