Evaluation of the VIDAS glutamate dehydrogenase assay for the detection of Clostridium difficile

We evaluated the performance of the VIDAS GDH assay for the detection of Clostridium difficile. In total, 350 fecal specimens collected from patients clinically suspected of having CDI were analyzed by C. difficile culture and enzyme-linked fluorescent immunoassay (VIDAS GDH); the results were compared with those of toxigenic C. difficile culture (TC), PCR (Xpert C. difficile assay), and toxin AB EIA (VIDAS CDAB). The numbers of culture-positive and culture-negative samples were 108 and 242, respectively. The concordance between the GDH assay and C. difficile culture was 90.3%. With PCR, 12 more samples were found to be positive in GDH-positive/C. difficile culture-negative specimens. Thus, the concordance between GDH assay and C. difficile culture/PCR was 93.7%. The sensitivity, specificity, positive predictive value, and negative predictive value of the VIDAS GDH assay were 97.2%, 87.2%, 77.2%, and 98.6%, respectively, based on the C. difficile culture, and 97.5%, 91.7%, 86.0%, and 98.6%, respectively, based on C. difficile culture/PCR. Positivity rates of the GDH assay were partially associated with those of semi-quantitative C. difficile cultures, which were maximized in grade 3 (>100 colony-forming unit [CFU]) compared with grade 1 (<10 CFU). We evaluated the two-step or three-step algorithm using GDH assay as a first step. No toxin EIA-positive case was found among GDH-negative samples, and 60.8% (48/79) were TC- and/or PCR-positive among the GDH-positive/toxin EIA-negative samples. Thus, approximately 25% of the 350 samples required a confirmatory test (TC or PCR) in the GDH-toxin EIA algorithm, whereas only 2.3% of the total samples in GDH-PCR algorithm was discrepant and required another confirmatory test like TC.