Clostridium Difficile Infection (CDI)

The Public Health Laboratory Network have developed a standard case definition for the diagnosis of diseases which are notifiable in Australia. This page contains the laboratory case definition for Clostridium difficile infection (CDI).

Page last updated: 09 February 2016

PDF printable version of Clostridium Difficile Infection (CDI) - Laboratory Case Definition (PDF 178 KB)

Version: 1.0
Authorisation: PHLN
Consensus Date: May 2016


1 PHLN Summary Laboratory Definition

1.1 Condition:

Clostridium difficile infection (CDI).

A diagnosis of CDI implies:

  1. laboratory detection of C. difficile toxins and/or toxigenic C. difficile in faeces, rectal swab or bowel contents PLUS
  2. relevant clinical manifestations: diarrhoea (usually defined as 3 or more loose stools in a 24 hour period) or, less commonly, ileus, toxic megacolon or pseudomembranous colitis (identified by colonoscopy).

1.1.1 Definitive laboratory criteria

Direct identification of preformed C. difficile toxin(s) in an unformed (diarrhoeal) stool sample (i.e. one that takes the shape of the container).

1.1.2 Suggestive laboratory criteria

  1. Direct detection of gene(s) encoding C. difficile toxin production - tcdA and/or tcdB - in an unformed (diarrhoeal) stool sample or in bowel tissue OR
  2. Isolation, from an unformed (diarrhoeal) stool sample or bowel tissue, of C. difficile, which EITHER
    • carries one or more C. difficile toxin-related gene(s) carried on the pathogenicity locus (PaLoc) and/or the
      C. difficile transferase locus (CdtLoc) OR
    • produces toxin A and/or B in vitro.

2 Introduction

2.1 The organism and its toxins

C. difficile is a Gram positive spore-forming bacterium, which was shown to be the cause of pseudomembranous colitis – a condition often associated with use of the (then) new antibiotic, clindamycin - in 1978 1. It is widely distributed in the environment and faecal flora of humans and animals. With rare exceptions, colonisation and infection are limited to the gastrointestinal tract.

Local and systemic symptoms are due to the effects of one or both of two toxins, encoded by genes carried on a pathogenicity locus (PaLoc), which is absent from non-toxigenic strains: toxin A (TcdA or “enterotoxin”, encoded by tcdA) and toxin B (TcdB or “cytotoxin”, encoded by tcdB). Although both are cytotoxic, TcdB is 100-1000-fold more potent; TcdA causes fluid accumulation, similar to cholera toxin, in the rabbit ileal loop 2. The PaLoc also carries tcdC and tcdR, which are, respectively, negative and positive regulators of toxin production, and tcdE, which is involved in toxin release 3. A minority (3-12%) of C. difficile strains infecting humans produce binary toxin or C. difficile transferase (CDT), which is a classic two component toxin with binding (CDTb) and enzyme/toxic (CDTa) subunits encoded by cdtA and cdtB, respectively; they are carried on the CDT locus (CdtLoc) along with a regulatory gene, cdtR. CDT production is generally associated with increased virulence of C. difficile, although the mechanism is uncertain; it is cytotoxic in vitro and believed to enhance adherence of C. difficile to the intestinal epithelium 3.

Various combinations of toxin genes occur in different C. difficile strains; the majority produce both TcdA and TcdB or TcdB alone, but not binary toxin. By contrast, ribotype (RT) 027 (refer below) and related strains, whichoriginate from animals, produce TcdA, TcdB and binary toxin 2,4.

2.2 Infection and colonisation

Symptoms of CDIrange in severity from a few loose stools to severe bloody diarrhoea with colonic ulceration, pseudomembrane formation, fever, leukocytosis and systemic toxicity, sometimes further complicated by life-threatening toxic megacolon or perforation and peritonitis, requiring emergency colectomy. The average attributable mortality is reported to be ~5%, but it varies widely according to patient characteristics – particularly age – and the C. difficile strain involved. Recurrent CDI – i.e. diarrhoea recurring after improvement of symptoms, within 8 weeks of the onset of the original incident - occurs in around 20% of cases, on average, but this also varies. Recurrence can be due to relapse of the original infection or reinfection with a different strain, particularly in individuals with ongoing risk factors e.g. reduced immune competence or a continuing or repeated course of antibiotic therapy 5.

