OA Epidemiology

Modeling within-host and population transmission of OXA-48 in K. pneumoniae and E. coli

Proceedings of the 2013 annual meeting of the Netherlands Epidemiology Society

Volume 1 Issue S1 Abstract 22


M.R. Haverkate (1), M.J.D. Dautzenberg (1,2,3), J.M. Ossewaarde (3), A. van der Zee (3), J.G. den Hollander (4), A. Troelstra (2), M.J.M. Bonten (1,2), M.C.J. Bootsma (1,5)

1. Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, the Netherlands
2. Department of Medical Microbiology, University Medical Center Utrecht, Utrecht, the Netherlands
3. Department of Medical Microbiology, Maasstad Ziekenhuis, Rotterdam, the Netherlands
4. Department of Internal Medicine, Maasstad Ziekenhuis, Rotterdam, the Netherlands
5. Department of Mathematics, Utrecht University, Utrecht, the Netherlands

Detection of OXA-48 producing bacteria with non-molecular tests is hampered because of phenotypic variation, ranging from highly resistant (mostly in Klebsiella pneumoniae) to susceptible (e.g., in Escherichia coli) to carbapenems. Using detailed molecular microbiological data from an outbreak in a Dutch hospital involving 118 patients in combination with mathematical modeling we investigated the likelihood of large-scale (unnoticed) spread of OXA-48 through non-Klebsiella isolates.

The model consisted of a hospital and its catchment population, divided into two populations with different readmission rates based on duration since hospital discharge. Individuals were uncolonized with OXA-48 producing Enterobacteriaceae, colonized with E. coliOXA-48, with K. pneumoniaeOXA-48, or with both. Transmission occurred only in the hospital, loss of colonization after discharge, and horizontal gene transfer both in- and outside the hospital. The effects of interventions were expressed as a reduction in R0.

Results were derived from 868 longitudinal microbiological screening results from patients carrying K. pneumoniaeOXA-48 (n=24), E. coliOXA-48 (n=17), or both (n=40). The horizontal gene transfer rate from K. pneumoniae to E. coli was 0.0091/day and 0.0015/day vice versa. Mean durations of colonization were 278 and 225 days for K. pneumoniaeOXA-48 and E. coliOXA-48, respectively. Eradication of E. coliOXA-48 would reduce R0 by 1.9%. To achieve 20% reduction in R0, the duration of colonization with E. coliOXA-48 should be 6.1 years or both horizontal gene transfer rates should be 7.3 times higher.

It is unlikely that an undetected reservoir of E. coliOXA-48 will contribute significantly to the spread of OXA-48.

Published: 06 Jun, 2013

Licensee OA Publishing London 2013. Creative Commons Attribution License (CC-BY)