Evaluation of Cycle Threshold, Toxin Concentration, and Clinical Characteristics of Clostridioides difficile Infection in Patients with Discordant Diagnostic Test Results.

Background: Clostridioides difficile infection (CDI) is one of the most common health-care associated infections that can cause significant morbidity and mortality. CDI diagnosis involves laboratory testing in conjunction with clinical assessment. The objective of this study was to assess the performance of various C. difficile tests and to compare clinical characteristics, Xpert® C. difficile /Epi (PCR) cycle threshold (CT), and Singulex Clarity® C. diff toxins A/B (Clarity) concentrations between groups with discordant test results. Methods : Unformed stool specimens from 200 hospitalized adults (100 PCR positive and 100 negative) were tested by cell cytotoxicity neutralization assay (CCNA), C. diff Quik Chek Complete® (Quik Chek), Premier® Toxins A and B, and Clarity. Clinical data, including CDI severity and CDI risk factors, were compared between discordant test results. Results : When compared to CCNA, PCR had the highest sensitivity at 100% and Quik Chek had the highest specificity at 100%. Amongst clinical and laboratory data studied, prevalence of leukocytosis, prior antibiotic use, and hospitalizations were consistently higher across all subgroups comparing toxin positive to toxin negative patients. Among PCR positive samples, median CT was lower in toxin positive samples compared to toxin negative, however CT ranges overlapped. Among Clarity positive samples, the quantitative toxin concentration was significantly higher in toxin positives samples as compared to toxin negative samples determined by CCNA and Quik Chek Toxin A and B. Conclusions : Laboratory tests for CDI vary in sensitivity and specificity. Quantitative toxin concentration may offer value in guiding CDI diagnosis and treatment. Presence of leukocytosis, prior antibiotic use, and previous hospitalizations may assist with CDI diagnosis while other clinical parameters may not be consistently reliable.