Asymptomatic carriage of C. difficile (including toxigenic strains) is not uncommon, particularly in young children or, in adults, after a period of hospital admission and among residents of long term care facilities; carriage often persists for several weeks after symptomatic recovery from an episode of CDI 6. Up to 50-60% of healthy infants are colonised with C. difficile in the first year of life; this is most likely if establishment of the normal gut microbiome is delayed or modified because of Caesarean section delivery, exposure to antibiotics or artificial feeding. Infants are resistant to the effect of C. difficile toxins, apparently because they lack relevant epithelial receptors, and rarely develop symptomatic CDI 7.

top of page

Colonisation rates decline in the second year of life as the gut microbiome becomes more complex and immune competence matures and is uncommon in healthy older children and adults (3-5%). Subsequent exposure can result in asymptomatic carriage or CDI, in individuals who are susceptible because of waning immunity - due to increasing age or immunosuppression associated with chronic illness or chemotherapy – and/or disruption of gut microbiota due to antibiotic therapy - especially with broad spectrum agents, such as clindamycin and third or fourth generation cephalosporins or fluorquinolones. Gastrointestinal surgery, inflammatory bowel disease, acid suppressive therapy, especially with protein pump inhibitors, and excessive use of laxatives or enemas are all associated with increased risk of CDI in some studies 6,8

2.3 Transmission

C. difficile is spread via the faecal-oral route by ingestion of spores, which are resistant to drying, heat and many disinfectants and so persist in the environment, particularly in the vicinity of individuals with diarrhoea due to C. difficile. CDI was once thought to be almost exclusively healthcare-associated, with spores spreading, via a contaminated environment or on the hands of healthcare workers, from a patient with diarrhoea to other susceptible patients, who would subsequently develop similar symptoms. It is now recognised that at least one third of cases is community-acquired 9,10 and the organism can spread from asymptomatic carriers, including colonised infants who may be an important reservoir of community-acquired infection 10. The strains they carry are often toxigenic and similar to those that cause CDI in susceptible adults 7. Recently, whole genome sequencing has demonstrated that C. difficile strains causing CDI – even when apparently hospital-acquired (i.e. those with onset of symptoms ≥48 hours after admission) - often cannot be genetically linked to another symptomatic patient or other hospital or community source, indicating that sources are diverse and often unrecognised.11 Some apparently hospital-acquired infections presumably result from an existing carrier being exposed, after admission to hospital, to risk factors such as antibiotic therapy or bowel surgery.

C. difficile can colonise and cause diarrhoea in food-producing, especially young, animals; meat and poultry at the point of sale are not infrequently contaminated with C. difficile; vegetables can be contaminated by animal manure and ready-to eat food can be contaminated during preparation. Even low-level contamination with spores may resist short periods of cooking. Food-borne transmission is probably a significant source of community-acquired CDI, but is difficult to prove in individual cases 12,13. The distribution of C. difficile strains (RTs) that colonise and infect animals differs from that of those that are most prevalent in humans, but there is overlap and direct transmission between animals and humans e.g. of RT 078 has been postulated. 14

2.4 Pathogenesis

After ingestion, spores survive the acid environment of the stomach, germinate on exposure to bile in the small intestine and proliferate in, and colonise or infect, the colon 4. In favourable circumstances – disruption of the gut microbiota and/or reduced immunity - C. difficile penetrates the mucus layer and adheres to the colonic epithelium. Toxin production increases during the stationary growth phase; cellular intoxication causes disruption of the cytoskeleton and tight junctions, increased epithelial permeability, fluid accumulation and an intense local and systemic inflammatory response. Toxin B (and, to a lesser extent, toxin A) alone can produce these effects but toxins A and B usually act together. Disease severity varies with different C. difficile strains; binary toxin production and differences in germination, sporulation and biofilm production between strains can contribute to virulence 4.

2.5 Changing CDI epidemiology

Until about 10 years ago, CDI was regarded as a relatively uncommon cause of diarrhoea, occurring mainly in elderly hospitalised patients with multiple co-morbidities. Outbreaks were reported, but uncommon. In 2004, a report from Quebec 15, drew attention to a dramatic increase in the incidence, severity and mortality of CDI in the region, including in younger (<65 yrs) age-groups. Similar increases had occurred in some parts of the USA 16 and, in both countries, were associated with emergence of a previously rare, fluoroquinolone-resistant C.difficile strain, 16,17 identified by pulse-filed gel electrophoresis (PFGE) as NAP1, restriction endonuclease analysis (REA) as BI, ribotyping as RT 027 and toxinotyping as toxinotype III.

This strain (now generally referred as 027) is characterised by production of binary toxin/CDT, and an 18-bp deletion and a deletion at position 117 (tcdCD117) in tcdC, which affect the negative regulatory function of TcdC and lead to increased toxin production. 6 Most, but not all, studies show that CDI due to RT 027 and other strains with similar toxin profiles (mostly belonging to clades 2 and 5) are associated with higher complication rates and mortality than strains with more typical profiles 18.

C. difficile RT 027 rapidly spread to the UK where widespread hospital outbreaks occurred, and subsequently to several European countries 19. In most countries, in which major hospital CDI outbreaks due to RT 027 have occurred, it has become the predominant strain. However, after the incidence and mortality from CDI in the UK reached a peak in 2006-8, major national infection control programs were implemented throughout the UK, which resulted in sharp falls in CDI rates.

