Clostridium difficile--beyond the usual suspects.

One of the central concepts guiding efforts to prevent Clostridium difficile infection has been that symptomatic patients in hospitals are the major source of transmission.1 Infected patients with diarrhea shed large numbers of spores, contaminating their skin, clothing, and surrounding surfaces. The basic measures that are used to prevent transmission include placement of infected patients under contact precautions until diarrhea resolves and the disinfection of surfaces and equipment with sporicidal products such as sodium hypochlorite. Unfortunately, these measures have often proved to be ineffective during the past decade, when infection rates have risen dramatically in association with large outbreaks of the 027/BI/ NAP1 strain of C. difficile.2 Control of outbreaks has frequently required sequential implementation of multiple control measures, including antimicrobial interventions.3 Even when outbreaks subside, many hospitals continue to struggle with high endemic infection rates. The difficulty in controlling C. difficile has raised several important questions about traditional models of transmission. First, are we missing important sources of transmission? For example, asymptomatic carriers of toxin-producing strains of C. difficile outnumber infected patients,4 but their role in transmission has been uncertain. Curry et al.5 recently examined C. difficile transmission in a hospital with well-thought-out practices for the prevention of transmission from symptomatic patients. On the basis of molecular typing, incident infections were as frequently linked to asymptomatic carriers as to symptomatic patients (30% and 29%, respectively). Second, are we missing novel routes of dissemination not addressed by current control strategies? It has been proposed, for example, that airborne dispersal of spores could contribute to transmission. Finally, are we underestimating the proportion of cases acquired outside the hospital? No hospital is an island: increasingly, patients enter and leave hospitals colonized or infected with C. difficile.4,6 In the United States, as many as 75% of infections have their onset in long-term care facilities or the community.6 The administration of antimicrobial agents during a previous hospitalization is a predisposing factor in many of these cases, because antimicrobial-induced alteration of the intestinal microbiota increases susceptibility to infection for several weeks. However, the time from the acquisition of colonization to the onset of C. difficile–associated illness is short,6 which suggests that C. difficile is frequently acquired outside the hospital, even in cases occurring after a recent hospitalization. In addition to long-term care and outpatient health care settings, other potential reservoirs outside the hospital include colonized infants, food, and animals.7 In this issue of the Journal, Eyre et al.8 present the results of an impressive 3.6-year study in which they used whole-genome sequencing, a highly discriminatory typing method, to study the epidemiology of C. difficile infection in Oxfordshire, United Kingdom. Of 1223 isolates sequenced, 71% were obtained from inpatients, 25% from outpatients, and 4% from patients at other hospitals. Only 35% of isolates obtained from patients with C. difficile infection were genetically linked. Moreover, only 38% and 54% of genetically linked cases shared ward-based and hospital-wide contacts, respectively. Remarkably, 45% of isolates were genetically distinct from all others, suggesting diverse sources of acqui sition. The results of this study challenge the tradi-