Application of diagnostic algorithms for Clostridioides difficile infection using a GDH/ toxin test for five years from a single center in Korea: a diagnostic accuracy study

Abstract

Background: Accurate diagnosis of Clostridioides difficile infection (CDI) requires both microbiologic confirmation and clinical correlation. Current guidelines recommend a two-step algorithm combining a sensitive screening test with a specific confirmatory assay. This study evaluated the diagnostic performance of the glutamate dehydrogenase (GDH)/toxin enzyme immunoassay (EIA) over five years and assessed its suitability as an initial screening test.

Methods: We retrospectively analyzed 8,685 C. difficile-related tests conducted between March 2020 to February 2025. The GDH/toxin EIA was performed using the C. DIFF QUIK CHEK COMPLETE (TechLab). Toxigenic culture involved alcohol-shocked stool samples plated on chromogenic agar and incubated anaerobically for 48 hours. Toxin gene polymerase chain reaction (PCR) was done using the BD MAX Cdiff assay and the Xpert C. difficile assay.

Results: The GDH test showed a sensitivity of 77.0% and negative predictive value (NPV) of 95.1% compared with culture. The toxin EIA showed 35.0% sensitivity and 96.9% positive predictive value relative to PCR. The combined GDH+Toxin EIA achieved 82.6% sensitivity and 96.9% NPV compared with PCR. Most discordant results involved low bacterial burden or non-toxigenic isolates. GDH positivity correlated with growth quantity, and toxin EIA positivity varied by ribotype. Algorithm modeling suggested the GDH/toxin test as a cost-effective first-line option.

Conclusion: The GDH/toxin EIA demonstrated high NPV and may be appropriate as an initial test in CDI diagnostic algorithms. These findings support its role in diagnostic stewardship and provide evidence to inform the development of national diagnostic guidelines in Korea.