Antibiotic Resistance in Pseudomonas aeruginosa Related to Formulary Quinolone Changes: An Interrupted Time Series Analysis

Background: The use of fluoroquinolones has been linked to increasing bacterial resistance based on volume of use and can involve resistance to other anti-pseudomonal agents. As risks of this effect may vary among fluoroquinolones, changes in utilization of individual fluoroquinolones, based on antibiotic formulary, in a given institution may also be related to this clinical problem. Changes in our formulary afforded an opportunity to examine possible relationships of this type. Gatifloxacin was added to our formulary in 2001, replacing levofloxacin, and moxifloxacin was in turn substituted for gatifloxacin in 2006.

Objective: To study the effect of changes in our hospital’s fluoroquinolone formulary on tobramycin, cefepime, and piperacillin/tazobactam susceptibility rates in Pseudomonas aeruginosa isolates.

Methods: Aggregated hospital-wide susceptibility data retrieved from hospital-wide antibiogram reports and aggregated hospital antibiotic use data, representing the period 2000 through 2008 were used for this analysis.  Interrupted time series analysis, which allows for assessment of the associations of an outcome attributable to a specific event in time, was used to analyze relationships. The Durbin-Watson statistic was used to test for autocorrelation. Significance was set, a priori, at 0.05.

Results: A statistically significant positive change was seen with cefepime susceptibility (increased susceptibility) when gatifloxacin was substituted for levofloxacin (p=0.0099). A positive trend was noted for piperacillin/tazobactam and cefepime susceptibility rates when gatifloxacin was replaced by moxifloxacin, but results were slightly short of statistical significance (p = 0.0589 and 0.0571). No association was detected between fluoroquinolone formulary changes and tobramycin susceptibility rates.

Conclusions: Individual quinolones may have varying influence on resistance of Pseudomonas aeruginosa to other classes of antibiotics. Such differences may have relevance to quinolone formulary and/or therapeutic decisions.