Advertisement

Overcoming barriers to effective recognition and diagnosis of Clostridium difficile infection

  • M.H. Wilcox
    Correspondence
    Corresponding author: M. H. Wilcox, Department of Microbiology, Old Medical School, Leeds General Infirmary, Great George Street, Leeds, LS1 3EX, UK
    Affiliations
    Department of Microbiology, Old Medical School, Leeds Teaching Hospitals NHS Trust & University of Leeds, Leeds General Infirmary, Leeds, UK
    Search for articles by this author

      Abstract

      With the frequency of cases of Clostridium difficile infection (CDI) increasing in many developed countries, accurate and reliable laboratory diagnosis of CDI is more important than ever. However, the diagnosis of CDI has been handicapped by the existence of two reference standards, one of which detects C. difficile toxin (cytotoxin assay) and the other only toxigenic strains (cytotoxigenic culture). Being relatively slow and laborious to perform, these reference methods were largely abandoned as routine diagnostic methods for toxin detection in favour of stand-alone rapid enzyme immunoassays (EIAs), which have suboptimal sensitivity and specificity. The management of CDI is undermined by high rates of both false-positive and false-negative test results. More recently developed nucleic acid amplification tests (NAATs) for toxin gene detection offer improved sensitivity over immunoassays, but fail to discriminate between CDI and asymptomatic colonization with C. difficile, and have clear drawbacks as stand-alone diagnostic tests. Two-step or three-step diagnostic algorithms have been proposed as a solution. In a large study of the effectiveness of currently available tests, a diagnostic algorithm was developed that combines available tests to more effectively distinguish patients with CDI from uninfected patients. This two-test protocol, which is now used in National Health Service laboratories in England, comprises an EIA for glutamate dehydrogenase detection or NAATs for toxin gene detection, followed by a relatively sensitive toxin EIA. This algorithm also identifies ‘potential C. difficile excretors’, individuals with diarrhoeal samples that contain C. difficile but without demonstrable toxin, who may be a source of transmission of C. difficile to susceptible patients.

      Keywords

      Introduction

      Clostridium difficile is recognized as the leading cause of infectious nosocomial diarrhoea in developed countries [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ], and is the main cause of antibiotic-associated pseudomembranous colitis [
      • Bartlett JG.
      Historical perspective on studies of Clostridium difficile and C. difficile infection.
      ]. The emergence of a hypervirulent strain of C. difficile (ribotype 027), which first appeared about a decade ago in North America and then spread to parts of Europe, has raised the profile of C. difficile in many countries [
      • Freeman J.
      • Bauer MP.
      • Baines SD
      • et al.
      The changing epidemiology of Clostridium difficile infections.
      ]. Nonetheless, in many parts of Europe, C. difficile infection (CDI) remains an under-recognized disease and is therefore under-diagnosed and under-reported. A lack of disease awareness leads to a reduced level of clinical suspicion of CDI and, potentially, a cycle of reduced requests for CDI testing, less diagnosis of cases, and potentially increased transmission of infection.
      A lack of consistent diagnostic and screening practices for CDI across European countries (and even from hospital to hospital within countries) makes it difficult to estimate the true number of cases of CDI that occur across Europe each year. The lack of consistent and accurate diagnostic data hampers our understanding of the epidemiology of CDI in Europe, with current incidence and prevalence rates more likely to reflect the testing practices in a particular country than the true frequency of CDI cases [
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ]. Countries where frequent testing does not occur tend to have a low estimated rate of diagnosis, whereas countries with frequent testing for CDI tend to have higher rates of diagnosis—in other words, the more you test, the more you find. This point is particularly well illustrated by data from a recent pan-European epidemiological survey of rates of CDI in Europe carried out in 2008 [
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ]. The survey, which covered between one and six hospitals in each of the 34 countries surveyed, showed that testing frequency varied between countries by more than 40 times, ranging from as few as three tests per 10 000 patient-days to as many as 141 tests per 10 000 patient-days. The measured incidence of CDI was highest where the frequency of testing was highest (Fig. 1).
      Figure thumbnail gr1
      FIG. 1Correlation between the frequency of Clostridium difficile infection (CDI) testing and measured CDI incidence [
      • Bauer MP.
      • Notermans DW.
      • van Benthem BH
      • et al.
      Clostridium difficile infection in Europe: a hospital-based survey.
      ].
      Throughout Europe, there is a need to raise awareness of C. difficile as a cause of nosocomial diarrhoea, and to standardize when and how CDI testing occurs. This will help to ensure that suspected cases of CDI can be diagnosed with greater accuracy and reliability than at present. It is difficult to manage patients well and to implement effective infection control measures without accurate diagnosis of CDI, and, in turn, interrupting the transmission of infection may be compromised. Inaccurate diagnostic protocols also lead to poor epidemiological data and undermine effective disease surveillance. Accurate diagnosis of CDI is a prerequisite for determining whether patients need specific antibiotic therapy for this infection and, potentially, whether there is also a need for antibiotic susceptibility testing. With the possible exception of rifaximin, which is used to treat CDI in some European countries and for which resistance has been identified in up to one-quarter of C. difficile isolates [
      • Huhulescu S.
      • Sagel U.
      • Fiedler A
      • et al.
      Rifaximin disc diffusion test for in vitro susceptibility testing of Clostridium difficile.
      ], there does not yet seem to be a requirement for susceptibility testing as part of routine diagnosis.
      Based on a presentation that took place during the 22nd European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), held in London, UK, from 31 March to 3 April 2012, this review will look at the current state of diagnostic testing for CDI from a European perspective, and specifically at new approaches being adopted in England to improve the outcome of diagnostic testing for CDI. This review is not intended to be a comprehensive or systematic review of the published literature on CDI testing. In addition, issues relating to the costs and logistics of using two-step diagnostic testing algorithms in preference to single-test procedures, although important, are beyond the scope of the review.

