Simulating interventions to prevent gastro-intestinal transmission of Clostridioides difficile and antimicrobial-resistant bacteria in healthcare settings: a model calibration study

Background Hospital-acquired infections caused by pathogens such as Clostridioides difficile and antimicrobial-resistant bacteria are a major burden to healthcare systems. Infection prevention and control (IPC) measures are used to limit transmission, but quantifying the impact of individual interventions remains challenging. Methods Here, we describe an agent-based model (ABM) that simulates faecal-oral transmission of pathogens under antimicrobial selection pressure in hospital settings. The model was calibrated to a C. difficile outbreak using approximate Bayesian computation. Findings The calibrated model reproduced the observed outbreak dynamics with rates of nosocomial infection that were similar to the true outbreak. In simulated scenarios, the pre-outbreak baseline rate was a median of 1.22 cases per 1000 patient days [interquartile range (IQR) range 1.02–1.44], compared with the real rate of 1.25 cases per 1000 [IQR 0.94–1.56]. During the outbreak period, simulations had a median rate of 2.42 [IQR 1.51–3.92] cases, compared with the real median rate of 3.74 [IQR 3.32–4.41]. Sensitivity analyses revealed that toilet cleaning frequency had the greatest impact on reducing nosocomial infections; antimicrobial stewardship and patient cohorting into single-occupancy rooms also reduced cases. Conclusion This ABM provides a flexible framework for evaluating IPC interventions and supporting outbreak preparedness. The approach is readily adaptable to other faecal-orally transmitted pathogens including vancomycin-resistant Enterococcus and carbapenemase-producing Enterobacterales.