Objectives: It has been suggested that colonization with C. difficile protects from infection. Nevertheless, the association between carriage of toxinogenic strains and ensuing C. difficile infections (CDIs) has not been studied.
Methods: We searched PubMed and EMBASE databases up to 20 June 2014, using the term "difficile". Our primary outcomes of interest included the prevalence of isolation of toxinogenic C. difficile or its toxins from asymptomatic patients on hospital admission through stool or rectal swab testing and the risk of ensuing infection among colonized and noncolonized patients. Data on previous hospitalization, antibiotic, and proton pump inhibitor (PPI) use and prior CDIs among colonized and noncolonized patients were also extracted.
Results: Nineteen out of 26,081 studies on 8,725 patients were included. The pooled prevalence of toxinogenic C. difficile colonization was 8.1% (95% confidence interval (CI) 5.7-11.1%), with an increasing trend over time (P=0.003), and 10.0% (95% CI 7.1-13.4%) among North American studies. Patients colonized upon hospital admission had a 5.9 times higher risk of subsequent CDIs compared with noncolonized patients (relative risk (RR) 5.86; 95% CI 4.21-8.16). The risk of CDI for colonized patients was 21.8% (95% CI 7.9-40.1%), which was significantly higher than that of noncolonized patients (3.4%; 95% CI 1.5-6.0%; P=0.03), with an attributable risk of 18.4%. History of hospitalization during the previous 3 months was associated with a higher risk of colonization (RR 1.63; 95% CI 1.13-2.34), as opposed to previous antibiotic (RR 1.07; 95% CI 0.75-1.53) and PPI use (RR 1.44; 95% CI 0.94-2.23), as well as history of CDI (RR 1.45; 95% CI 0.66-3.18) that had no impact.
Conclusions: Over 8% of admitted patients are carriers of toxinogenic C. difficile with an almost 6 times higher risk of infection. These findings update current knowledge regarding the contribution of colonization in CDI epidemiology and stress the importance of preventive measures toward colonized patients.