      Laboratory Diagnosis of CDI—Still in a State of Flux

      Most patients with CDI present with profuse watery diarrhoea that has a distinctive, foul-smelling odour. However, on its own, this is insufficiently reliable for a definitive diagnosis of CDI, because of the many infective and non-infective causes of diarrhoea. Additional laboratory tests are therefore needed to confirm a presumptive clinical diagnosis of CDI [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ] (Table 1).
      TABLE 1.Current diagnostic tests for Clostridium difficile infection
      Diagnostic targetTesting method
      C. difficile toxin (TcdA or TcdB)Cell cytotoxicity assay
      Reference test methods detect different targets, and are not directly comparable.
      Enzyme immunoassays
      Membrane assays
      C. difficileAnaerobic culture
      Glutamate dehydrogenase (common antigen) enzyme immunoassays
      Toxigenic C. difficileCytotoxigenic culture
      Reference test methods detect different targets, and are not directly comparable.
      Nucleic acid amplification tests, PCR
      TcdA, C. difficile toxin A; TcdB, C. difficile toxin B.
      a Reference test methods detect different targets, and are not directly comparable.
      The cell cytotoxicity assay and cytotoxigenic culture have traditionally been considered to be the reference methods for the diagnosis of CDI. The first detects the presence of C. difficile toxins, toxin B and toxin A, in a patient's faecal sample [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. By contrast, cytotoxigenic culture detects C. difficile strains that have the capacity to produce toxins. However, the availability of two reference standards is problematic, because each method detects a different target, and the two methods are not directly comparable (Table 1).
      Experience from the UK suggests that many laboratories abandoned the standard reference methods in favour of more rapid toxin detection kits based on enzyme immunoassays (EIAs). A survey carried out in the UK in 1997 showed that, at that time, only 45% of laboratories used the reference standards, as compared with 52% that used toxin detection kits and 31% that used traditional anaerobic culture [
      • Wilcox MH.
      • Smyth ETM.
      Incidence and impact of Clostridium difficile infection in the UK, 1993-1996.
      ]. For the past 10-15 years, the laboratory diagnosis of CDI has relied mainly on the use of EIAs to detect the presence of the major C. difficile toxins [
      • Wilcox MH.
      Laboratory diagnosis of Clostridium difficile infection: in a state of transition or confusion or both.
      ]. This is despite evidence from Brazier (1998), who reviewed data from several publications and found that toxin detection kits are approximately 10-20% less sensitive than the cell cytotoxicity test [
      • Brazier JS.
      The diagnosis of Clostridium-difficile-associated disease.
      ]. With the advent of more sensitive methods, it is now clear that detection rates with EIA toxin tests may be even lower than previously published, with as many as 50% of cases being missed by some assays [
      • Sloan LM.
      • Duresko BJ.
      • Gustafson DR
      • et al.
      Comparison of real-time PCR for detection of the tcdC gene with four toxin immunoassays and culture in diagnosis of Clostridium difficile infection.
      ,
      • Larson AM.
      • Fung AM.
      • Fang FC.
      Evaluation of tcdB real-time PCR in a three-step diagnostic algorithm for detection of toxigenic Clostridium difficile.
      ]. Diagnostic accuracy is undermined by both false-positive and false-negative test results. False-positive test results for CDI can lead to unnecessary antibiotic treatment and isolation of patients, and false-negative test results can lead to a delay in treatment, with the risk of adverse clinical sequelae, such as fulminant colitis, and of disease transmission to other patients.
      A recent systematic review of six commercial C. difficile toxin detection assays found that the overall positive predictive value (PPV) was unacceptably low, being <50% in some cases [
      • Planche T.
      • Aghaizu A.
      • Hollman R
      • et al.
      Diagnosis of Clostridium difficile by toxin detection kits: a systematic review.
      ]. Such low PPVs increase the risk of false-positive results, especially when the prevalence of CDI is low. These findings are supported by data from a large study in which nine commercial toxin detection assays were compared with standard reference methods. On the basis of a hospital setting with an assumed CDI prevalence rate of 10%, the toxin detection tests had low-moderate PPVs ranging from 48.6% to 86.8% [
      • Eastwood K.
      • Else P.
      • Charlett A
      • 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.
      ]. In this study, false-positive and false-negative results were generally not obtained in the same samples that were tested by different assays, suggesting that incorrect diagnoses were attributable to inaccuracies in the toxin detection kits rather than to other factors. Overall, these analyses demonstrate that no single immunoassay for the detection of C. difficile toxin is adequately sensitive or specific for the accurate diagnosis of CDI.
      Given that EIA toxin detection tests may, at best, fail to detect CDI in 20% of true cases of CDI, while also falsely identifying CDI in one or two samples of every ten positives [
      • Wilcox MH.
      Laboratory diagnosis of Clostridium difficile infection: in a state of transition or confusion or both.
      ,
      • Wilcox MH.
      • Eastwood M.
      Evaluation report.
