1316. The Biology and Epidemiology of Clostridium difficile in Oxfordshire Hospitals 2007-2010
Session: Oral Abstract Session: Challenges in C. difficile Infection Surveillance
Saturday, October 20, 2012: 11:30 AM
Room: SDCC 29 ABCD
Background: Clostridium difficileis a major cause of healthcare-associated diarrhoea and controlling its spread is the focus of significant public health effort. However, its biology and transmission epidemiology are incompletely understood.

Methods: Admission records and ward movements for inpatient stays in hospitals in Oxfordshire were combined with results of C. difficile EIA testing and culture between 1 September 2007 and 1 March 2010 (ca. 750,000 hospital admissions, 931 culture-positive patients). C. difficile cases were resolved into 70 distinct types using Multi-Locus Sequence Typing. A stochastic compartmental model for transmission of C. difficile between hospital contacts (EIA positive, EIA negative and not tested), including potential transmission within and between wards, and ward contamination, was fitted to the available data using Markov Chain Monte Carlo. Whole genome sequencing was used to validate inferences from the model and confirmed it was well-calibrated.

Results: Use of a probabilistic modelling approach allows novel insights into the biology and epidemiology of C. difficile infection. We see strong evidence for heterogeneity between strains, particularly that NAP1/ST1/Ribotype 027 is more transmissible, being responsible for 50% of transmissions but only 13% of new introductions to the hospital. We find that a minority (22%) of patients continue to transmit for several months after initial diagnosis, and identify a potentially significant role for a median of 14 days (IQR 6-30 days) post-ward-discharge contamination leading to post-ward-discharge transmission. We find limited evidence for onward transmission prior to EIA test (likely onset of symptoms). We find no evidence for the existence of “superspreader” patients (the largest observed number of onward transmissions is 6), and find evidence for reductions in the amount of transmission over calendar time, and differences between the hospitals in the study. We confirm earlier findings that the majority of C. difficile cases cannot be explained by contact with symptomatic EIA-positive patients.

Conclusion: Transmission declined over time and varied by genotype. Statistical modelling provides a useful metric for assessing case-to-case transmission.

Madeleine Cule, PhD1, Rory Bowden, PhD2, David Eyre, BM, BCh3, A. Sarah Walker, PhD1, David Griffiths, BSc4, John Finney1, David Wyllie, PhD1, Derrick Crook, MB, BCh5, Tim Peto, MB BS, DPhil5, Peter Donnelly6 and Infections in Oxfordshire Research Database, (1)NIHR Oxford Biomedical Research Centre, Oxford, United Kingdom, (2)Statistics, University of Oxford, Oxford, United Kingdom, (3)Nihr Oxford Biomedical Research Centre, Oxford, United Kingdom, (4)National Institute for Health Research Oxford Biomedical Research Centre, Oxford, United Kingdom, (5)NIHR Oxford Biomedical Research Centre, University of Oxford, Oxford, United Kingdom, (6)University of Oxford, Oxford, United Kingdom


M. Cule, None

R. Bowden, None

D. Eyre, None

A. S. Walker, None

D. Griffiths, None

J. Finney, None

D. Wyllie, None

D. Crook, None

T. Peto, None

P. Donnelly, None

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 17th with the exception of research findings presented at the IDWeek press conferences.