Incidence and predictive biomarkers of Clostridioides difficile infection in hospitalized patients receiving broad-spectrum antibiotics

Research output: Contribution to journalArticlepeer-review


  • Cornelis H. van Werkhoven
  • Annie Ducher
  • Matilda Berkell
  • Christine Lammens
  • Julian Torre-Cisneros
  • Jesús Rodríguez-Baño
  • Delia Herghea
  • Oliver A. Cornely
  • Lena M. Biehl
  • Louis Bernard
  • M. Angeles Dominguez-Luzon
  • Sofia Maraki
  • Olivier Barraud
  • Maria Nica
  • Nathalie Jazmati
  • Frederique Sablier-Gallis
  • Jean De Gunzburg
  • Jean de Gunzburg
  • Surbhi Malhotra-kumar
  • Marc J.M. Bonten
  • Maria J.G.T. Vehreschild

Institutes & Expert groups

  • UA - Universiteit Antwerpen
  • UCO - University of Cordoba
  • Hospital Universitario Virgen Macarena
  • The Oncology Institute - Laboratory of Radiotherapy - Radiobiology and Tumor Biology
  • University of Cologne/Universität zu Köln
  • Centre hospitalo-universitaire de Tours
  • University of Barcelona
  • University General Hospital of Heraklion
  • UL - University of Limoges
  • Clinical Hospital of Infectious and Tropical Diseases “Dr. Victor Babes”
  • Da Volterra
  • UMC - University Medical Center Utrecht

Documents & links


Trial enrichment using gut microbiota derived biomarkers by high-risk individuals can improve the feasibility of randomized controlled trials for prevention of Clostridioides difficile infection (CDI). Here, we report in a prospective observational cohort study the incidence of CDI and assess potential clinical characteristics and biomarkers to predict CDI in 1,007 patients ≥ 50 years receiving newly initiated antibiotic treatment with penicillins plus a betalactamase inhibitor, 3rd/4th generation cephalosporins, carbapenems, fluoroquinolones or clindamycin from 34 European hospitals. The estimated 90-day cumulative incidences of a first CDI episode is 1.9% (95% CI 1.1-3.0). Carbapenem treatment (Hazard Ratio (95% CI): 5.3 (1.7-16.6)), toxigenic C. difficile rectal carriage (10.3 (3.2-33.1)), high intestinal abundance of Enterococcus spp. relative to Ruminococcus spp. (5.4 (2.1-18.7)), and low Shannon alpha diversity index as determined by 16 S rRNA gene profiling (9.7 (3.2-29.7)), but not normalized urinary 3-indoxyl sulfate levels, predicts an increased CDI risk.


Original languageEnglish
Article number2240
Pages (from-to)1-10
Number of pages10
JournalNature Communications
Publication statusPublished - 14 Apr 2021


  • Microbiota, Biomarkers, Clostridioides difficile infection (CDI), Antibiotic treatment

ID: 7114786