      ], the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) advocated a two-step approach to diagnosis [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. The ESCMID two-step diagnostic algorithm is based on an initial screening test with a high negative predictive value (NPV) to identify those individuals for whom a confirmatory test must be performed in order to identify a truly positive case of CDI [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ] (Fig. 2). The greatest strength of a two-step diagnostic algorithm as advocated by the ESCMID is that screening tests, such as assays for glutamate dehydrogenase (GDH) or nucleic acid amplification tests (NAATs) for C. difficile toxin genes, have high NPVs at low CDI prevalence, and can thus be used to reliably exclude patients without infection [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. This means that negative results can be issued promptly, and patients with non-CDI diarrhoea managed appropriately [
      • Planche T.
      • Aghaizu A.
      • Hollman R
      • et al.
      Diagnosis of Clostridium difficile by toxin detection kits: a systematic review.
      ]. Some authors have recently questioned the reliability of assays for GDH and their use in two-step diagnostic algorithms [
      • Novak-Weekley SM.
      • Marlowe EM.
      • Miller JM
      • et al.
      Clostridium difficile testing in the clinical laboratory by use of multiple testing algorithms.
      ]. Although the choice of assay method may influence both sensitivity and specificity, as illustrated in the ESCMID analysis, the reported sensitivity and specificity of GDH assays generally exceed 90% [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. It has also been suggested that the sensitivity of the GDH assay may be influenced by the clonal mix of C. difficile strains in circulation in the hospital environment [
      • Tenover FC.
      • Novak-Weekley SM.
      • Woods CW
      • et al.
      Impact of strain types on detection of toxigenic Clostridium difficile: comparison of molecular diagnostic and enzyme immunoassay approaches.
      ], although published data on this are limited. When a positive test result is obtained on initial testing, it can only be reported as provisionally positive, and further testing must be performed with, for example, a toxin detection test [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ,
      • Planche T.
      • Aghaizu A.
      • Hollman R
      • et al.
      Diagnosis of Clostridium difficile by toxin detection kits: a systematic review.
      ]. Another potential benefit of a two-step approach to CDI diagnosis is that it discourages the submission of multiple faecal samples for repeat testing, which studies have shown can lead to a significant increase in the number of false-positive test results, as well as being highly wasteful of resources [
      • Drees M.
      • Snydman DR.
      • O'Sullivan CE.
      Repeated enzyme immunoassays have limited utility in diagnosing Clostridium difficile.
      ,
      • Litvin M.
      • Reske KA.
      • Mayfield J
      • et al.
      Identification of a pseudo-outbreak of Clostridium difficile infection (CDI) and repeated testing, sensitivity, and specificity on perceived prevalence of CDI.
      ].
      Figure thumbnail gr2
      FIG. 2Two-step diagnostic algorithm for Clostridium difficile infection (CDI) as recommended by the European Society of Clinical Microbiology and Infectious Diseases [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. EIA, enzyme immunoassay; GDH, glutamate dehydrogenase; TcdA, C. difficile toxin A; TcdB, C. difficile toxin B. *A positive toxinogenic culture always indicates the presence of toxin-producing C. difficile and makes further testing unnecessary. Reproduced with permission from Crobach MJT 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: 1053-1066. ©2009 The Authors. ©2009 European Society of Clinical Microbiology and Infectious Diseases.
      Despite widespread acceptance that a two-step approach can help to address many of the deficiencies that exist with currently available diagnostic tests for CDI, it has been inconsistently applied. A comprehensive survey of 170 hospital trusts in England in 2010, 99% of which responded, revealed that 70% of their microbiology laboratories used an EIA toxin detection test as a stand-alone method for CDI, even though PPVs were as low as 20% in some cases [
      • Goldenberg SD.
      • French GL.
      Diagnostic testing for Clostridium difficile: a comprehensive survey of laboratories in England.
      ]. Failure to provide an accurate and reliable diagnostic service for CDI undermines mandatory reporting of CDI [
      • Chand MA.
      • Fleming MJ.
      • Wellsteed S
      • et al.
      Impact of changes in Clostridium difficile diagnostic testing on detection of C. difficile infection and all England mandatory surveillance data.
      ], which is a requirement in the UK, as well as affecting patient services. For example, the survey showed that, in over one-third of laboratories using stand-alone EIAs for C. difficile toxins, the probability that a reported positive test result represented a true case of CDI was <50% [
      • Wilcox MH.
      Laboratory diagnosis of Clostridium difficile infection: in a state of transition or confusion or both.
      ].
      Given that infection control requires CDI patients to be placed in isolation, cohorting of cases is sometimes used if there is insufficient single-room capacity. However, if patients are managed according to diagnostic results of below optimal accuracy, such a practice risks mixing CDI patients with patients whose diarrhoea has other causes. False-positive results can lead to unnecessary antibiotic treatment, with an increased risk of adverse events, the potential acquisition of vancomycin-resistant enterococci, and, because of the loss of colonization resistance, an increased risk of CDI.

