Incidence of Bloodstream Infections, Length of Hospital Stay, and Survival in Patients With Recurrent Clostridioides difficile Infection Treated With Fecal Microbiota Transplantation or Antibiotics: A Prospective Cohort Study

Abstract

Background: Clostridioides difficile infection (CDI) is a risk factor for bloodstream infection (BSI). Fecal microbiota transplantation (FMT) is more effective than antibiotics in treating recurrent CDI, but its efficacy in preventing CDI-related BSI is uncertain.

Objective: To assess incidence of primary BSI in patients with recurrent CDI treated with FMT versus antibiotics.

Design: Prospective cohort study. Patients treated with FMT and those treated with antibiotics were matched on propensity score.

Setting: Single academic medical center.

Patients: 290 inpatients with recurrent CDI (57 patients per treatment in matched cohort).

Intervention: FMT or antibiotics.

Measurements: The primary outcome was primary BSI within 90 days. Secondary outcomes were length of hospitalization and overall survival (OS) at 90 days.

Results: Of the 290 patients, 109 were treated with FMT and 181 received antibiotics. Five patients in the FMT group and 40 in the antibiotic group developed BSI. Because of differences in the patients treated with FMT versus antibiotics in many baseline characteristics, including number of recurrences and CDI severity, comparative analyses were limited to the matched cohort. Risk for BSI was 23 percentage points (95% CI, 10 to 35 percentage points) lower in the FMT group; the FMT group also had 14 fewer days of hospitalization (CI, 9 to 20 fewer days) and a 32-percentage point increase in OS (CI, 16 to 47 percentage points) compared with the antibiotic group.

Limitation: Nonrandomized study with potential for unmeasured or residual confounding; limited generalizability of the propensity score-matched cohort.

Conclusion: In a propensity score-matched cohort, patients with recurrent CDI treated with FMT were less likely to develop primary BSI.

Primary funding source: None.