384. Assessing the Multistep Approach for Clostridioides difficile Infection Diagnosis

Abstract

Abstract Background Clostridioides difficile infection (CDI) is a spore-forming anaerobic Gram-positive bacillus and the leading cause of nosocomial infectious diarrhea. 2017 IDSA/SHEA C. difficile guidelines recommend a multistep algorithm approach for CDI diagnosis. This algorithm entails stool testing via enzyme-linked immunosorbent assay (EIA) for glutamate dehydrogenase (GDH) and C. difficile toxins A/B (Tox). Discordant samples are arbitrated by nucleic acid amplification test (NAAT) for tcdB gene. GDH EIA has sensitivities >90%, while Tox EIA has sensitivities ranging from 50-90%. We aimed to evaluate the diagnostic performance of GDH and Tox EIA using the multistep algorithm as the gold standard. Methods We performed a secondary analysis of an inpatient dataset collected in a tertiary community health system in the Southern US, including 9-months of the incident population from 2019 to 2020. All adults ≥ 18-years old from Princeton Baptist Medical Center (Birmingham, AL) tested for C. difficile were included. Alere TechLab C. difficile Quik Chek Complete® was used to determine GDH and Tox EIA, and Alethia C. difficile DNA Amplification Assay®, a NAAT, for discordant samples. CDI case was defined as positivity for Tox and GDH, or positivity of either Tox or GDH in addition to positive NAAT. Retesting within the same admission was allowed. Results Overall, 288 tests were performed. Concordant positivity for GDH & Tox was found in 29 (10.1%) tests; concordant negativity in 226 (78.5%). Discordant positivity for GDH or Tox was observed in 33 (11.5%) tests; all were tested with NAAT. Only 1 of 33 (3.03%) discordant tests was GDH negative & Tox positive, which had a positive NAAT. The remaining 32 (96.97%) discordant tests were GDH positive and Tox negative, of which 18 (56.3%) were positive for NAAT (Table 1). Performance characteristics for the above tests were calculated (Table 2). Table 1:Clostridioides difficile infection case distributions per GDH and ToxinTable 2:Test performance characteristics Conclusion Taking the multistep algorithm as a gold standard, the performance characteristics of GDH and Tox are concordant to previous descriptions in the literature, with GDH being highly sensitive and Tox being highly specific. A diagnostic algorithm lacking GDH and NAAT would miss over one-third of CDI cases. A diagnostic algorithm lacking Tox and NAAT could potentially overdiagnose about one-fifth of tested patients. Disclosures All Authors: No reported disclosures.