      Can Molecular Diagnostic Methods Help?

      In the search for more accurate diagnostic tests for CDI, molecular tests that detect the genes encoding C. difficile toxin A and toxin B have been developed and are now commercially available. NAATs are more sensitive than EIA detection kits. In the comparative study of nine toxin detection assays, a PCR for the toxin B gene was more sensitive than the cell cytotoxicity assay (92.2%) and cytotoxigenic culture (88.5%). Specificity in relation to the standard reference tests was 94.0% and 95.4%, respectively [
      • Eastwood K.
      • Else P.
      • Charlett A
      • 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.
      ]. However, although detection of toxin genes with NAATs is highly sensitive and allows differentiation between faecal samples with and without C. difficile that has toxigenic potential [
      • Belanger S.
      • Boissinot M.
      • Clairoux N
      • et al.
      Rapid detection of Clostridium difficile in feces by real-time PCR.
      ], this does not discriminate between CDI and asymptomatic colonization with C. difficile. This is probably the greatest weakness of NAATs, because these assays do not detect the presence of faecal toxin, without which a diagnosis of CDI cannot be made with confidence [
      • Wilcox MH.
      • Planche T.
      • Pang PC
      • et al.
      Point-counterpoint. What is the current role of algorithmic approaches for diagnosis of Clostridium difficile infection.
      ].
      The distinction between active infection and asymptomatic colonization with toxigenic C. difficile is important, because there can be high rates of asymptomatic carriage, especially among elderly inpatients and those residing in long-term care or nursing homes, where colonization rates of 10-50% have been reported [
      • Kyne L.
      • Warny M.
      • Qamar A
      • et al.
      Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
      ,
      • Riggs MM.
      • Sethi AK.
      • Zabarsky TF
      • et al.
      Asymptomatic carriers are a potential source for transmission of epidemic and non-epidemic strains of Clostridium difficile strains among long-term care facilities.
      ]. Although only diarrhoeal specimens should be submitted for diagnostic testing for CDI, diarrhoea is a common symptom among elderly inpatients, many of whom have received antibiotics or laxatives, or have been exposed to other pathogens, such as norovirus. Considerable potential exists to detect toxigenic C. difficile as an innocent, and possibly protective, bystander. As studies have demonstrated, colonization by C. difficile may be protective against the development of CDI when accompanied by a robust anti-toxin IgG antibody response [
      • Kyne L.
      • Warny M.
      • Qamar A
      • et al.
      Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
      ]. Current treatment guidelines also advise against antibiotic treatment of asymptomatic carriers [
      • Bauer MP.
      • Kuijper EJ.
      • van Dissel JT.
      European Society of Clinical Microbiology and Infectious Diseases (ECSMID): treatment guidance document for Clostridium difficile infection (CDI).
      ].
      The difficulty in distinguishing between asymptomatic carriage of toxigenic C. difficile and CDI in a patient presenting with unexplained diarrhoea is illustrated by data from a survey of elderly hospitalized patients with viral gastroenteritis, the incidence of which has increased dramatically in recent years, owing to the emergence of new norovirus variants [
      • Wilcox M.
      • Fawley W.
      Viral gastroenteritis increases the reports of Clostridium difficile infection.
      ]. On wards affected by viral gastroenteritis, there was a significant increase in the detection of cytotoxin-positive cases of CDI. In such cases where the patient presents with diarrhoea (with or without vomiting), the significance of a C. difficile-positive result, especially where C. difficile toxin is not sought, is difficult to interpret. For example, do such patients have CDI and require specific anti-CDI antibiotic therapy, or are they just carriers of the bacterium who can be managed solely with supportive therapy, as for viral gastroenteritis?
      We know that up to 25% of CDI patients experience recurrence within 30 days following antibiotic treatment with either oral metronidazole or vancomycin [
      • Louie TJ.
      • Miller MA.
      • Mullane KM
      • et al.
      Fidaxomicin versus vancomycin for Clostridium difficile infection.
      ,
      • Bouza E.
      • Dryden M.
      • Mohammed R
      • et al.
      Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
      ,
      • Lowy I.
      • Molrine DC.
      • Leav BA
      • et al.
      Treatment with monoclonal antibodies against Clostridium difficile toxins.
      ]. However, long-term colonization with C. difficile is also common post-treatment. In a study of 52 patients with CDI, 56% were asymptomatic carriers 1-4 weeks after treatment cessation [
      • Sethi AK.
      • Al-Nassir WN.
      • Nerandzic MM
      • et al.
      Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection.
      ]. These findings illustrate the difficulties in interpreting a positive C. difficile NAAT result after a primary episode of CDI if such tests are used as the sole method of laboratory diagnosis.
      The availability of molecular diagnostic tests for CDI represents an important addition to the existing array of tests, but they are not a solution for all of the weaknesses of earlier testing methods. The new tests deliver relatively rapid results and high NPVs, but specificity for CDI is still an issue, as they detect strains (with toxin genes) rather than the presence of free toxins. In the author's opinion, this precludes their use as stand-alone tests for CDI. Nonetheless, it is important to acknowledge that this view is not universally held, and that some laboratories consider NAAT-only testing to be suitable for the diagnosis of CDI [
      • Musher DM.
      • Stager C.
      Diagnosis of Clostridium difficile infection.
      ], and would recommend specific CDI treatment for patients who are PCR-positive but EIA toxin-negative. Such an approach is open to debate, but certainly merits independent verification in a prospective study. Given that NAATs detect the toxin gene and not the toxin, the diagnostic accuracy of molecular-based testing can probably be enhanced by careful selection of patients to include only those with true diarrhoea and a strong clinical suspicion of CDI and, importantly, exclude patients who have been treated following a laboratory-confirmed diagnosis [
      • Kufelnicka AM.
      • Kirn TJ.
      Effective utilization of evolving methods for the laboratory diagnosis of Clostridium difficile infection.
      ].

