Mathematical modeling of clostridium difficile transmission in healthcare settings Public Deposited

Downloadable Content

Download PDF
Last Modified
  • March 22, 2019
  • Lofgren, Eric T.
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • Clostridium difficile is a frequent source of healthcare-associated infection, especially among patients on antibiotics or proton pump inhibitors (PPIs). The rate of C. difficile infection (CDI) has been steadily rising since 2000 and now represents a major burden on the healthcare system in terms of both morbidity and mortality. However, despite its public health importance, there are few mathematical models of C. difficile which might be used to evaluate our current evidence base or new control measures. Three different data sources were analyzed to provide parameters for a mathematical model: a cohort of incident CDI cases in the Duke Infection Control Outreach Network (DICON), a hospital-level surveillance time series, also from DICON, and inpatient records from UNC Healthcare, all from 7/1/2009 to 12/31/2010. Using estimates from these data, as well as from the literature, a pair of compartmental transmission models, one deterministic and the other stochastic, were created to evaluate the potential effect of the use of fecal transplantation as a treatment to prevent CDI. The analysis of the cohort of incident cases suggested that ICU patients experience a greater burden of mortality while infected with C. difficile and have longer lengths of stay and times until death, suggesting this population as one of special interest. Two interventions were simulated using the stochastic model: the use of fecal transplantation to treat CDI and prevent recurrent cases and the use of fecal transplantation after treatment with antibiotics or PPIs to prevent the development of CDI. Simulation results showed that treating patients with CDI was effective in preventing recurrence but not in reducing the overall number of incident cases of CDI. Transplantation after treatment with antibiotics or PPIs had no effect on preventing recurrence and a statistically significant reduction in incident cases that did not reach clinical significance. These results suggest that routine fecal transplantation for patients with CDI may be an effective treatment to prevent recurrence. Mathematical models such as the one described in this dissertation are powerful tools to evaluate potential interventions, suggest new directions for study, and understand the dynamics of infection on a population level.
Date of publication
Resource type
Rights statement
  • In Copyright
  • Weber, David
  • Doctor of Philosophy
Graduation year
  • 2013

This work has no parents.