Lessons learned from implementing Clostridium difficile-focused antibiotic stewardship interventions

To determine whether controlling the prescription of targeted antibiotics would translate to a measurable reduction in hospital-onset Clostridium difficile infection (CDI) rates.

A multicenter before-and-after intervention comparative study.

Ten medical centers in the greater New York region. Intervention group comprised of 6 facilities with early antimicrobial stewardship programs (ASPs). The 4 facilities without ASPs made up the nonintervention group.

Intervention facilities identified target antibiotics using case-control studies and implemented ASP-based strategies to control their use. Pre- and postintervention hospital-onset CDI rates and antibiotic consumption were compared for a 20-month period from June 2010 to January 2012. Antibiotic usage was compared using defined daily dose, days of therapy, and number of courses prescribed. Comparisons used bivariate and regression techniques.

Intervention facilities identified piperacillin/tazobactam, fluoroquinolones, or cefepime (odds ratio, 2.0-9.8 in CDI case patients compared with those without CDI) as intervention targets and selected several interventions (all included a component of audit and feedback). Varying degrees of success were observed in reducing antibiotic consumption over time. Total target antibiotic use significantly decreased (P < .05) when measured by days of therapy and number of courses but not by defined daily dose. Intravenous moxifloxacin and oral ciprofloxacin use showed significant reduction when measured by defined daily dose and days of therapy (P ≤ .01). Number of courses with all forms of these antibiotics was reduced (P < .005). Intervention hospitals reported fewer hospital-onset CDI cases (2.8 rate point difference) compared with nonintervention hospitals; however, we were unable to show statistically significant decreases in aggregate hospital-onset CDI either between intervention and nonintervention groups or within the intervention group over time.

Although decreases in target antibiotic consumption did not translate into reductions of hospital-onset CDI in this study, many valuable lessons (including implementation strategies and antibiotic consumption measures) were learned. The findings can inform potential policy decisions regarding incorporating control of CDI and ASP as healthcare quality measures.