C. difficile RT 027 was also associated with an increase in the incidence and recognition of community-acquired infections, not only due to RT 027, but also less virulent strains. In some European countries, another virulent RT, 078, which is commonly associated with animals, especially pigs, has been increasingly recognised causing community-acquired infection in humans 14. Typically, community-acquired infections occur in younger age-groups than hospital-acquired CDI, often in the absence of obvious risk factors; because those affected generally have fewer comorbidities, the overall mortality is lower than that of hospital-acquired infection, but up to 40% of patients with community-acquired CDI require admission to hospital and rapidly progressive, severe disease can occur in previously healthy young people 5, 10.

There has been relatively little spread of RT 027 in Australia; only a few individual imported 20 and hospital-acquired 21 cases have been reported. However, limited strain typing of C. difficile isolates has identified the emergence of several previously unknown RTs with increased virulence, toxin profiles similar to and belonging to the same clade as, RT 027, such as RT 244 22 and RT 251. In addition, a multitude of clade 5 strains has been detected in Australian food animals but not RT 078. Prospective national surveillance of public hospital-diagnosed cases of CDI, indicate that the incidence of CDI, including community-acquired disease, is increasing 23. However, changes in testing criteria and laboratory methods have probably contributed to some but not all this apparent increase.

top of page

2.6 Diagnosis

Indications for and methods of laboratory diagnosis of CDI have changed significantly in the last 10 years. Previously, it was based on detection of preformed C. difficile toxin(s) in, or culture of toxigenic C. difficile from, faeces; both methods are slow, relatively difficult and usually done only in response to a specific request on specimens from hospital patients. Subsequently, immunoassays for detection of toxins A and/or B and glutamate dehydrogenase (GDH) – a C. difficile “common antigen” - and nucleic acid amplification tests (NAATs) became commercially available and widely used. However, there is wide variation in their sensitivities, specificities, predictive values and costs 2,24,25 and in their use in different Australian laboratories 26, which affect the reliability of diagnosis and national CDI surveillance data. One source of variability and confusion is the fact that different types of test measure different things and there are two “gold standards” against which newer methods are compared. While rapid diagnosis is important to allow prompt implementation of appropriate treatment and infection control measures, interpretation of results requires an understanding of the strengths and limitations of individual tests.

a) Tests for preformed C. difficile toxin in faeces: The cellular cytotoxicity neutralisation assay (CCNA) was once regarded as the “gold standard” for diagnosis of CDI; however, although highly specific, its sensitivity is only 75-85% that of than toxigenic culture 27. CCNA requires cell culture capability, the turnaround time is 24-48 hours and it is now rarely performed. Preformed C. difficile toxins A and/or B can also be identified in stool by immunoassays. The best of the commercial assays are also specific (>95%) but less sensitive than CCNA 24. Early assays that tested only for toxin A – because a) it was thought that all toxigenic strains produced it and b) toxin A is more immunogenic than toxin B - are now recognised as unsuitable since a significant minority of C. difficile strains in Australia produce only toxin B (e.g. the major ribotype found in Asia, RT 017, is toxin A negative); most assays are now designed to detect both toxins A and B.

b) Tests for toxigenic C. difficile in faeces. Toxigenic culture is an alternative “gold standard” for diagnosis of CDI, which is more sensitive than CCNA, but less specific for CDI, since it detects the genetic potential for toxin production, but not necessarily actual toxin production in vivo. The turnaround time for culture is 24-48 hours and a confirmatory test for toxigenicity is required. While culture is impractical for routine diagnosis, it is required for strain typing and antibiotic susceptibility testing. Toxigenic C. difficile can also be detected in faecal specimens by rapid methods, such as a combination of GDH immunoassay plus supplementary immunoassay to detect preformed toxin or by direct NAAT to detect toxin genes.

The presence of toxigenic C. difficile is a necessary, but not sufficient, condition for diagnosis of CDI – diarrhoea in a colonised patient can be due to another cause. Although detection of preformed toxin results in fewer positive tests, it may correlate better with clinical findings. A recent prospective, multicentre study evaluated toxigenic culture, CCNA and rapid methods (alone and in combination) based on clinical measures of CDI severity including routine blood tests, length of hospital stay and 30-day mortality 28. Results, in >6,500 faecal specimens, were grouped according to test results: group 1, CCNA (and toxigenic culture) positive (7% of specimens); group 2, toxigenic culture positive, CCNA negative (3%); group 3 both CCNA and toxigenic culture negative (90%). Mortality was significantly higher in group 1 (16.6%) than in either group 2 (9.7%; p=0.002) or 3 (8.6%; p=<0.0001) and not significantly different between groups 2 and 3 (p=0.53). The authors concluded that patients from whom group 2 isolates originated were C. difficile excretors, who probably did not have CDI but were potential sources of cross-infection. None of the rapid tests, in isolation, gave acceptable predictive values but the combinations that correlated best with gold standard methods were:

