534. Clostridium difficile Reduction: An Agent-Based Simulation Modeling Approach to Evaluating Intervention Comparative Effectiveness at Pediatric Hospitals

Abstract

Background Clostridium difficile surveillance data are lacking from pediatric facilities and there are few pediatric-centered guidelines or studies evaluating C. difficile targeted pediatric interventions. Compared with the adult setting, C. difficile control in pediatric healthcare facilities is also further complicated by epidemiologic variability across the age spectrum and increased patient-to-patient and patient-to-family interactions.MethodsWe constructed an agent-based simulation model of C. difficile transmission at a freestanding children’s hospital. The 80-bed hospital model included interactions between the physical environment, patients, visitors, family caregivers, nurses, and physicians. The model was then used to evaluate the comparative effectiveness of nine infection control interventions and six multiple-intervention bundles at reducing hospital-onset C. difficile infections and asymptomatic C. difficile colonization.ResultsThe most effective two-intervention bundle, composed of daily cleaning with sporicidal disinfectant and an asymptomatic C. difficile screening protocol, reduced hospital-onset C. difficile infection by 62.0% and asymptomatic colonization by 88.4%. Six of the nine single-intervention strategies also significantly reduced both outcomes, including daily and terminal cleaning, asymptomatic C. difficile screening, healthcare worker and patient hand hygiene, and reducing room transfers. The remaining three single-intervention strategies, visitor hand hygiene and visitor and healthcare worker contact precautions, did not significantly reduce either measure.ConclusionThis is the first mathematical model to evaluate pediatric C. difficile transmission. Hospitals can achieve a high rate of reduction for hospital-onset C. difficile infections by prioritizing implementation of a small number of interventions with high fidelity.Disclosures All authors: No reported disclosures.