The perceived urgency of the need to decrease C. difficile superinfections warranted a restrictive component to the fluoroquinolone intervention as determined by the ASP. Limited hospital resources including a lack of 24-hour ID coverage as well as the role of fluoroquinolones as an important non-beta-lactam antipseudomonal agent, however, precluded a more austere restriction. As a result, an immediate impact of the fluoroquinolone intervention indicative of a restrictive strategy was observed but not to the magnitude that might be expected from a strict preauthorization or formulary restriction. It is noteworthy that there was already a downward trend in fluoroquinolone usage in the 44 months immediately preceding the intervention. This likely reflects efforts by the ASP and prescribers in light of accumulating fluoroquinolone safety concerns including C. difficile. In contrast to more draconian restrictions, trends of decreasing fluoroquinolone prescribing increased in the short- and long-term periods, potentially owing to the continued reinforcement provided by the PAF component. A statistically significant decrease in fluoroquinolone consumption was observed during the long-term post-intervention period, indicating that the policy had a sustained impact. The ultimate result was a dramatic reduction of fluoroquinolone prescribing from 166.68 DDD/1000 APD in December 2011 to 14.89 DDD/1000 APD in December 2019.
The use of an ITS design is an increasingly popular method for evaluating the impact of AMS interventions allowing for statistical analysis of the magnitude of change resulting from an intervention both immediately and over time [12]. Our study had a sufficient number of data points to evaluate the effects over two separate time periods. Including a gap allowed for the evaluation of both short- and long-term trends of fluoroquinolone consumption.
The impact of the fluoroquinolone restriction policy on consumption of alternative classes of antibiotics was evaluated in order to elucidate the overall impact of the intervention on antibiotic usage (i.e., did we “squeeze the antibiotic balloon”). Shifts within the antipseudomonal beta-lactams were likely multifactorial including drug and IV fluid shortages and a similar aztreonam restriction implemented by the ASP during the study period. It is likely that the fluoroquinolone restriction contributed to the overall increase in antipseudomonal beta-lactam prescribing. It is likely that the significant upwards trend in usage of 3rd generation cephalosporins, macrolides, and tetracyclines is chiefly due to fluoroquinolones being relegated to “Alternative Therapy” status in CAP order sets as well as the impact of the sustained effort and educational campaign to reduce fluoroquinolone prescribing resulting in prescribers substituting these agents for chronic obstructive pulmonary disease (COPD) exacerbations.
P. aeruginosa susceptibilities were determined to be an appropriate microbiological marker given the clinical relevance and declining institutional susceptibility rates for levofloxacin. The increase in levofloxacin P. aeruginosa susceptibilities from 62% in 2014 to 76% in 2018 could have a significant clinical impact regarding its viability for empiric double-coverage as well as targeted therapy. Improvements in P. aeruginosa susceptibilities for antipseudomonal beta-lactams despite increased usage for most of these agents suggests that decreased levofloxacin usage may have been a contributing factor.
These results are consistent with that of Lafaurie et al, in which a sustained reduction of fluoroquinolone use (118.2 DDD/1000 PD to 77.5 DDD/1000 PD; P = 0.0002) following a primarily enabling intervention led to a significant decrease in fluoroquinolone-resistant P. aeruginosa (42–26%; P = 0.001) in a 600-bed acute care university hospital [8]. Sarma et al evaluated the effects of fluoroquinolone restriction on the resistance of Enterobacterales and observed a large decline in the percentage of ciprofloxacin-resistant ESBL-producing urinary E. coli isolates in both hospital (risk ratio, 0.473; 95% CI, 0.315 to 0.712) and community settings (risk ratio, 0.098; 95% CI, 0.315 to 0.712) and community settings (risk ratio, 0.098; 95% CI, 0.062 to 0.157) [5]. Additionally, there were no observable increases in resistance of E. coli to other antibiotics as a result of their intervention.
Limitations of this study include its single-center setting and the retrospective nature of the analysis. Determination of whether the prescriber would have ordered a fluoroquinolone if not prohibited by the intervention can only be inferred. In 2016 and 2018, the Food and Drug Administration (FDA) issued Drug Safety Communication warnings associated with fluoroquinolones [13–14]. This is a potential confounder as it could have influenced fluoroquinolone prescribing practices. Antibiotic consumption was measured by DDD using purchasing data although current recommendations from the IDSA are to measure days of therapy (DOT) using antibiotic administrations as there can be dissimilarity between administered doses and the DDD recommended by the WHO as well as a lack of precision in utilizing purchasing data. We anticipate that this potential lack of precision should not have a profound impact on long-term effects as measured in our analysis [15].
In addition, other ASP interventions such as the aforementioned aztreonam restriction would be expected to have an indirect effect on fluoroquinolone prescribing. Despite these limitations, we feel that the results of this analysis can be impactful in demonstrating that ASP interventions combining restrictive and enabling components can be effective for the immediate determined need while also having sustained consequences, can be implemented in resource limited settings, and can lead to considerable improvements in susceptibilities. In conclusion, the multifaceted approach to targeting fluoroquinolone use at a tertiary, non-teaching hospital led to both an immediate and sustained impact on fluoroquinolone consumption, with only minor effects on consumption of alternative antimicrobials with improvements in P. aeruginosa susceptibility rates.