      Optimizing the Diagnosis of CDI

      Currently available rapid tests for the diagnosis of CDI include EIAs/membrane tests for toxins A and B, immunoassays for the common antigen, GDH, and, most recently, NAATs that detect the genes encoding C. difficile toxins A and B. However, in the author's opinion, none is yet accurate enough to be used as a stand-alone diagnostic test for CDI, a view endorsed in a recent review [
      • Kufelnicka AM.
      • Kirn TJ.
      Effective utilization of evolving methods for the laboratory diagnosis of Clostridium difficile infection.
      ]. With a single diagnostic test, improving sensitivity usually compromises specificity. Theoretically, with combined tests it is possible to get the best of both worlds, with improved sensitivity (screening test) and improved specificity. This rationale underpins the two-step approach to the diagnosis of CDI, as demonstrated in the current ESCMID two-step diagnostic algorithm [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ].
      Understanding the source of C. difficile is also important for diagnostic purposes, as many clinicians still believe that a majority of CDI cases occur endogenously, with patients already harbouring C. difficile on admission to hospital, and CDI developing following subsequent antibiotic therapy. This is a common misconception, as asymptomatic carriers of toxigenic C. difficile are significantly less likely than non-carriers to develop CDI [
      • Kyne L.
      • Warny M.
      • Qamar A
      • et al.
      Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
      ]. However, asymptomatic carriers may be a potential reservoir of onward transmission. A recently published study that used molecular typing to match cases of nosocomially acquired CDI in >14 000 patients (1200 cases of CDI) with diarrhoea has highlighted how relatively few cases of CDI could be linked to a ward-based contact. The study revealed that, overall, no more than 25% of cases could be linked to a potential ward-based inpatient source; these ‘linkage’ rates varied from 37% in renal/transplant units, to 29% in haematology/oncology units and 28% in acute/elderly medicine units, with only 6% arising in specialist surgery units [
      • Walker AS.
      • Eyre DW.
      • Wyllie DH
      • et al.
      Characterisation of Clostridium difficile hospital ward-based transmission using extensive epidemiological data and molecular typing.
      ]. For these linked cases, the time for putative onward transmission of CDI for most patients was between 1 and 4 weeks, with the incubation period from ward-based contact with donor to recipient typically being up to 4 weeks. As <25% of CDI cases could be plausibly linked to other inpatients with CDI, other sources of transmission of C. difficile in the hospital should be explored.
      Several studies have alluded to the importance of asymptomatic C. difficile carriers as a potential source of transmission [
      • Riggs MM.
      • Sethi AK.
      • Zabarsky TF
      • et al.
      Asymptomatic carriers are a potential source for transmission of epidemic and non-epidemic strains of Clostridium difficile strains among long-term care facilities.
      ,
      • Sethi AK.
      • Al-Nassir WN.
      • Nerandzic MM
      • et al.
      Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection.
      ,
      • Johnson S.
      • Homann SR.
      • Bettin KM
      • et al.
      Treatment of asymptomatic Clostridium difficile carriers (fecal excretors) with vancomycin or metronidazole.
      ]. In a study of elderly patients in a long-term care facility affected by an outbreak of CDI, asymptomatic carriers outnumbered symptomatic patients by seven to one. However, levels of C. difficile contamination on the skin and in the surrounding environment of carriers approached those for symptomatic patients, suggesting that the former may be an important source of onward transmission [
      • Sethi AK.
      • Al-Nassir WN.
      • Nerandzic MM
      • et al.
      Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection.
      ]. In this respect, it is noteworthy that many CDI patients in whom diarrhoea resolves following a course of specific antibiotic therapy become asymptomatic carriers, and may continue shedding C. difficile spores for several weeks after treatment has ended [
      • Sethi AK.
      • Al-Nassir WN.
      • Nerandzic MM
      • et al.
      Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection.
      ]. In optimizing diagnostic algorithms for CDI, we need to be aware of patients who may not have a laboratory diagnosis of CDI (the primary purpose of testing), but are nonetheless important as potential sources of bacterial transmission.
      To address current deficiencies in diagnostic testing for CDI and improve mandatory reporting of laboratory detection of C. difficile, the Department of Health and the Health Protection Agency in England commissioned a large-scale, observational study involving four National Health Service (NHS) laboratories and diarrhoeal faecal samples from >12 500 patients. With such a large sample size, the accuracy of the diagnostic testing algorithms could be determined with high precision [
      Department of Health
      ] (Health Protection Agency website: http://www.hpa.org.uk/webc/HPAwebFile/HPAweb_C/1317132979562). The tests incorporated into the diagnostic algorithms were those routinely used by NHS laboratories, and included EIAs for C. difficile toxins, toxin gene detection by PCR, and EIA for GDH, with the cell cytotoxicity assay and cytotoxigenic culture being used as reference tests (Davies et al. 22nd ECCMID, 2012, Abstract LB-2817; http://registration.akm.ch/einsicht.php?XNABSTRACT_ID=151898&XNSPRACHE_ID=2&XNKONGRESS_ID=161&XNMASKEN_ID=900).
      An important objective of the study was to determine the relative clinical values of the two reference tests for CDI with respect to patient outcomes, such as 30-day mortality and morbidity-associated laboratory measurements. The results showed that the presence of C. difficile toxin in faecal samples was significantly associated with poor clinical outcome. Conversely, culture of toxigenic C. difficile in faeces in the absence of a positive toxin assay result (cytotoxigenic culture-positive but cytotoxin-negative) was not associated with any significantly worse clinical outcome than that of negative samples. However, samples containing C. difficile but with no demonstrable toxin can indicate ‘potential C. difficile excretors’, and knowledge of this may aid infection prevention and control measures for CDI [
