Changes in the gut microbiota and risk of colonization by multidrug-resistant bacteria, infection, and death in critical care patients.

OBJECTIVES: This study aimed to investigate whethehr the diversity and composition of the intestinal microbiota determines the risk of multidrug-resistant organism (MDRO) acquisition, infection, and mortality in patients admitted to a liver intensive care unit (ICU). METHODS: This prospective study included patients admitted to a 12-bed ICU between July and December 2018. Rectal swabs to detect MDRO intestinal colonization were obtained at ICU admission and weekly thereafter during the ICU stay. The 16S rRNA gene sequencing was performed on 138 rectal swabs from 62 patients. We evaluated the potential association between gut microbiota composition and diversity and the risk of MDRO colonization, infection, and hospital mortality. RESULTS: Of the patients studied, 19 of 62 (30.65%) presented with MDRO colonization at admission, 16 (25.81%) were colonized during their stay, and 27 (43.55%) were not colonized; 45 of 62 patients (72.58%) developed an infection, and mortality was 29.03% (18 of 62). Higher bacterial diversity and abundance of Bacillales Family XI incertae sedis and Prevotella families were associated with a lower risk of colonization by MDRO, infection, and death (linear discriminant analysis effect size score >4), whereas the Enterococcaceae family was associated with an increased risk of infection and death (linear discriminant analysis effect size score >4). The LASSO regression and multivariate analysis identified Family XI incertae sedis to be associated with a lower risk of infection (OR: 0.997; 95% CI, 0.996-0.999; p = 0.001) and microbial evenness index to be associated with lower mortality risk (OR: 0.68; 95% CI, 0.49-0.95; p = 0.02). DISCUSSION: Microbial diversity and abundance of certain bacterial taxa could have prognostic value in patients admitted to a critical care unit. Larger perspective studies should address the value of these markers in clinical practice.