RESUMO
BACKGROUND: There is a paucity of data on the effect of early de-escalation of antimicrobial therapy on rates of Clostridioides difficile infection (CDI). This retrospective cohort study evaluated impact of de-escalation from antipseudomonal ß-lactam (APBL) therapy within 48 hours of Enterobacteriaceae bloodstream infections (BSIs) on 90-day risk of CDI. METHODS: Adult patients hospitalized for >48 hours for treatment of Enterobacteriaceae BSI at Palmetto Health hospitals in Columbia, South Carolina, from 1 January 2011 through 30 June 2015 were identified. Multivariable Cox proportional hazards regression was used to examine time to CDI in patients who received >48 hours or ≤48 hours of APBL for empirical therapy of Enterobacteriaceae BSI after adjustment for the propensity to receive >48 hours of APBL. RESULTS: Among 808 patients with Enterobacteriaceae BSI, 414 and 394 received >48 and ≤48 hours of APBL, respectively. Incidence of CDI was higher in patients who received >48 hours than those who received ≤48 hours of APBL (7.0% vs 1.8%; log-rank P = .002). After adjustment for propensity to receive >48 hours of APBL and other variables in the multivariable model, receipt of >48 hours of APBL (hazard ratio [HR], 3.56 [95% confidence interval {CI}, 1.48-9.92]; P = .004) and end-stage renal disease (HR, 4.27 [95% CI, 1.89-9.11]; P = .001) were independently associated with higher risk of CDI. CONCLUSIONS: The empirical use of APBL for >48 hours was an independent risk factor for CDI. Early de-escalation of APBL using clinical risk assessment tools or rapid diagnostic testing may reduce the incidence of CDI in hospitalized adults with Enterobacteriaceae BSIs.