Precise Identification of C difficile Transmission

A recent study uncovers some useful information about C difficile by using whole-genome sequencing on 1250 separate C difficile cases (NEJM 2013; 369: 1195-205).

Between 2007-2011, the authors used genetic sequencing to determine the similarity of C difficile cases.  They were able to successfully sequence 1223 (98%) of the identified cases.  Of these isolates, 71% were from inpatients, 25% from outpatients, and 4% from patients at other hospitals. To determine similarity, they compared single-nucleotide variants (SNVs) between the isolates.  If isolates were related, it was anticipated that there would be 0 to 2 SNVs between transmitted isolates (95% prediction if less than 124 days apart).

This study was from the Oxford University Hospitals which provide all acute care and 90% of hospital services in Oxfordshire, United Kingdom (~600,000 population).


  • Only 35% of cases were genetically related to at least one previous case.
  • Of the 333 (35%) with ≤2 SNVs (consistent with transmission), and 126 (38%) had close hospital contact with another patient, 120 (36%) had no hospital or community contact with another patient.
  • 13% of cases were genetically related (≤2 SNVs) but without any evidence of plausible contact.
  • 45% of C difficile cases were genetically distinct (>10 SNVs from any previous case) from all previous cases.  This indicates that the source of the infection was not from another symptomatic case; most likely these cases were acquired from asymptomatic persons or an environmental reservoir.
  • There were reductions in the rate of C difficile infection during the 4-year study.  The authors relate this to changes in antibiotic prescribing behavior, specifically the restriction of fluoroquinolones and cephalosporins.

Aspects of the setting may limit some of the conclusions.  For example, the study was conducted in a nonoutbreak setting and the hospitals had established measures to limit transmission from symptomatic patients.  These included the following:

  • isolation of patients with suspected C difficile
  • daily hypochlorite disinfection
  • monitoring of compliance

These measures will decrease nosocomial transmission.  However, at the same time, some of the genetically distinct cases could still have been acquired in the hospital setting from asymptomatic hospital sources. While the authors concede a number of limitations, this study is quite helpful in understanding the role of hospitals in controlling C difficile infection.

Take-home message: the fact that only 35% of cases were related to other symptomatic cases indicates that hospital control measures by themselves will not be effective.  The most important aspect in reducing C difficile infection will be optimizing antibiotic usage.

Related news media: On the same day of that I read this study, there was an article in the Atlanta Journal Constitution (“PLEASE wash your hands … Please.”) which describes a first-hand account of C difficile infection which contributed to a slow recovery from intestinal surgery.  The article contained some questionable assertions including that C difficile was “impossible to eradicate” and that her immune system trapped C difficile in peritoneal abscesses which required drainage (these abscesses were more likely related to the initial operation). However, the article is a stark reminder that many hospital staff do not follow basic hand washing recommendations.  The article is really bad PR for the named hospital.

Related blog posts: