Since the first description of Clostridium difficile as the cause of pseudomembranous colitis in 1978 (1), historically, this infection was known to occur in hospitalized patients after exposure to broad-spectrum antibiotics. More recently, however, both the incidence of C. difficile infection and the conditions under which infection may occur have increased dramatically. In 2006, McDonald et al. (2) reported that the rate of C. difficile-associated disease (CDAD) doubled over an 8-year span. The list of antibiotics known to predispose to infection has expanded, with fluoroquinolones being particularly problematic (3). Moreover, in addition to the classic risk factors of older age, hospital length of stay, and multiple comorbidities (4, 5 and 6), higher risk has been reported in a wider array of patient types including those in the community (7), those taking immunosuppressants (8) and proton pump inhibitors (9). Lastly, epidemic strains of C. difficile, such as NAP1/BI, produce a more potent toxin and have been implicated in specific outbreaks in North America (10).

This shift in the nature of C. difficile infection has similarly affected patients with IBD, and several features of the infection unique to IBD are now appreciated. For example, whereas older studies found that C. difficile was an infrequent cause of a flare of disease (11,12), it may now complicate as many as 40% of hospitalized IBD patients (13) and is likely a common cause of exacerbation of the underlying IBD. In 2007, two studies reported an increase in the incidence of CDAD in both UC and CD (14,15). Interestingly, C. difficile infection in CD doubled over 7 years, matching that of the non-IBD population, but in UC, the rate of CDAD tripled over the same time period. There has been widespread agreement among various studies on the overall rates of infection and the higher rates in UC (16). In CD, risk of CDAD is higher with isolated colonic disease compared to terminal ileal involvement alone (14,16), which suggests that the inflammatory process from IBD itself somehow predisposes to C. difficile infection. IBD-CDAD patients usually do not manifest classic pseudomembranes on colonoscopic examination (14), a feature with important diagnostic implications. Unlike in the non-IBD population, CDAD in IBD is more often community acquired rather than nosocomial. In one study of hospitalized IBD patients, 67% tested positive within 48 hours of admission (15), an insufficient time for infection to have been acquired nosocomially. Not surprisingly, CDAD complicating IBD results in longer hospital stay, increased cost, and higher mortality (13,16). Further studies are needed to determine whether the risk of infection is related to severity of IBD, use of specific immunosuppressants, or other factors.

In the non-IBD and IBD patients alike, diagnosis of C. difficile infection remains a challenge, notwithstanding 30 years of clinical experience with this organism (17). Any examination of the CDAD diagnostic considerations in the IBD population must also include a broader discussion of the general approach to the diagnosis of CDAD and the controversies therein.

There is no consensus on the ideal diagnostic strategy for C. difficile infection. The most widely used test is the enzyme immunoassay (EIA) to detect toxins A and B in stool. EIA and other rapid, low-cost tests have largely eliminated the availability of confirmatory testing such as the cell cytotoxicity assay or C. difficile culture. However, EIA performance characteristics vary widely, with reported sensitivity as low as 36% (18), but typically in the range of 75% to 95% (19). Local variations in laboratory technique, degree of toxin production by certain C. difficile strains, and differences in CDAD prevalence are some of the factors that can influence EIA test characteristics. To date, no studies examining the sensitivity, specificity, or negative and positive predictive values of EIAs or other tests for C. difficile have been conducted in an IBD population. Thus, both interpreting the meaning of a positive EIA test and deciding how often to repeat testing in the IBD patient remain significant challenges.

As with all diagnostic testing, interpretation of C. difficile test results must take into account the pretest probability for disease, which is most troublesome in the IBD patient because the symptoms of infection are identical to those of an IBD flare. Deciding whether diarrhea is caused by C. difficile infection or an exacerbation of underlying IBD or both lies at the heart of what makes the IBD-CDAD patient difficult to manage. In the non-IBD patient, the presence of gross pseudomembranes seen on endoscopic examination often serves as a confirmatory test when routine testing leads to diagnostic uncertainty. Unfortunately, pseudomembranes are often absent in the C. difficile-infected IBD patient and, thus, are not a reliable sign of true infection (14). The problem is compounded further by the frequent practice of repeating C. difficile testing routinely, which has been advocated as a means to overcome a perceived lack of sensitivity (20). This approach erroneously assumes that testing has 100% positive predictive value for true disease and has been further undermined by studies showing that only about 2% of repeat EIA tests result in conversion from a negative to a positive test when performed within 7 days of the initial test (21,22). A similar lack of value of repeat testing was found using the cell cytotoxicity assay (23), which adds to the evidence that routine multiple testing may now be a larger problem than underdiagnosis (21).

In light of the many problems inherent in C. difficile testing, the following approach to diagnosis of CDAD in the IBD patient is suggested. An initial C. difficile test should be performed in most IBD patients presenting with diarrhea and presumed exacerbation of disease. Commonly, rapid tests with reasonable sensitivity, such as EIA, will be available. Initial results tend to have high positive and negative predictive values (24). Hence, a positive test should prompt antibiotic treatment and a negative result should be interpreted as the absence of infection. Following an initial negative test, clinical suspicion for infection should be re-evaluated on a daily basis. Rather than perform routine multiple or daily C. difficile testing, it is recommended that repeat testing be performed only if the clinical suspicion for infection remains high and not sooner than 48 hours after the initial negative test (21). In time, newer diagnostic methods, such as polymerase chain reaction (PCR) or a two-step algorithm that combines a high-sensitivity common antigen assay with a confirmatory assay for the presence of toxin, may supplant the current popularity of EIA (24,25

Stay updated, free articles. Join our Telegram channel

Join