What is the recommended diagnostic test and treatment for Clostridium difficile (C. diff) infection, comparing Enzyme Immunoassay (EIA) to toxin detection?

C. difficile Diagnostic Testing: EIA vs Toxin Detection

Direct Answer

EIA for toxin A/B alone should not be used as a stand-alone diagnostic test for C. difficile infection due to unacceptably low sensitivity (32-98%), despite its speed and convenience. 1

Recommended Diagnostic Approach

Two-Step Algorithm (Preferred Strategy)

The optimal diagnostic strategy is a two-step algorithm that balances sensitivity and specificity: 1, 2

Step 1 - High Sensitivity Screening:

• Use either NAAT (nucleic acid amplification test/PCR) OR GDH (glutamate dehydrogenase) EIA as the initial screening test 1, 2
• Both have excellent sensitivity (80-100% for NAAT; >90% for GDH) 1
• Negative results can be reported as negative without further testing 1

Step 2 - Confirmatory Toxin Detection:

• All positive screening tests must be followed by toxin A/B EIA for confirmation 1, 2
• This step differentiates active infection from asymptomatic colonization 1
• The two-step approach achieves combined sensitivity of 91% and specificity of 98% 2

Why EIA for Toxin A/B Fails as Stand-Alone Test

The fundamental problem with toxin EIA alone is its poor sensitivity: 1

• Sensitivity ranges from 32-98% (unacceptably variable) 1
• Specificity is high (84-100%) but this doesn't compensate for missed cases 1
• Positive predictive values are unacceptably low (0.28-0.77) at typical CDI prevalence of 5-10% 1
• Missing active CDI cases directly impacts mortality and morbidity through delayed treatment and continued transmission 1

Critical Distinction: Toxin Detection vs Organism Detection

Understanding what each test detects is essential for proper interpretation: 3

• Toxin EIA: Detects free toxins A and/or B in stool (indicates active toxin production) 1, 3
• GDH EIA: Detects C. difficile enzyme but cannot distinguish toxigenic from non-toxigenic strains 1
• NAAT/PCR: Detects toxin genes (indicates presence of toxigenic organism, not necessarily active toxin production) 1
• Culture: Detects organism presence but is too slow (48+ hours) and cannot detect toxin presence 1, 2

The Colonization Problem

NAAT alone as a single-step test risks overdiagnosis: 1

• Up to 7% of asymptomatic hospitalized patients are colonized with toxigenic C. difficile 1
• NAAT detects toxin genes in both colonized and infected patients 1
• Using NAAT alone leads to overtreatment, unnecessary antibiotic exposure, and delayed recognition of other diarrheal causes 1
• This is why confirmatory toxin testing is critical after positive NAAT 1, 2

Practical Implementation Algorithm

For symptomatic patients with ≥3 unformed stools in 24 hours: 1

1. Test only unformed stool (formed stool causes false positives) 1

2. First-line: GDH EIA + Toxin A/B EIA simultaneously 1

   • If both negative → Report negative
   • If GDH positive + Toxin positive → Report positive CDI
   • If GDH positive + Toxin negative → Perform NAAT for confirmation 1, 2
   • If GDH negative + Toxin positive → Repeat testing or clinical correlation 1

3. Alternative: NAAT first, then Toxin EIA on positives 1, 2

   • NAAT negative → Report negative
   • NAAT positive → Perform toxin A/B EIA
   • If toxin positive → Active CDI confirmed
   • If toxin negative → Likely colonization vs early infection (clinical correlation required) 1

Critical Pitfalls to Avoid

Do not repeat testing during the same diarrheal episode unless in outbreak situations or very high clinical suspicion 1

Do not test patients with ileus using standard stool samples - consider perirectal swabs (95.7% sensitivity, 100% specificity by PCR) 1

Do not use culture for routine diagnosis - reserve for epidemiological typing and outbreak investigation only 1, 2

Do not test formed stools - this dramatically increases false positives and leads to unnecessary treatment 1

When Discordant Results Occur

If toxin is negative but organism/genes are detected: 1

• Cannot definitively differentiate colonization from infection 1
• Requires clinical correlation (symptoms, risk factors, alternative diagnoses) 1
• Consider treating if high clinical suspicion (recent antibiotics, healthcare exposure, severe symptoms) 1