Health care–associated Clostridium difficile infection correlates with different host and pathogen variables than does health care–associated C. difficile colonization, according to a report in the Nov. 3 issue of the New England Journal of Medicine.
These findings "add to the understanding of C. difficile infection and colonization, and have implications for prevention and therapy," said Dr. Vivian G. Loo of McGill University Health Centre, Montreal, and her associates.
Patients who develop C. difficile infection during hospitalization are likely to be older, to have a history of recent antibiotic use, to be taking proton pump inhibitors, and to carry the NAP1 strain of the organism.
In contrast, patients who show gut colonization with C. difficile but no infection during hospitalization are likely to have been hospitalized during the preceding 2 months, to have a recent history of chemotherapy, to be taking H2 blockers or PPIs, and to carry antibodies against toxin B.
The link to previous hospitalization "suggests previous exposure to C. difficile and possibly the subsequent development of immunity," whereas the link to chemotherapy, H2 blockers, and PPIs suggests that disruption of the bowel flora allowed for C. difficile colonization, Dr. Loo and her associates said.
Dr. Loo and associates examined host risk factors for C. difficile in 4,143 consecutive adults who were admitted to six university-affiliated hospitals in Canada during a recent 6-month period. The one hepatobiliary, eight general medicine, and five general surgery units from which patients were recruited for the study were chosen because they historically had either very high or very low rates of C. difficile infection.
The study subjects provided rectal swabs or stool samples cultured for toxigenic C. difficile upon admission, once a week during their hospitalization, at the onset of diarrhea (in the cases in which diarrhea developed), and at hospital discharge. They also were contacted 60 days after hospital discharge to determine whether diarrhea developed during that time.
For the assessment of antibody levels, subjects also provided serum samples at admission.
A total of 184 patients (4.4%) had asymptomatic colonization at admission. Another 117 (2.8%) developed the infection during hospitalization, and 123 (3.0%) developed asymptomatic C. difficile colonization during hospitalization.
The incidence of health care–associated C. difficile infection was 28.1 cases per 10,000 patient-days, and that of health care–associated C. difficile colonization was 29.5 per 10,000 patient-days.
There were 14 deaths within 60 days of C. difficile infection diagnosis, for a crude mortality of 12.0%. Death was attributed directly to C. difficile infection in two cases (1.7%), and the infection was considered a contributing cause of death in another six cases (5.1%).
C. difficile infection necessitated intensive care in one patient, but no patients required colectomy. Of the 117 infected patients, 29 developed at least one recurrence during the 60-day follow-up.
Older age was particularly significantly associated with C. difficile infection. For every year of age older than 18, the risk of health care–associated C. difficile infection rose by approximately 2%, the investigators said (N. Engl. J. Med. 2011;365:1693-703).
This study was supported by the Consortium de Recherche sur le Clostridium difficile. Dr. Loo reported receiving consulting fees from Merck, and her associates reported ties to Pfizer, Spartan Biosciences, Cubist Pharmaceuticals, Cepheid, and Genome Canada.