What is the recommended approach for laboratory diagnosis of Clostridium difficile?

Laboratory Diagnosis of Clostridium difficile Infection

A two-step diagnostic algorithm is the recommended best practice, starting with a high-sensitivity screening test (either GDH enzyme immunoassay or NAAT/PCR) followed by confirmatory testing with toxin A/B detection to distinguish active infection from asymptomatic colonization. 1, 2

Recommended Diagnostic Algorithms

Three evidence-based approaches are recommended as best practices for C. difficile laboratory diagnosis 1:

Option 1: Two-Step GDH/NAAT Algorithm

• Step 1: Screen with GDH enzyme immunoassay (sensitivity 90-94%, specificity 87-90%) 3
• Step 2: If GDH positive, confirm with NAAT/PCR (sensitivity 92%, specificity 96%) 3
• This combined approach achieves sensitivity of 91% and specificity of 98% 2, 4

Option 2: Two-Step GDH/Toxin/NAAT Algorithm

• Step 1: Screen with GDH enzyme immunoassay 2, 4
• Step 2: If GDH positive, test for toxin A/B by EIA 2, 4
• Step 3: If toxin negative but GDH positive, perform NAAT as tiebreaker 2, 4
• This achieves PPV of 82-85% and NPV of 98-99% at endemic prevalence rates 3

Option 3: NAAT-Only Testing

• Single-step NAAT/PCR testing is acceptable as a standalone best practice 1
• Provides rapid results with high sensitivity and specificity 1

Critical Testing Principles

Only test symptomatic patients with clinically significant diarrhea (≥3 unformed stools in 24 hours) to avoid detecting asymptomatic colonization rather than active infection 2, 4. Testing formed stools or asymptomatic patients leads to false positive results 2.

Never rely on toxin EIA alone as a standalone test due to unacceptably low sensitivity 2, 3:

• Well-type toxin A/B EIA: only 66% sensitivity compared to toxigenic culture 3
• Membrane-type toxin A/B EIA: only 52% sensitivity, missing nearly half of true cases 3
• At endemic CDI prevalence (5-10%), toxin EIA has poor PPV of only 28-77%, meaning most positive results could be false positives 1, 3

Reference Standard Methods (Not for Routine Use)

Cell culture cytotoxicity assay (CCA) detects toxin B by observing cytopathic effects on cultured cells after 24-48 hours 1, 3. This remains the historical reference standard but is time-consuming and requires specialized facilities 1.

Toxigenic culture (TC) involves culturing C. difficile on selective media (cycloserine-cefoxitin-fructose agar) followed by in vitro toxin detection of isolated strains 1, 3. This provides higher sensitivity than CCA but takes at least 48 hours for anaerobic incubation 1.

Both reference methods are impractical for routine clinical use, which is why rapid EIAs and molecular tests have replaced them 3.

Common Pitfalls to Avoid

Do not use toxin A-only detection assays, as they demonstrate even lower sensitivity of 61-82% and will miss substantial proportions of true CDI cases 3. Always use assays detecting both toxins A and B 3.

Do not perform "test of cure" after treatment, as patients may continue to shed C. difficile spores for up to six weeks after successful treatment 2. Repeat testing is not indicated unless symptoms recur 2.

Do not perform repeat testing within short intervals using NAAT, as insufficient evidence exists to support this practice and it may detect prolonged colonization rather than reinfection 1.

Avoid stool culture as a routine diagnostic test due to slow turnaround time (≥48 hours), inability to distinguish toxigenic from non-toxigenic strains without additional testing, and detection of asymptomatic colonization 4. Culture should be reserved for epidemiological typing during outbreak investigations 4.

Understanding Test Targets

GDH (glutamate dehydrogenase) is a metabolic enzyme produced by all C. difficile strains—both toxigenic and non-toxigenic—making it highly sensitive for screening but unable to distinguish pathogenic from non-pathogenic strains 3.

Toxins A and B are the primary virulence factors that cause disease through direct cellular damage by glucosylating Rho family GTPases, leading to cytoskeletal disorganization and colonocyte death 3. Only toxigenic strains cause disease, as CDI is a toxin-mediated infection 3.

NAAT/PCR detects toxin genes (tcdA and tcdB) rather than actual toxin production, which explains why it may be more sensitive than toxin detection but can identify colonized patients without active toxin production 3, 5.

Performance Characteristics Summary

At 5% CDI prevalence in a population of 10,000 patients, the two-step algorithm using GDH screening followed by toxin confirmation results in only 1% overall error rate 3. Single toxin EIA testing at this prevalence would yield unacceptably high false positive rates 1, 3.

The diagnostic approach must balance sensitivity (to avoid missing true cases and prevent transmission) with specificity (to avoid overdiagnosis and inappropriate treatment) 1.