  • Toxigenic culture: GDH EIA (Techlab C diff Chek-60 glutamate dehydrogenase enzyme immunoassay) plus PCR (GeneXpert, Cepheid): sensitivity, 92%; specificity, 98%; PPV, 81%; NPV 99%;
  • CCNA: Toxin A/B EIA (Techlab C difficile Tox A/B II toxin enzyme immunoassay) plus PCR or GDH EIA plus toxin A/B EIA: sensitivity, 82%; specificity, 99.5%; PPV 92%; NPV 99%.

It is now generally recommended that the diagnosis of CDI be based on a combination of rapid tests performed sequentially or in parallel 2, 8, 26, 29 and supplemented by culture in at least some larger laboratories so that strain typing can be performed for surveillance. The choice will depend not only on predictive values and clinical validation, but also on cost, turnaround times and laboratory workflow.

2.7. Strain typing and surveillance

PCR ribotyping is the most common method used for C. difficile strain typing in a small number of Australian laboratories. Other methods include toxinotyping - a restriction fragment length polymorphism assay for variations in the PaLoc of toxigenic C. difficile strains; multilocus sequence typing; pulse filed gel electrophoresis and restriction endonuclease analysis 25.

3 Laboratory Diagnosis/Tests

Several clinical practice guidelines for diagnosis of CDI have been published 26,29,30.

3.1. Testing criteria

  • Tests for toxigenic C. difficile should only be performed on unformed stool specimens (or gut contents from patients with diarrhoea), unless ileus is suspected.
  • All adults and children over 2 years, who have been hospitalized for ≥48 hours and develop diarrhoea (≥3 unformed stools on a 24-hour period) should be tested for CDI.
  • All adults and children over 2 years, in whom diarrhoea has persisted for >48 hours and no other enteropathogen has been identified should be tested for CDI.
  • Repeat testing of faecal specimens during the same episode of diarrhea is not recommended a) within 4 weeks of a positive test (response to treatment is determined by clinical criteria) or b) following a negative test – unless CDI is strongly suspected and a more sensitive method (e.g. NAAT) is used after a negative immunoassay.
  • Tests for C. difficile in children <2 years old should be performed in consultation with a paediatrician.
top of page

3.2 Cellular cytotoxicity neutralisation assay (CCNA) 27

Cell monolayers (usually Vero or Hep2 cells) are cultured in the presence of a faecal (or culture) filtrate, with and without neutralising antitoxin antibodies (usually C. sordellii  antiserum) and examined at 24 and 48 hours for cytopathic effect (cell rounding) that is absent in monolayers with neutralising antibody. This assay is now rarely used either for primary diagnosis or confirmation of in vitro toxin production by C. difficile isolates.

3.2.1. Suitable specimen

Unformed stool or culture filtrate. Specimens should be tested as soon as possible after receipt or stored at 4oC for no more than 48 hours. Prolonged storage and repeated freezing and thawing will result in degradation of toxin.

3.2.2 Test sensitivity

Compared with toxigenic culture 75-85% 27.

3.2.3. Test specificity

Compared with toxigenic culture, theoretically, 100%. However, note that neither method is standardised, both require considerable expertise and results vary between laboratories. Rarely, CCNA is positive in culture-negative specimens, which is most likely to indicate insensitive culture methods 27.

3.2.4. Predictive values

In the presence of relevant clinical symptoms, a positive result (neutralisable cytopathic effect) is generally regarded as definitive evidence of CDI, but a negative result does not exclude it. Reduced sensitivity can be due to degradation of toxin in the specimen due to delayed testing or repeated freeze-thawing.

3.2.5. Suitable acceptance criteria.

Cytopathic effect (cell rounding) produced by faecal filtrate in a cell monolayer without antitoxin but not in a monolayer with added antitoxin. Note that a very high concentration of toxin in faecal filtrate can overwhelm the neutralising capacity of antitoxin. If typical cytopathic effect is produced, but not inhibited by antitoxin, the assay should be repeated with dilutions of faecal filtrate.

3.3. C. difficile culture

3.3.1 Suitable specimens

Unformed faeces; rectal swab (e.g., from patient with ileus or toxic megacolon) or intestinal contents. Heat or alcohol pretreatment of the specimenis generally recommended to kill vegetative organisms and minimise overgrowth of other faecal flora, whilstsparing C. difficile spores.

