Clostridium difficile
Clostridium difficile bacteria from stool sample (credit: Janice Haney Carr, CDC)

Whole-genome sequencing tracks hospital infections

Researchers in the High-Throughput Genomics core have sequenced the complete genomes of more than 1200 Clostridium difficile bacteria isolated from patients in Oxford hospitals as part of a major effort to track the transmission of infection. The results suggest that sources of infection are much more diverse than previously suspected.

Outbreaks of bacterial infection are a huge problem in hospitals, and controlling them depends on understanding how the microbes are transmitted. The Modernising Medical Microbiology initiative, led by Professor Derrick Crook of the Nuffield Department of Medicine’s Experimental Medicine Division, is demonstrating the potential of whole-genome sequencing in tracing sources of hospital-based infection for four different pathogenic organisms. Much of the sequencing for the project, which to date has investigated around 13,000 samples, has taken place at WTCHG.

Results on C. difficile, based on samples isolated from all patients with suspected infections in four Oxford hospitals over three years, have just been published in the New England Journal of Medicine. The researchers examined how the sequences of bacteria extracted from stool samples differed both from reference C. difficile genomes and from each other. They applied statistical techniques to compared these data with epidemiological information about possible contacts between infected patients who might have occupied the same wards or been in hospital at the same time.

Doctors had previously assumed that most patients catch C. difficile in hospital from other infected patients, and infection-control practices were traditionally based on this assumption. However, whole-genome sequencing has shown that only 35 per cent of cases of infection were genetically related to a previous case in the same hospital and same time period. A majority of cases of infection were not acquired from another infected patient. Indeed, nearly half of the isolates studied could not, on the basis of their genetic sequences, have come from any of the other 1200 patients in the study.

The authors conclude that there are many diverse sources of C. difficile infection in the environment, each of which might contribute to only a small number of cases. These might include people who carry the infection but do not develop symptoms. More puzzlingly, 13 per cent of patients carried genetically similar or identical infections, yet could not be linked to one another either in the hospital, in the community or through health care workers.

‘Fast, accurate genetic sequencing can help to tell you whether a patient brought an infection into the hospital or acquired it on arrival’, says Dr Rory Bowden, Deputy Head of High Throughput Genomics. ‘It also allows hospitals to monitor the effectiveness of their infection control by counting only genetically-related cases, rather than all cases. In addition, investigating genetically-related cases where there is no apparent transmission route between patients might shed new light on the source of currently unexplained infections.’


David W. Eyre, Madeleine L. Cule, Daniel J. Wilson, David Griffiths, Alison Vaughan, Lily O'Connor, Camilla L.C. Ip, Tanya Golubchik, Elizabeth M. Batty, John M. Finney, David H. Wyllie, Xavier Didelot, Paolo Piazza, Rory Bowden, Kate E. Dingle, Rosalind M. Harding, Derrick W. Crook, Mark H. Wilcox, Tim E.A. Peto, and A. Sarah Walker. Diverse sources of C. difficile infection identified on whole-genome sequencing. N Engl J Med 2013; 369:1195-1205.

Most C. diff infections are ‘not hospital spread’, BBC News Online 26 September 2013.