Efficacy of Oral Vancomycin, Oral Metronidazole, or IV Metronidazole Prophylaxis at Reducing the Risk of Clostridium difficile Recurrence

BACKGROUND: Secondary prophylaxis (SP) for Clostridium difficile infection (CDI) with oral vancomycin or oral/IV metronidazole when initiating antibiotics is common, though few studies are available to support this practice. The purpose of this study was to assess the efficacy of prophylaxis within a year of index CDI. METHODS: This retrospective chart review looks at subsequent courses of antibiotics and CDI in patients with initial positive CDI testing in 2013–16. A positive CDI test within 90 days of antibiotics was a recurrence. The use of antibiotics for SP was noted, along with other factors associated with CDI relapse. Non-parametric and exact tests were used for univariate analysis. These variables were included in a multivariate proportional hazards model. RESULTS: We found 597 antibiotic episodes in 230 patients. 130 episodes (21.8%) received SP. The difference of recurrence rates with and without antibiotics, 9.2 % vs 10.7%, was not statistically significant. No difference was seen when metronidazole was used, but vancomycin SP reduced the rate to 7.5% (6/80, P = 0.45). Probiotics were associated with a higher rate of recurrence (16.7 vs. 8.9%, P = 0.025). Proton pump inhibitors were also associated with a slightly higher rate of CDI recurrence (13.0% vs. 8.4%). The rate of relapse fell significantly with increasing time since the index case of CDI by logistic regression (P = 0.011). In multivariate regression, relapse was associated with shorter time from index CDI, shorter durations of antibiotics, and the use of probiotics. CONCLUSION: This retrospective study does not support the routine use of metronidazole in subsequent antibiotic courses following CDI. The use of probiotics paradoxically increased the rate of CDI relapse in this study. The limitations of this retrospective study do not eliminate the possibility of utility of vancomycin as prophylaxis, but this requires further evaluation. DISCLOSURES: All authors: No reported disclosures.