      Department of Health
      ].
      These findings illustrate the importance of redesigning diagnostic algorithms to facilitate improved understanding of C. difficile cases among hospitalized patients. To this end, the results of the 12-month study showed that a diagnostic algorithm based on a test for GDH in combination with a NAAT for toxigenic C. difficile was optimal for sensitivity [
      Department of Health
      ] (Davies et al. 22nd ECCMID, 2012, Abstract LB-2817; http://registration.akm.ch/einsicht.php?XNABSTRACT_ID=151898&XNSPRACHE_ID=2&XNKONGRESS_ID=161&XNMASKEN_ID=900). However, the specificity and corresponding PPV were poor for this test combination, such that, in approximately four of 10 of positive results, the patients did not have a confirmed diagnosis of CDI. Thus, an algorithm based on these tests would be useful for excluding CDI but poor at determining whether CDI was truly present. Conversely, a diagnostic algorithm that combined a toxin EIA with a NAAT for toxigenic C. difficile yielded the highest specificity. The PPV was correspondingly higher (89.0-90.8%) than for the GDH-based algorithm (59.6-80.7%) according to the two reference tests. However, high specificity was achieved at the expense of sensitivity, so too many cases of CDI would be missed with this approach.
      The most advantageous solution was an algorithm based on either an EIA for GDH or a NAAT for toxigenic C. difficile for initial screening, with a follow-up toxin EIA as a confirmatory test [
      Department of Health
      ] (Davies et al. 22nd ECCMID, 2012, Abstract LB-2817; http://registration.akm.ch/einsicht.php?XNABSTRACT_ID=151898&XNSPRACHE_ID=2&XNKON-GRESS_ID=161&XNMASKEN_ID=900). The caveat here is that toxin EIAs have variable sensitivity, and so an assay with (relatively) high sensitivity needs to be used in this algorithm. In their diagnostic guidance, the ESCMID publish sensitivity and specificity rates for a number of EIA toxin detection kits, in which the sensitivity exceeded 90% for at least two kits [
      • Crobach MJT.
      • Dekkers OM.
      • Wilcox MH
      • et al.
      European Society of Clinical Microbiology and Infectious Diseases (ESCMID): data review and recommendations for diagnosing Clostridium difficile-infection (CDI).
      ]. In our study, the overall performance of the GDH/EIA toxin algorithm yielded a highly respectable NPV of 98.9%, demonstrating the algorithm's ability to correctly rule out CDI cases, and a PPV of 91.4%, ensuring confidence in the detection of most cases of CDI. As a result, this diagnostic algorithm for CDI was recommended for introduction to all NHS laboratories in England from April 2012, with results (consistent with CDI) being used for mandatory reporting (Fig. 3). Also, with this new NHS diagnostic algorithm, it was possible to identify not only when CDI was or was not likely to be present, but when C. difficile was probably present (in an individual with diarrhoea) and the patient was therefore a ‘potential C. difficile excretor’. A PCR test for toxin genes could be used as an optional third test to determine with more certainty which GDH-positive, toxin-negative patients are potential C. difficile excretors [
      Department of Health
      ] (Davies et al. 22nd ECCMID, 2012, Abstract LB-2817; http://registration.akm.ch/einsicht.php?XNABSTRACT_ID=151898&XNSPRACHE_ID=2&XNKONGRESS_ID=161&XNMASKEN_ID=900). Also, consistent with reports from earlier studies, this research again showed that C. difficile toxin EIAs are not suitable as stand-alone tests for the diagnosis of CDI or the detection of C. difficile [
      Department of Health
      ].
      Figure thumbnail gr3
      FIG. 3Two-step diagnostic algorithm for Clostridium difficile infection (CDI) in a patient with unexplained diarrhoea (Bristol Stool Chart types 5-7) [
      Department of Health
      ]. EIA, enzyme immunoassay; GDH, glutamate dehydrogenase; NAAT, nucleic acid amplification test. SIGHT: S = suspect that a case may be infective when there is no clear alternative cause for diarrhoea; I = isolate the patient within 2 h; G = gloves and aprons must be used for all contacts with the patient and their environment; H = hand-washing with soap and water should be carried out before and after each contact with the patient and the patient's environment; T = test the stool for C. difficile by sending specimen immediately. *Performance of a toxin EIA is not necessary in cases with a negative GDH EIA (or NAAT) result. Notes: A cell cytotoxin assay may be considered as an alternative to a sensitive toxin EIA, but yields slower results (l-2 days), and this will need to be taken into account in making decisions about infection control. Reproduced with permission. ©Crown copyright 2012; http://www.nationalarchives.gov.uk/doc/open-government-licence/open-government-licence.htm
      In conclusion, laboratory diagnosis is a crucial part of the management of patients with suspected CDI, but has been handicapped by a plethora of testing methods, the lack of a single reference method, and a growing reliance on stand-alone rapid diagnostic procedures, such as EIAs, for toxin detection. The result has been confusion and suboptimal testing. A growing body of evidence supports the use of a combination of tests for optimal laboratory diagnosis of CDI. An algorithm developed in UK laboratories and based on a two-step approach of GDH EIA (or PCR) to screen samples followed by a (relatively) sensitive confirmatory toxin EIA test is now recommended in NHS laboratories in England as the basis for testing and mandatory reporting of CDI. This algorithm has the additional advantage of being able to identify (symptomatic) potential C. difficile excretors. Although no specific treatment is indicated for such individuals, their identification offers a possible way of reducing CDI risk for other patients. Finally, as this review has highlighted, the more you test for C. difficile, the more you will find.

