IntroductionClostridioides difficile infection (CDI) is the main cause of nosocomial diarrhea in developed countries. A key challenge in CDI is the lack of objective methods to ensure more accurate diagnosis, especially when differentiating between true infection and colonization/diarrhea of other causes. The main objective of this study was to explore the role of the microbiome as a predictive biomarker of CDI.MethodsBetween 2018 and 2021, we prospectively included patients with CDI, recurrent CDI (R-CDI), non-CDI diarrhea (NO-CDI), colonization by C. difficile, and healthy individuals. Clinical data and fecal samples were collected. The microbiome was analyzed by sequencing the hypervariable V4 region of the 16S rRNA gene on an Illumina Miseq platform. The mothur bioinformatic pipeline was followed for pre-processing of raw data, and mothur and R were used for data analysis.ResultsDuring the study period, 753 samples from 657 patients were analyzed. Of these, 247 were from patients with CDI, 43 were from patients colonized with C. difficile, 63 were from healthy individuals, 324 were from NOCDI, and 76 were from R-CDI. We found significant differences across the groups in alpha and beta diversity and in taxonomic abundance. We identified various genera as the most significant biomarkers for CDI (Bacteroides, Proteus, Paraprevotella, Robinsoniella), R-CDI (Veillonella, Fusobacterium, Lactobacillus, Clostridium sensu stricto I), and colonization by C. difficile (Parabacteroides, Faecalicoccus, Flavonifractor, Clostridium XVIII).DiscussionWe observed differences in microbiome patterns between healthy individuals, colonized patients, CDI, R-CDI, and NOCDI diarrhea. We identified possible microbiome biomarkers that could prove useful in the diagnosis of true CDI infections. Further studies are warranted.