Test Catalog


Laboratory Diagnosis  of Clostridium difficile Infection

September 2013

Clostridium difficile is an anaerobic spore-forming gram-positive bacillus and is the most important cause of antibiotic-associated diarrhea and pseudomembranous colitis. Clostridium difficile is not part of the normal stool flora but may be present in many normal infants and up to 5% to 10% of asymptomatic adults. Both toxigenic (enterotoxin-producing) and nontoxigenic strains exist. The development of C difficile infection (CDI) requires acquisition of a toxigenic strain of C difficile and alteration of the normal bowel flora allowing the organism to proliferate. Although CDI is most often associated with antimicrobial use in a hospital or other health care setting, an increasing number of infections are “community acquired.” Proton pump inhibitor use has also been recognized as a predisposing factor for CDI.

Pathogenic C difficile strains produce toxins A and B and possibly a recently recognized binary toxin. Outbreaks of severe CDI in recent years have been related to a hypervirulent strain of C difficile known as North American pulsed-field type 1 (NAP1), also known as ribotype 027. Many 027 strains contain binary toxin and are also resistant to quinolone antimicrobials. It is now recognized that strains other than 027 may cause severe disease and that many patients infected with the 027 strain do not have severe CDI.

The pathogenicity locus (PaLoc region) of the C difficile genome contains the genes for toxins A and B, which are designated tcdA and tcdB, and the tcdC gene, which actually modulates or downregulates toxin production. There are nucleic acid base pair deletions and a mutation at position 117 in the tcdC gene that are thought to be associated with increased toxin production.1

The laboratory diagnosis of CDI depends upon several important principles. Only those patients who actually have diarrhea—defined as having 3 or more unformed stools within 24 hours—should be tested. The specimen should be loose enough to take the shape of the container. There is no need to test more than 1 specimen when using a sensitive and specific method.2 After a positive result, tests should not be repeated unless the patient does not improve on treatment after 5 to 7 days or relapses after improvement. It is not necessary to perform a “test of cure.”

Culture of stool on cycloserine-cefoxitin-fructose agar (CCFA) and toxin testing of the isolates recovered is the gold standard method for detecting C difficile. This is a sensitive and specific method but is too complex and time-consuming for routine use. The cell culture cytotoxicity neutralization assay or CCNA was widely used in the past but is less sensitive than toxigenic culture and is also complex and time-consuming. It is also dependent on subjective interpretation of the result.

There is a commercially available immunoassay for glutamate dehydrogenase (GDH), the common antigen found in all strains of C difficile. This is a sensitive test but it is not specific for toxigenic strains. Numerous enzyme immunoassays (EIA) for toxins A and B are available and 1 version is combined with the GDH test. However, the toxin EIAs are only 40% to 70% as sensitive as toxigenic culture and are not suitable as definitive tests for the diagnosis of CDI.3

Polymerase chain reaction (PCR) assays have recently  been developed that detect genetic markers in the PaLoc region of C difficile, including tcdA, and/or tcdB or tcdC and/or the binary toxin and the 117 mutation. A number of these are available commercially, including the BD GeneOhm Cdiff assay, Cepheid Xpert, Prodesse ProGastro CD, and Meridian Illumigene assay. The Clinical Microbiology Laboratory at Mayo Clinic developed a real-time PCR method using fluorescence resonance energy transfer probes which will detect the presence of tcdC. PCR assays are 80% to 90% as sensitive as toxigenic culture.3 In a recent Mayo Clinic study, a number of the commercially available PCR assays were compared with a Mayo Clinic laboratory-developed assay for detection of toxigenic C difficile. In detection of 36 true-positive and 310 true-negative results, the assays were essentially equivalent.4 Other studies have reported similar results.

Some investigators have advocated an algorithmic approach to the testing for toxigenic C difficile5  It starts with screening for the GDH common antigen using a single test or a version that also includes a toxin EIA. (Figure) If the GDH antigen test is negative, the result can be reported as negative for C difficile. If the GDH test is positive, the result of the toxin EIA test must be considered, which, if positive, can result in a presumptive positive report. If the toxin EIA is negative, a more sensitive confirmatory test must be performed—either PCR or CCNA. If the confirmatory test is positive, the diagnosis of CDI is confirmed, or, if negative, CDI is not present and presumably only a nontoxigenic strain of C difficile was detected in the specimen. The decision to use the algorithmic approach or directly test all specimens by PCR depends upon the needs and resources of the individual laboratory. At Mayo Clinic, all specimens are directly tested using a laboratory-developed PCR method, which is efficient and economical in a high test volume setting.

With the introduction of a commercial PCR (Xpert C. difficile/Epi) to identify the 027 strain of C difficile, the clinical laboratory must decide if it is important to identify and report this particular strain. This assay is 87% sensitive for the presumptive identification of the 027 strain but is not a typing method; it presumptively detects 027 by a PCR that detects tcdB, the 117 mutation in tcdC, and the binary toxin. It is not approved by the Food and Drug Administration (FDA) to guide therapy but is FDA approved for epidemiologic use. As noted above, the significance of the 027 strain as the only or primary cause of severe outbreaks of CDI is now somewhat diminished. Moreover, the spectrum of disease caused by this strain varies from mild to severe, and reporting of the presumptive detection of the 027 strain might actually mislead clinicians to employ unnecessary antimicrobials or infection control measures. Guidelines from the Society for Hospital Epidemiology of America and the Infectious Diseases Society of America indicate that treatment and infection control should be based on disease severity and not on identification of a particular strain type.6 In addition, in an outbreak situation it is preferable to perform actual strain typing rather than presumptive identification of only 1 type. For these reasons, it does not seem necessary or desirable to presumptively report the presence of the 027 strain.

Stool Specimen Algorithm

In summary, CDI may be caused by any of numerous toxin-producing strains. Clostridium difficile culture with isolate toxin testing is the gold standard for laboratory detection but is not suitable for routine use. PCR is the most efficient direct method for laboratory diagnosis of CDI, although some laboratories may prefer the algorithmic approach using PCR as a confirmatory test for the presence of toxin. Mayo Clinic standard of care for diagnosis is direct testing by PCR because of the low cost of the laboratory-developed test and the efficiency of its use in a high test volume setting.

Authored by Rosenblatt JE


  1. Spigaglia P, Mastrantonio P: Molecular analysis of the pathogenicity locus and polymorphism in the putative negative regulator of toxin production (TcdC) among Clostridium difficile clinical isolates. J Clin Microbiol 2002 Sep;40(9):3470-3475
  2. Aichinger E, Schleck CD, Harmsen WS, et al: Nonutility of repeat laboratory testing for detection of Clostridium difficile by use of PCR or enzyme immunoassay. J Clin Microbiol 2008 Nov;46(11):3795-3797
  3. Burnham CA, Carroll, KC: Diagnosis of Clostridium difficile Infection: an Ongoing Conundrum for Clinicians and for Clinical Laboratories. Clin Microbiol Rev 2013 July;26(3):604-630
  4. Karre T, Sloan L, Patel R, et al: Comparison of two commercial molecular assays to a laboratory-developed molecular assay for diagnosis of Clostridium difficile infection. J Clin Microbiol 2011 Feb;49(2):725-727
  5. Kufelnicka AM, Kirn TJ: Effective utilization of evolving methods for the laboratory diagnosis of Clostridium difficile infection. Clin Infect Dis 2011 Jun;52:1451-1457
  6. Cohen S. Gerding D, Johnson S, et al: Clinical practice guidelines for Clostridium difficle infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect Control Hosp Epidemiol 2010;31:431-455