      Acknowledgements

      The author wishes to thank Elements Communications Ltd (Westerham, UK) for medical writing assistance, funded by Astellas Pharma Europe Ltd (Staines, UK).

      Provenance

      The development of the content and the printing of this supplement has been funded by Astellas Pharma Europe Ltd. This supplement was created in collaboration with the faculty from the Astellas-sponsored symposium at the 2012 ECCMID Congress.

      Transparency Declaration

      M. Wilcox has received funds for speaking, consultancy, advisory board membership and travel from Actelion, Astellas, Astra-Zeneca, Bayer, Cubist, Durata, J&J, Merck, Nabriva, Novacta, Novartis, Optimer, Pfizer, Sanofi-Pasteur, The Medicines Company, VH Squared, and Viropharma. M. Wilcox has received research funding from Actelion, Astellas, Biomerieux, Cubist, Pfizer, Summit, and The Medicines Company.

      References

        • Crobach MJT.
        • Dekkers OM.
        • Wilcox MH
        • 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: 1053-1066
        • Bartlett JG.
        Historical perspective on studies of Clostridium difficile and C. difficile infection.
        Clin Infect Dis. 2008; 46: S4-S11
        • Freeman J.
        • Bauer MP.
        • Baines SD
        • et al.
        The changing epidemiology of Clostridium difficile infections.
        Clin Microbiol Rev. 2010; 23: 529-549
        • Bauer MP.
        • Notermans DW.
        • van Benthem BH
        • et al.
        Clostridium difficile infection in Europe: a hospital-based survey.
        Lancet. 2011; 377: 63-73
        • Huhulescu S.
        • Sagel U.
        • Fiedler A
        • et al.
        Rifaximin disc diffusion test for in vitro susceptibility testing of Clostridium difficile.
        J Med Microbiol. 2011; 60: 1206-1212
        • Wilcox MH.
        • Smyth ETM.
        Incidence and impact of Clostridium difficile infection in the UK, 1993-1996.
        J Hosp Infect. 1998; 39: 181-187
        • Wilcox MH.
        Laboratory diagnosis of Clostridium difficile infection: in a state of transition or confusion or both.
        J Hosp Infect. 2011; 79: 1-3
        • Brazier JS.
        The diagnosis of Clostridium-difficile-associated disease.
        J Antimicrob Chemother. 1998; 41: 29-40
        • Sloan LM.
        • Duresko BJ.
        • Gustafson DR
        • et al.
        Comparison of real-time PCR for detection of the tcdC gene with four toxin immunoassays and culture in diagnosis of Clostridium difficile infection.
        J Clin Microbiol. 2008; 46: 1996-2001
        • Larson AM.
        • Fung AM.
        • Fang FC.
        Evaluation of tcdB real-time PCR in a three-step diagnostic algorithm for detection of toxigenic Clostridium difficile.
        J Clin Microbiol. 2010; 48: 124-130
        • Planche T.
        • Aghaizu A.
        • Hollman R
        • et al.
        Diagnosis of Clostridium difficile by toxin detection kits: a systematic review.
        Lancet Infect Dis. 2008; 8: 777-784
        • Eastwood K.
        • Else P.
        • Charlett A
        • 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: 3211-3217
        • Wilcox MH.
        • Eastwood M.
        Evaluation report.
        Clostridium difficile toxin detection assays. NHS Purchasing and Supply Agency, London2009
        • Novak-Weekley SM.
        • Marlowe EM.
        • Miller JM
        • et al.
        Clostridium difficile testing in the clinical laboratory by use of multiple testing algorithms.
        J Clin Microbiol. 2010; 48: 889-893
        • Tenover FC.
        • Novak-Weekley SM.
        • Woods CW
        • et al.
        Impact of strain types on detection of toxigenic Clostridium difficile: comparison of molecular diagnostic and enzyme immunoassay approaches.
        J Clin Microbiol. 2010; 48: 3719-3724
        • Drees M.
        • Snydman DR.
        • O'Sullivan CE.
        Repeated enzyme immunoassays have limited utility in diagnosing Clostridium difficile.
        Eur J Clin Microbiol Infect Dis. 2008; 27: 397-399