Alcohol shock pretreatment. Equal parts (1mL) of specimen and industrial methylated spirit or absolute alcohol are homgenised in a vortex mixer and the left at room temperature of 1 hour before inoculation.31.

Heat shock pretreatment. The specimen is heated at 80oC for 10 minutes before inoculation.

3.3.2 Media

A variety of media can be used for broth enrichment and/or direct plating. Recent comparisons of media have produced contradictory results 2,31-33. Those most commonly used and generally satisfactory include cefoxitin, cycloserine fructose agar (CCFA), cefoxitin cycloserine egg yolk agar (CCEY), and chromogenic C. difficile agar. Taurocholate or lysozyme is generally added (or included in commercial media) to enhance spore germination; broth enrichment (e.g using CCFB) may 31 or may not 33 increase culture sensitivity. Most media, except chromogenic C. difficile agar, need to be prereduced in anaerobic conditions for 2 hours before inoculation. Media are incubated anaerobically at 35oC. Growth of characteristic colonies is often visible after 24 hours incubation on chromogenic agar, but recovery can be increased by incubation for up to 72 hours.

Recent comparison of prereduced taurocholate-containing CCFA (TCCFA) - as recommended by the Infectious Diseases Society of America IDSA and Society for Healthcare Epidemiology of America SHEA 30 - with chromogenic C. difficile agar (CDIF, BioMerieux) 32,34 showed that the latter was more selective (less faecal flora grown), even without alcohol shock specimen pretreatment, and more sensitive (more specimens positive and more C. difficile colonies on CDIF at 24 hours than on TCCFA at 48 hours), even without prereduction of plates. Several chromogenic media are available; they are more expensive but more sensitive and convenient to use than conventional media.

3.3.3 Test sensitivity

As indicated by comparative studies, there is considerable variability in sensitivity depending on specimen treatment, broth enrichment, solid media used and duration of incubation.

3.3.4 Test specificity

Isolation of C. difficile from faeces indicates at least colonisation; confirmation of toxigenic potential by CCNA, immunoassay or NAAT is required for presumptive or definitive diagnosis of CDI.

3.3.5 Predictive values

Negative culture does not exclude CDI (although if appropriate methods are used it makes the diagnosis unlikely). Most commonly, culture is performed only on specimens in which one or more rapid tests for C. difficile is positive (see below).

top of page

3.3.6 Suitable acceptance criteria

Isolation of colonies with typical morphology (irregular, flat, with ‘ground glass’ appearance on TCCFA) allows presumptive identification of C. difficile. Black colonies typically isolated on CDIF are easily recognized, but not produced by a small minority of C. difficile strains (notably RT 023). Isolates should be subcultured for purity on horse blood agar, on which they produce chartreuse colony fluorescence under UV light. Further confirmatory identification is generally by nucleic acid testing and/or immunoassay for toxins A and B.

3.3.7 Suitable external QC program

No QC program is available for C. difficile culture per se, which is performed by a minority of laboratories.

top of page

3.4. Immunoassays for C. difficile common antigen - glutamate dehydrogenase (GDH)

GDH is a metabolic enzyme, encoded by gluD; it is produced by toxigenic and nontoxigenic C. difficile, and cross-reacts with GDH of C. sordellii. A number of enzyme immunoassays (EIAs) and immunochromatographic tests (ICTs) are commercially available, with similar performance characteristics 2. GDH assays are rapid, sensitive and inexpensive screening tests for CDI, but must be combined with or supplemented by more specific confirmatory tests.

3.4.1 Suitable specimens

Unformed faecal specimen.

3.4.2 Test sensitivity

There is considerable variation in the reported sensitivity (88-100%), which probably reflects differences in methodology between studies, including duration and methods of specimen storage before evaluation and the reference method used 2. They are generally highly sensitive (>95%) for detection of C. difficile.

3.4.3. Test specificity

Specificity is also variable (76-98%), depending on methods and reference standards 2.

3.4.4. Predictive values.

GDH has a very high NPV (98-99%) but relatively poor PPV (61-94%) for CDI 2,28. It is generally recommended that a negative result can be reported as excluding CDI. Specimens with positive results require additional testing.

3.4.5 Suitable acceptance criteria.

Tests should be performed and interpreted according to manufacturer’s instructions.

3.5. Immunoassays for C. difficile toxins

These assays use polyclonal or monoclonal antibodies against toxin A and/or B. Although most commercial assays now detect both toxin A and B they are relatively insensitive for diagnosis of CDI and are not recommended as standalone tests. However, when combined with a sensitive screening test such as a GDH assay, and/or a NAAT, they provide evidence of C. difficile toxin production in vivo (when testing faecal specimens) or in vitro (when testing isolates).

3.5.1 Suitable specimen

Unformed faecal specimens.

