A 2-step algorithm combining glutamate dehydrogenase and nucleic acid amplification tests for the detection of Clostridioides difficile in stool specimens.

The optimized diagnosis algorithm of Clostridioides difficile infection (CDI) is worldwide concerns. The purpose of this study was to assess the toxigenic C. difficile test performance and propose an optimal laboratory workflow for the diagnosis of CDI in mild virulent epidemic areas. Diarrhea samples collected from patients were analyzed by glutamate dehydrogenase (GDH), toxin AB (CDAB), and nucleic acid amplification test (NAAT). We assessed the performance of GDH, the GDH-CDAB algorithm, and the GDH-NAAT algorithm using toxigenic culture (TC) as a reference method. In this study, 186 diarrhea samples were collected. The numbers of TC-positive and TC-negative samples were 39 and 147, respectively. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and kappa of the GDH assay were 100%, 80.3%, 57.4%, 100%, and 0.63; of the GDH-CDAB algorithm were 48.7%, 97.3%, 82.6%, 87.7%, and 0.54; and of the GDH-NAAT algorithm were 74.4%, 100%, 100%, 93.6%, and 0.82, respectively. The GDH-NAAT algorithm has great concordance with TC in detecting toxigenic C. difficile (kappa = 0.82), while the sensitivity of the GDH-CDAB algorithm was too low to meet the demand of CDI diagnosis clinically. GDH-NAAT algorithm is recommended for the detection of toxigenic C. difficile with high specificity, increased sensitivity, and cost-effective.