BackgroundSecondary 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.MethodsThis 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.ResultsWe 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.ConclusionThis 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.

BackgroundWhile current guidelines suggest a total treatment duration of 7 to 14 days for gram-negative bloodstream infections (GN-BSI), there is mounting evidence to suggest that shorter durations may be sufficient. This study compared the treatment outcomes of patients who received short duration therapy (6–10 days) with those who received long durations (11–15 days).MethodsThis was a retrospective study of adult patients who grew an aerobic gram-negative organism from a blood culture while admitted at Strong Memorial Hospital between May 2016 and May 2018. The primary outcome was a composite of mortality and relapsed GN-BSI with the same organism within 90 days of index culture. Secondary outcomes included clinical resolution at end of therapy (EOT), length of stay (LOS), 30-day readmission rate, Clostridoides difficile infection (CDI), development of recurrent GN-BSI resistant to prior antibiotic therapy, and development of multi-drug-resistant (MDR) GN-BSI within 90 days. Appropriate therapy was defined as an antibiotic with confirmed in vitro susceptibility that was either parenteral or a highly bioavailable oral antibiotic (fluoroquinolones or sulfamethoxazole–trimethoprim).ResultsOf 600 patients screened, 116 were included in the long duration group and 34 patients in the short-duration group. The majority of patients had a urinary source of infection (59.3%). The primary composite outcome occurred in 11.8% of the short duration group compared with 10.3% in the long (P > 0.999). There was no difference in clinical resolution at EOT, LOS, or rates of CDI, MDR GN-BSI, recurrent GN-BSI resistant to prior therapy, or 30-day readmission. Patients in the long duration group were discharged with longer appropriate outpatient courses (8 days vs. 0.5 days, P < 0.001), which remained significant when including lower bioavailability agents (e.g., oral β lactams) (8 days vs. 5 days, P < 0.001).ConclusionThere was no difference in clinical outcomes between the long and short duration therapy for treatment of GN-BSI. This study may support shorter treatment durations for uncomplicated GN-BSI, but should be interpreted cautiously given the smaller sample size.Disclosures All authors: No reported disclosures.

No abstract