        • Litvin M.
        • Reske KA.
        • Mayfield J
        • et al.
        Identification of a pseudo-outbreak of Clostridium difficile infection (CDI) and repeated testing, sensitivity, and specificity on perceived prevalence of CDI.
        Infect Control Hosp Epidemiol. 2009; 30: 1166-1171
        • Goldenberg SD.
        • French GL.
        Diagnostic testing for Clostridium difficile: a comprehensive survey of laboratories in England.
        J Hosp Infect. 2011; 79: 4-7
        • Chand MA.
        • Fleming MJ.
        • Wellsteed S
        • et al.
        Impact of changes in Clostridium difficile diagnostic testing on detection of C. difficile infection and all England mandatory surveillance data.
        J Hosp Infect. 2011; 79: 8-12
        • Belanger S.
        • Boissinot M.
        • Clairoux N
        • et al.
        Rapid detection of Clostridium difficile in feces by real-time PCR.
        J Clin Microbiol. 2003; 41: 730-734
        • Wilcox MH.
        • Planche T.
        • Pang PC
        • et al.
        Point-counterpoint. What is the current role of algorithmic approaches for diagnosis of Clostridium difficile infection.
        J Clin Microbiol. 2010; 48: 4347-4353
        • Kyne L.
        • Warny M.
        • Qamar A
        • et al.
        Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A.
        N Engl J Med. 2000; 342: 390-397
        • Riggs MM.
        • Sethi AK.
        • Zabarsky TF
        • et al.
        Asymptomatic carriers are a potential source for transmission of epidemic and non-epidemic strains of Clostridium difficile strains among long-term care facilities.
        Clin Infect Dis. 2007; 45: 992-998
        • Bauer MP.
        • Kuijper EJ.
        • van Dissel JT.
        European Society of Clinical Microbiology and Infectious Diseases (ECSMID): treatment guidance document for Clostridium difficile infection (CDI).
        Clin Microbiol Infect. 2009; 15: 1067-1079
        • Wilcox M.
        • Fawley W.
        Viral gastroenteritis increases the reports of Clostridium difficile infection.
        J Hosp Infect. 2007; 66: 395-396
        • Louie TJ.
        • Miller MA.
        • Mullane KM
        • et al.
        Fidaxomicin versus vancomycin for Clostridium difficile infection.
        N Engl J Med. 2011; 364: 422-431
        • Bouza E.
        • Dryden M.
        • Mohammed R
        • et al.
        Results of a phase III trial comparing tolevamer, vancomycin and metronidazole in patients with Clostridium difficile-associated diarrhoea.
        Clin Microbiol Infect. 2008; 14: S103-S104
        • Lowy I.
        • Molrine DC.
        • Leav BA
        • et al.
        Treatment with monoclonal antibodies against Clostridium difficile toxins.
        N Engl J Med. 2010; 362: 197-205
        • Sethi AK.
        • Al-Nassir WN.
        • Nerandzic MM
        • et al.
        Persistence of skin contamination and environmental shedding of Clostridium difficile during and after treatment of C. difficile infection.
        Infect Control Hosp Epidemiol. 2010; 31: 21-27
        • Musher DM.
        • Stager C.
        Diagnosis of Clostridium difficile infection.
        Clin Infect Dis. 2012; 54: 1675-1676
        • Kufelnicka AM.
        • Kirn TJ.
        Effective utilization of evolving methods for the laboratory diagnosis of Clostridium difficile infection.
        Clin Infect Dis. 2011; 52: 1451-1457
        • Walker AS.
        • Eyre DW.
        • Wyllie DH
        • et al.
        Characterisation of Clostridium difficile hospital ward-based transmission using extensive epidemiological data and molecular typing.
        PLoS Med. 2012; 9: 1-11
        • Johnson S.
        • Homann SR.
        • Bettin KM
        • et al.
        Treatment of asymptomatic Clostridium difficile carriers (fecal excretors) with vancomycin or metronidazole.
        Ann Intern Med. 1992; 117: 297-302
        • Department of Health
        Updated guidance on the diagnosis and reporting of Clostridium difficile. 6 March 2012; (last accessed 4 May 2012).