3.5.2. Test sensitivity

Reported sensitivities vary with different kits, reference methods used - average values are ~75% and ~50% compared with CCNA and toxigenic culture, respectively) 24,28,29. Sensitivity also varies depending on RT. There is less difference in sensitivity between immunoassays and NAAT when RT027, which produces large amounts of toxin, is present than when other strains, which produce less - or less antigenic – toxin, are present 2.

3.5.3. Test specificity

Specificities also vary with different kits but are generally >90%.

3.5.4 Predictive values

PPV is reported to be unacceptably low - <50% - for most kits, in populations with a (fairly typical) CDI prevalence of <10% 24. However, the clinical significance of this is uncertain. A chart review of >6,000 patients with diarrhoea, in whom C. difficile toxin tests were performed, showed that only one toxin-negative patient had pseudomembranous colitis and none had other severe CDI complications 35.

3.5.5 Suitable acceptance criteria

Tests should be performed and interpreted according to manufacturer’s instructions.

3.6 Combined GDH and toxin assays

Combined GDH and toxin EIAs are also available. They are very rapid, convenient and less expensive than NAAT.

top of page

3.6.1. Test sensitivity and specificity.

The sensitivities of the GDH component are similar to those of stand-alone GDH assays and the toxin EIA components have the same limitations as conventional toxin EIAs. Nevertheless, they have the advantage, over NAATs of a similar sensitivity for detection of C. difficile and the ability to detect in vivo toxin production and thus generally high specificity for CDI 2.

3.6.2 Predictive values.

Specimens that are GDH and toxin negative or GDH and toxin positive have high NPV and PPV for CDI, respectively. Those that are GDH positive but toxin negative should be tested by a more sensitive method for detection of toxin (e.g. CCNA) or toxin genes (NAAT).

3.7. Nucleic acid amplification tests (NAAT)

In-house NAATs for C. difficile, targeting tcdA, tcdB and/or 16S rDNA in conventional PCR reactions using gel electrophoresis for product detection, were first described in the 1990s. Despite difficulties with faecal DNA extraction at the time, they were more sensitive than the CCNA and older culture methods, although some times less specific because of primer cross-reaction with other Clostridium species. Improved DNA extraction methods made these assays more practicable and their sensitivity comparable with, and turnaround time much better than, the more sensitive current culture methods. 36 The first commercial NAAT was licensed in 2009 and since then there has been a plethora of assays using many different (usually real time) platforms and product detection methods. Nearly all assays primarily target tcdB alone or with one or more additional targets e.g. tcdC D117 (to presumptively identify RT 027) and/or CdtLoc (binary toxin); one assay includes a gyrA target to detect fluoroquinolone resistance. Targets of the few assays that do not target tcdB include a conserved region of tcdA (present as a remnant in tcdA-negative/tcdB-positive strains) and a combination of cdtA and tcdCD117. 2, 25 C. difficile is included in several multiplex assays for several viral and bacterial enteropathogens. While this approach is theoretically attractive these assays are expensive and their performance quality and role in routine diagnosis is yet to be established 25.

3.7.1 Test sensitivity and specificity

When compared with toxigenic culture as gold standard, most NAATs have high analytical sensitivity and specificity (generally <90%); most are fully automated (including extraction) and rapid to perform (1-2 hours) but relatively expensive, compared with immunoassays.2, 25 However, they have limitations: some toxin A-negative strains(e.g. RT033, among others) are missed by assays that target only tcdA 37; RT 244 – which has recently emerged, in Australia 22 – has the tcdC D117 and is misidentified as 027 by assays targeting this deletion 38. Therefore it is important to recognise that the performance of different assays may vary depending on geographic and temporal difference in RT distribution.

3.7.2 Predictive values

In general the NPVs of NAATs are very high. However, even when analytical sensitivities and specificities are optimal, the clinical performance is not ideal. Like toxigenic culture, they detect toxigenic C. difficile but not necessarily in vivo toxin production so their specificities and PPV are poor compared with CCNA and their use may lead to over-diagnosis and inappropriate treatment 8. Their introduction into laboratories has resulted in sudden apparent increases in CDI rates, which are potentially misleading in the context of surveillance or public reporting of quality performance data 39, 40.

3.7.3 Cost-effectiveness

Their speed, convenience, sensitivity and specificity have led many laboratories to adopt NAAT as a single diagnostic test for CDI, despite its higher cost compared with immunoassays. A review of diagnostic methods (sponsored by several C. difficile NAAT manufacturers) concluded that, despite their greater cost, NAATs were cost-effective as standalone diagnostic tests on the basis of their high sensitivity and rapid turnaround time, which allows more rapid identification and isolation of CDI patients (or toxigenic C. difficile carriers with diarrhoea due to another cause or insignificant diarrhoea) and, hence, reduced risk of transmission 41.

