Dysbiosis of the gut microbiota is associated with in-hospital mortality in patients with antibiotic-associated diarrhoea: a metagenomic analysis.

Background: The increasing incidence of antibiotic-associated diarrhoea (AAD) is a serious health care problem. Dysbiosis of the gut microbiota is suspected to play a role in the pathogenesis of AAD, but its impact on the clinical outcomes of patients remains unclear.

Methods: Between May and October 2022, 210 patients with AAD admitted to a university hospital and 100 healthy controls were recruited. DNA extraction from stool specimens and shotgun sequencing were performed. Machine learning was conducted to assess profiling at different taxonomic levels and to select variables for multivariable analyses.

Results: Patients were classified into two groups: Clostridioides difficile infection (CDI, n = 39) and non-CDI AAD (n = 171). The in-hospital mortality rate for the patients was 20.0%, but the presence of C. difficile in the gut microbiota was not associated with mortality. Machine learning showed that taxonomic profiling at the genus level best reflected patient prognosis. The in-hospital mortality of patients was associated with the relative abundance of specific gut microbial genera rather than alpha-diversity: each of the five genera correlated either positively (Enterococcus, Klebsiella, Corynebacterium, Pseudomonas, and Anaerofustis) or negatively (Bifidobacterium, Bacteroides, Streptococcus, Faecalibacterium, and Dorea). Genes for vancomycin resistance were significantly associated with in-hospital mortality in patients with AAD (adjusted hazard ratios, 2.45; 95% CI, 1.20-4.99).

Conclusion: This study demonstrates the potential utility of metagenomic studies of the gut microbial community as a biomarker for prognosis prediction in AAD patients.