Characteristics of Gut Microbiota in Patients with Severe Pneumonia and its Potential Clinical Relevance

Abstract

Gut microbiota is associated with a variety of diseases, but its relationship with severe pneumonia remains to be explored. This study primarily analyzed the intestinal microbiota of patients with severe pneumonia and examined its association with clinical data. We collected clinical data from 96 patients with severe pneumonia for differential analysis and identified prognostic factors using logistic regression. Fecal samples from patients with severe pneumonia and healthy controls were collected and analyzed using 16S rRNA sequencing. We applied three machine learning algorithms (LASSO, Random Forest, and SVM) to identify microbial markers associated with severe pneumonia. The patients were grouped by “discharge status”. Significant differences were observed in age ( p = 0.025), total length of hospital stay ( p < 0.001), and C-reactive protein (CRP) ( p < 0.001). Logistic regression analysis identified age ( p = 0.024) and total hospital stay ( p < 0.001) as factors influencing the likelihood of improvement and discharge. Diversity analysis of collected stool samples revealed differences between the two groups. LDA Effect Size (LEfSe) analysis highlighted significant microbial differences at various taxonomic levels between the two populations. Three machine learning algorithms identified 9 microbial markers for severe pneumonia. A diagnostic prediction model was constructed, with an area under the Receiver Operating Characteristic (ROC) curve of 0.969 (95% CI: 0.946–0.992). Patients with severe pneumonia exhibit unique intestinal microbiota characteristics, which may be regulated by age and total length of hospital stay, thereby influencing the disease’s prognosis.