3.7.4 Suitable acceptance criteria

Tests should be performed and interpreted according to manufacturer’s instructions.

3.8 Testing algorithms

Although many laboratories have chosen to use a NAAT as a standalone diagnostic test for CDI, others use test combinations, simultaneously or sequentially to optimise clinical predictive values, cost and flexibility, without major increases in turnaround times.

3.8.1 Commonly used CDI testing algorithm:

Image displays 3.8.1 Commonly used CDI testing algorithmD top of page
  1. GDH and toxin assays can be combined in a single assay. Immunoassays are rapid and easily performed on individual specimens without batching.
  2. This algorithm allows all negative and most positive tests to be reported rapidly, with high NPV and PPV and reduces the number of specimens in which NAAT is required to a small proportion of specimens, without loss of sensitivity or significant increase in turnaround time.
  3. Clinical records of patients in whom GDH and NAAT are positive, but the toxin immunoassay is negative should be reviewed. These patients may have CDI or could be C. difficile excretors in whom diarrhoea is due to another cause.
  4. In laboratories in which C. difficile cultures are performed to allow strain typing and/or susceptibility testing, cultures can be limited to specimens in which the GDH assay is positive.

4 SNOMED_CT Terminology

SNOMED CT concept Code
Clostridium difficile (organism)
5933001
Clostridium difficile diarrhea (disorder)
186431008
Clostridium difficile infection (disorder)
5891000119102
Clostridium difficile colitis (disorder) 
423590009
Clostridium difficile assay (procedure)
72415005
Clostridium difficile culture (procedure)
122209009
Clostridium difficile toxin assay (procedure)
75332002
Clostridium difficile antigen assay (procedure)
118114008
top of page

