Antimicrobial stewardship programs (ASP) restricting fluoroquinolone (FQ) prescriptions have been shown to reduce hospital-onset Clostridium difficile infection (HO-CDI) rates. The impact of ASP promotion of doxycycline, proposed to have a protective effect, on HO-CDI rates is unknown. We conducted a promotional campaign containing both these initiatives within a non-ICU CAP guideline and measured the impact on antibiotic use and the HO-CDI rate.
We performed a single-center, prospective study comparing antibiotic use for adult inpatients before (October 2012 through December 2013) and after (January 2014 through March 2015) an ASP campaign consisting of non-ICU CAP guideline development, provider education, and prospective audit and feedback. Recommended empiric first-line therapy included ceftriaxone plus doxycycline, with moxifloxacin reserved only for those with a severe beta-lactam allergy or a recent history of beta-lactam or doxycycline treatment failure. No antibiotics were restricted and no changes in electronic order sets were made. Primary outcomes included antibiotic days of therapy (DOT)/1,000 patient days (PD) of moxifloxacin, doxycycline, ceftriaxone, and azithromycin, with fluconazole used as a non-equivalent dependent variable. The incidence of HO-CDI as defined by National Healthcare Safety Network was a secondary outcome. Mann Whitney U tests were applied where appropriate and the significance level was set to p<0.05.
Moxifloxacin and azithromycin DOT significantly decreased from 26.0/1,000PD to 10.9/1,000PD (p=0.007), and 43.1/1,000PD to 26.6/1,000PD (p=0.007), respectively. Doxycycline DOT significantly increased from 8.0/1,000PD to 25.6/1,000PD (p=0.007). No significant change in ceftriaxone or fluconazole DOT occurred between the two time periods. The incidence of HO-CDI declined from 8.6 cases/10,000PD pre-intervention to 6.8 cases/10,000PD post-intervention (p=0.095).
A non-restrictive ASP campaign effectively persuaded providers to use lower-risk antibiotics for CAP and succeeded in significantly reducing moxifloxacin use while significantly increasing doxycycline prescriptions. This may have contributed to a reduction in HO-CDI rates.
G. Dumyati, None
M. L. Brundige, None