Abstract
Clostridioides difficile infection (CDI) is one of the most common health care-associated infections, resulting in significant morbidity, mortality, and economic burden. Diagnosis of CDI relies on the assessment of clinical presentation and laboratory tests. We evaluated the clinical performance of ultrasensitive single-molecule counting technology for detection of C. difficile toxins A and B. Stool specimens from 298 patients with suspected CDI were tested with the nucleic acid amplification test (NAAT; BD MAX Cdiff assay or Xpert C. difficile assay) and Singulex Clarity C. diff toxins A/B assay. Specimens with discordant results were tested with the cell cytotoxicity neutralization assay (CCNA), and the results were correlated with disease severity and outcome. There were 64 NAAT-positive and 234 NAAT-negative samples. Of the 32 NAAT(+)/Clarity(-) and 4 NAAT(-)/Clarity(+) samples, there were 26 CCNA(-) and 4 CCNA(-) samples, respectively. CDI relapse was more common in NAAT(+)/toxin(+) patients than in NAAT(+)/toxin(-) and NAAT(-)/toxin(-) patients. The clinical specificity of Clarity and NAAT was 97.4% and 89.0%, respectively, and overdiagnosis was more than three times more common in NAAT(+)/toxin(-) than in NAAT(+)/toxin(+) patients. The Clarity assay was superior to NAATs for the diagnosis of CDI, by reducing overdiagnosis and thereby increasing clinical specificity, and the presence of toxins was associated with negative patient outcomes.