Background: A multistep algorithm using GDH antigen plus toxin with a reflex PCR is an acceptable method for detecting CDI. The use of the PCR in discordant cases can identify those patients who are colonized from those patients who have nontoxogenic strains of C. difficile. Identification of discordant patients has infection prevention implications. Treatment is not recommended for patients colonized with C. difficile. Methods: A line listing of patients with positive hospital-onset antigen/toxin positive and discordant PCR positive was created. Demographic information was extracted from medical records and the 2 cohorts were compared. Results: There were 59 discordant and 44 positive cases HO CDI cases from October 2017 through September 2019: (1) There was no difference in age and sex between the 2 groups. (2) Positive patients tended to have 3 loose stools before and after testing (57% vs 27%; P = .026). (3) Overall, 82% of positive patients had 1 of 3 signs or symptoms (leukocytosis, abdominal pain, and temperature >38°C) consistent with CDI compared to 66% of discordant patients (P = .038), and 55% of positive patients were more likely to have 2 of 3 signs or symptoms of CDI compared to 17% of discordant patients (P = .00003). (4) Also, 46% of discordant patients were either on the oncology ward or ICU compared to 32% of positive patients (P = .764). (5) There was no difference between in discordant compared to positive patients in non-CDI antimicrobial therapy within 7 days of CDI test submission (81% vs 84%, respectively). Conclusions: (1) Screening for CDI testing should include 3 loose stools and at least 2 of 3 signs or symptoms of CDI. (2) Discordant cases most likely represents colonization because only 17% of discordant patients had 2 of 3 CDI signs or symptoms at presentation. (3) Discordant cases without clinical features of CDI should not receive treatment to minimize antibiotic exposure. (4) Identification of discordant patients have infection prevention ramifications because CD can be indirectly transmitted by colonized patients; therefore, using PCR in addition to toxin testing is favored. (5) Antimicrobial therapy highly associated with CDI should be avoided, should antimicrobial therapy be necessary in PCR-positive discordant patients.

Funding: None

Disclosures: None