2352. Clinician Assessment of Pretest Probability for Clostridioides difficile Infection and Disease Severity While Using Multiplex, Syndromic Molecular Panel in Patients Presenting with Diarrhea

Abstract

BackgroundThe role of nucleic acid amplification tests (NAAT) for diagnosing Clostridioides difficile (CD) infection remains controversial. Adding CD to multiplex molecular panels (GIPCR) that detects multiple GI pathogens of community origin, has the potential to introduce confusion leading to delayed diagnosis and unnecessary antibiotic use especially if pretest probability is not considered.MethodsWe conducted a retrospective study to determine the frequency at which clinicians characterize pretest probability and disease severity in adult patients with diarrhea who tested positive for CD by GIPCR (BioFire, Inc.) from July 1, 2017 to October 16, 2018. We excluded immunocompromised patients. Routine testing includes reflex to GDH and toxin A/B detection when GIPCR is positive for CD. Charts were reviewed and clinical suspicion (PTP) was assigned as high, medium, low, or not done. Disease severity was classified as mild, moderate and severe. Exposure to systemic antibiotic within 90 days prior to testing and stool frequency was also captured.ResultsIn total, 447 patients were included in the analysis: 110 (24.6%) were positive for both GDH and Toxin (G+/T+), 158 (35.3%) were G+/T−, 179 (40%) were G−/T−, and 149 (33%) were not classified. Toxin positivity was highest in the setting of high PTP (67%) (figure). In contrast, toxin was negative in most cases when suspicion for CDI was low or not characterized (81%). For medium suspicion, only 36% were T+. Antibiotic exposure prior to testing was observed in 203 (45%) of the cases. More G+/T+ patients received antibiotics (63%) before testing and 66% of G−/T− did not receive antecedent antibiotics. Clinicians did not characterize frequency of diarrhea in 261 (58%) of the patients tested and 95% of cases did not undergo severity classification. When documented, 24% of tested patients had < 3 diarrheal episodes/day (Table 1). Most cases where multiple pathogens were detected were T− (84.5%) and G−/T− (44%) (Table 2).ConclusionOverall, characterization of diarrheal illness was poor and PTP was frequently omitted. A large proportion of GIPCR results positive for CD (40%) were negative for both GDH and Toxin. CD results in molecular testing with syndromic panels should be interpreted with caution. Disclosures All authors: No reported disclosures.