5 References

  1. Bartlett, J.G., et al., Role of Clostridium difficile in antibiotic-associated pseudomembranous colitis. Gastroenterology, 1978. 75(5): p. 778-82.
  2. Burnham, C.A. and K.C. Carroll, Diagnosis of Clostridium difficile infection: an ongoing conundrum for clinicians and for clinical laboratories. Clin Microbiol Rev, 2013. 26(3): p. 604-30.
  3. Carroll, K.C. and J.G. Bartlett, Biology of Clostridium difficile: implications for epidemiology and diagnosis. Annu Rev Microbiol, 2011. 65: p. 501-21.
  4. Peniche, A.G., T.C. Savidge, and S.M. Dann, Recent insights into Clostridium difficile pathogenesis. Curr Opin Infect Dis, 2013. 26(5): p. 447-53.
  5. Leffler, D.A. and J.T. Lamont, Clostridium difficile infection. N Engl J Med, 2015. 372(16): p. 1539-48.
  6. Freeman, J., et al., The changing epidemiology of Clostridium difficile infections. Clin Microbiol Rev, 2010. 23(3): p. 529-49.
  7. Adlerberth, I., et al., Toxin-producing Clostridium difficile strains as long-term gut colonizers in healthy infants. J Clin Microbiol, 2014. 52(1): p. 173-9.
  8. Planche, T. and M.H. Wilcox, Diagnostic pitfalls in Clostridium difficile infection. Infect Dis Clin North Am, 2015. 29(1): p. 63-82.
  9. Honda, H. and E.R. Dubberke, The changing epidemiology of Clostridium difficile infection. Curr Opin Gastroenterol, 2014. 30(1): p. 54-62.
  10. Wilcox, M.H., et al., A case-control study of community-associated Clostridium difficile infection. J Antimicrob Chemother, 2008. 62(2): p. 388-96.
  11. Eyre, D.W., et al., Diverse sources of C. difficile infection identified on whole-genome sequencing. N Engl J Med, 2013. 369(13): p. 1195-205.
  12. Rupnik, M. and J.G. Songer, Clostridium difficile: its potential as a source of foodborne disease. Adv Food Nutr Res, 2010. 60: p. 53-66.
  13. Hoover, D.G. and A. Rodriguez-Palacios, Transmission of Clostridium difficile in foods. Infect Dis Clin North Am, 2013. 27(3): p. 675-85.
  14. Goorhuis, A., et al., Emergence of Clostridium difficile infection due to a new hypervirulent strain, polymerase chain reaction ribotype 078. Clin Infect Dis, 2008. 47(9): p. 1162-70.
  15. Pepin, J., et al., Clostridium difficile-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ, 2004. 171(5): p. 466-72.
  16. McDonald, L.C., et al., An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med, 2005. 353(23): p. 2433-41.
  17. Loo, V.G., et al., Host and pathogen factors for Clostridium difficile infection and colonization. N Engl J Med, 2011. 365(18): p. 1693-703.
  18. Gerding, D.N. and S. Johnson, Does infection with specific Clostridium difficile strains or clades influence clinical outcome? Clin Infect Dis, 2013. 56(11): p. 1601-3.
  19. Kuijper, E.J., et al., Emergence of Clostridium difficile-associated disease in North America and Europe. Clin Microbiol Infect, 2006. 12 Suppl 6: p. 2-18.
  20. Riley, T.V., et al., First Australian isolation of epidemic Clostridium difficile PCR ribotype 027. Med J Aust, 2009. 190(12): p. 706-8.
  21. Richards, M., et al., Severe infection with Clostridium difficile PCR ribotype 027 acquired in Melbourne, Australia. Med J Aust, 2011. 194(7): p. 369-71.
  22. Eyre, D.W., et al., Emergence and spread of predominantly community-onset Clostridium difficile PCR ribotype 244 infection in Australia, 2010 to 2012. Euro Surveill, 2015. 20(10): p. 21059.
  23. Slimings, C., et al., Increasing incidence of Clostridium difficile infection, Australia, 2011-2012. Med J Aust, 2014. 200(5): p. 272-6.
  24. Eastwood, K., et al., Comparison of nine commercially available Clostridium difficile toxin detection assays, a real-time PCR assay for C. difficile tcdB, and a glutamate dehydrogenase detection assay to cytotoxin testing and cytotoxigenic culture methods. J Clin Microbiol, 2009. 47(10): p. 3211-7.
  25. Collins, D.A., B. Elliott, and T.V. Riley, Molecular methods for detecting and typing of Clostridium difficile. Pathology, 2015. 47(3): p. 211-8.
  26. Ferguson, J.K., et al., Clostridium difficile laboratory testing in Australia and New Zealand: national survey results and Australasian Society for Infectious Diseases recommendations for best practice. Pathology, 2011. 43(5): p. 482-7.
  27. Planche, T. and M. Wilcox, Reference assays for Clostridium difficile infection: one or two gold standards? J Clin Pathol, 2011. 64(1): p. 1-5.
  28. Planche, T.D., et al., Differences in outcome according to Clostridium difficile testing method: a prospective multicentre diagnostic validation study of C difficile infection. Lancet Infect Dis, 2013. 13(11): p. 936-45.
  29. Crobach, M.J., et al., European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI). Clin Microbiol Infect, 2009. 15(12): p. 1053-66.
  30. Cohen, S.H., et al., Clinical practice guidelines for Clostridium difficile 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(5): p. 431-55.
  31. Hink, T., C.A. Burnham, and E.R. Dubberke, A systematic evaluation of methods to optimize culture-based recovery of Clostridium difficile from stool specimens. Anaerobe, 2013. 19: p. 39-43.
  32. Carson, K.C., et al., Isolation of Clostridium difficile from faecal specimens--a comparison of chromID C. difficile agar and cycloserine-cefoxitin-fructose agar. J Med Microbiol, 2013. 62(Pt 9): p. 1423-7.
  33. Lister, M., et al., Comparison of culture based methods for the isolation of Clostridium difficile from stool samples in a research setting. Anaerobe, 2014. 28: p. 226-9.
  34. Eckert, C., et al., Evaluation of the chromogenic agar chromID C. difficile. J Clin Microbiol, 2013. 51(3): p. 1002-4.
  35. Polage, C.R., et al., Outcomes in patients tested for Clostridium difficile toxins. Diagn Microbiol Infect Dis, 2012. 74(4): p. 369-73.
  36. Peterson, L.R., et al., Detection of toxigenic Clostridium difficile in stool samples by real-time polymerase chain reaction for the diagnosis of C. difficile-associated diarrhea. Clin Infect Dis, 2007. 45(9): p. 1152-60.
  37. Androga, G.O., et al., Evaluation of the Cepheid Xpert C. difficile/Epi and meridian bioscience illumigene C. difficile assays for detecting Clostridium difficile ribotype 033 strains. J Clin Microbiol, 2015. 53(3): p. 973-5.
  38. Kok, J., et al., Presumptive identification of Clostridium difficile strain 027/NAP1/BI on Cepheid Xpert: interpret with caution. J Clin Microbiol, 2011. 49(10): p. 3719-21.
  39. Fong, K.S., et al., Impact of PCR testing for Clostridium difficile on incident rates and potential on public reporting: is the playing field level? Infect Control Hosp Epidemiol, 2011. 32(9): p. 932-3.
  40. Goldenberg, S.D., Public reporting of Clostridium difficile and improvements in diagnostic tests. Infect Control Hosp Epidemiol, 2011. 32(12): p. 1231-2; author reply 1233.
  41. Tenover, F.C., et al., Laboratory diagnosis of Clostridium difficile infection can molecular amplification methods move us out of uncertainty? J Mol Diagn, 2011. 13(6): p. 573-82.
top of page

In this section