2364. Surprising Prescribing: Treatment Practices following Addition of Toxin Testing to an Existing Molecular Test for Clostridioides difficile

BACKGROUND: Diagnosis of Clostridioides difficile infection (CDI) is problematic. Adding toxin enzyme immunoassay (EIA) to molecular tests may differentiate disease from colonization and reduce treatment. Detailed descriptions of prescribing patterns following the introduction of EIA testing are not well characterized, particularly in PCR+/ EIA− patients. METHODS: In June 2018, Cleveland Clinic added EIA testing to PCR+ specimens. We conducted a retrospective cohort study on all adult inpatients who were PCR+/ EIA- from June–December 2018. Patients were placed into 3 groups for comparison: (1) complete treatment (guideline concordant); (2) incomplete treatment; and (3) no treatment. Associations with prescribing complete treatment were determined. RESULTS: We identified 240 patients (Figure 1). Mean age was 60 years, and 122 (51%) were female. Baseline conditions included many high severity comorbidities (Figure 2). 38 (16%) had history of prior CDI. 110/ 240 (46%) patients were receiving concomitant systemic antibiotics. 173 (72%) patients received complete CDI treatment, 41 (17%) incomplete treatment, and 26 (11%) none. 158/ 173 (91%) were prescribed vancomycin amounting to 2,107 days of therapy, averaging 13 PO vancomycin days each. Hematologic malignancy (P = 0.03) and leukocytosis > 15,000/ mm(3) (P = 0.04) were significantly associated with complete treatment whereas Infectious Disease (I.D.) consultation was associated with not prescribing complete treatment (P = 0.001). Prior CDI was also associated with not prescribing complete treatment but did not reach statistical significance due to small sample with prior CDI (P = 0.09). There was no association with concomitant systemic antibiotic use or use in the past 1 month. CONCLUSION: In the first 6 months after adding toxin EIA to an existing molecular test for CD, 89% of PCR+/ EIA− patients were prescribed some treatment. Hematologic malignancy and leukocytosis were associated with treatment. I.D. consultation was associated with not prescribing treatment. Although we suspect a learning curve with the 2-step approach to testing, it appears the challenge of optimizing testing and management remains. Improved understanding of the clinical significance of a sensitive test on a non-sterile specimen is required. [Image: see text] [Image: see text] DISCLOSURES: All authors: No reported disclosures.