Fluoroquinolones (FQs) are commonly prescribed for urinary tract and respiratory infections. Overprescription of FQs may cause collateral damage, such as increased rates of FQ resistance in gram-negative bacilli and Clostridiodes difficile infections.Reference Paterson1 Fluoroquinolone use has been associated with tendon rupture, QTc prolongation, aortic aneurysm rupture, and aortic dissection. Selective reporting of antibiotic susceptibility may reduce inappropriate and unnecessary antibiotic prescriptions when only first-line agents are reported, especially for samples in which colonization is possible (eg, urine or sputum). Reporting of antibiotic susceptibility may better guide prescribers to choose the most appropriate antibiotic; this strategy has been shown to effectively improve susceptibility rates in some settings.Reference Al-Tawfiq, Momattin, Al-Habboubi and Dancer2,Reference Langford, Seah, Chan, Downing, Johnstone and Matukas3 Implementation of this strategy may have challenges, such as the lack of human resources and healthcare system support.Reference Pulcini, Tebano and Mutters4 Selective reporting may have little impact on empiric antibiotic use.
The National University Hospital (NUH) in Singapore is a 1,200-bed tertiary-care hospital. As part of a robust antimicrobial stewardship program (ASP), selectively reporting of FQ susceptibility started in 2011, and results were released when requested by a clinician. The disclosure of FQ susceptibility results was sometimes delayed, particularly after office hours and on weekends, even when FQ use may have been appropriate. This practice was reviewed, and in April 2016, the practice of selective reporting was halted due to manpower constraints. In this study, we analyzed the impact of this change on prescribing patterns and resistance rates.
Methods
Study design
We conducted a time series analysis to compare FQ use between 2 time periods: (1) when selective reporting was in place (ie, FQ masked), and (2) after selective reporting was halted (ie, FQ unmasked) starting in April 2016. We conducted a chart review from 2015 to 2017 of patients receiving intravenous and oral FQ prescriptions for culture-directed therapy in all inpatient adult wards to determine appropriateness of therapy. Patients <18 years old and patients enrolled in clinical trials during the study period were excluded. Aggregate data on inpatient FQ use, ciprofloxacin resistance (the predominant FQ prescribed in this setting), and inpatient C. difficile rates were collected between 2014 and 2018 for comparison during masked and unmasked periods.
Inpatient FQ utilization data, measured in defined daily doses (DDD) per 100 inpatient days, were collected from electronic pharmacy dispensing records and were tabulated monthly from April 2014 to December 2018.
Appropriate versus inappropriate fluoroquinolone use in culture-directed therapy
We conducted chart reviews of inpatients regarding culture-directed FQ treatment. Culture-directed therapy was defined as antibiotics administered (newly prescribed or continued) on days 3–7 after a positive microbiological culture. Fluoroquinolone use was deemed appropriate if alternative agents had poor site penetration, were unsuitable (eg, allergy, intolerance, or unacceptable risk of adverse events, particularly renal impairment), or if carbapenems were the only alternative agents. Oral FQ use was considered appropriate if it had been prescribed within 48 hours of discharge and was the only suitable oral agent. Fluoroquinolone use was deemed inappropriate if other suitable antimicrobial agents were available based on site of infection and patient factors. Fluoroquinolone prescriptions for culture-directed therapy were audited for appropriateness for a 2-month period each year (May and June) in 2015, 2016, and 2017.
Fluoroquinolone resistance
Ciprofloxacin susceptibility rates for inpatient clinical isolates of Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa were reviewed from laboratory records between January 2014 and December 2018. Duplicate isolates within a calendar year were excluded. Susceptibility testing was routinely performed using the Vitek 2 system (bioMérieux, Marcy l’Etoile, France). Due to changes in ciprofloxacin breakpoints during this period, interpretation of minimum inhibitory concentration was standardized using European Committee on Antimicrobial Susceptibility Testing (EUCAST) 2019 break points.
Clostridium difficile rates
New inpatient C. difficile cases between January 2014 and December 2018 were obtained from surveillance data, per 10,000 patient days.
Outcomes
The outcomes of this study were comparisons of inpatient consumption of FQs in DDD per 100 inpatient days, inappropriate FQ prescriptions for culture-directed therapy, rates of ciprofloxacin susceptibility, and rates of inpatient C. difficile between the 2 periods: FQ susceptibilities “masked” versus “unmasked.”
Statistical analysis
Rates of FQ consumption, appropriate usage, and C. difficile rates between “masked” and “unmasked” periods were assessed using independent 2-sample t tests. Results were reported as incidence rate ratios (RR) and 95% confidence intervals (CI). Percentages for organism susceptibility were compared using the Fisher exact test. All statistical tests were 2-sided, and P < .05 was considered statistically significant. Analyses were conducted using STATA version 12 software (StataCorp, Cary, NC).
Results
In total, 438 patient charts were reviewed; patient demographics and diagnoses are reported in Table 1. The total number of culture-directed FQ prescriptions increased from 95 in 2015 to 153 in 2016 and 190 in 2017; the proportion of inappropriate prescriptions also increased. Overall, inappropriate FQ prescriptions for culture-directed treatment increased from 0.069 to 0.173 prescriptions per 100 inpatient days after FQ unmasking, with a relative risk (RR) of 2.50 (95% confidence interval [CI], 1.84–3.39; P < .001). Fluoroquinolone prescriptions for culture-directed treatment comprised only 20% of total FQ prescriptions over the study period.
Table 1. Patient Demographics, Antimicrobial Indications and Appropriateness, in Patients on Culture-Directed Fluoroquinolone (FQ) Therapy
a Epididymitis, tubular-ovarian abscess, line sepsis, head and neck infections, and febrile neutropenia.
Although usage of other commonly prescribed inpatient antibiotics remained fairly constant over the study period, inpatient consumption of FQs decreased from 10.90 to 9.93 DDD per 100 inpatient days after FQ unmasking (Fig. 1). The RR of 0.91 (95% CI, 0.90– 0.92; P < .001) showed statistical significance.
Fig. 1. Inpatient fluoroquinolone (FQ) usage (defined daily doses per 100 inpatient days), compared to other commonly prescribed inpatient antimicrobials; before and after unmasking of FQ susceptibilities.
FQ usage (masked) 
FQ usage (unmasked)
Average FQ usage (masked) 
Average FQ usage (unmasked)
Pipercillin-tazobactam usage 
Amoxicillin-clavulnate usage
Ceftriaxone usage
Ciprofloxacin susceptibility of E. coli decreased from 56.8% to 55.8% (95% CI, 0.881–1.044; P = .335), K. pneumoniae decreased from 66.1% to 63.0% (95% CI, 0.783– 0.973; P = .013) and P. aeruginosa increased from 83.9% to 85.4% (95% CI, 0.941–1.328; P = .197). Changes in E. coli and P. aeruginosa susceptibility were not statistically significant; however, the change in susceptibility of K. pneumoniae had a RR of 0.873, which was statistically significant.
Inpatient C. difficile infections decreased from 7.64 to 5.35 per 10,000 patient days during the study period. The RR of 0.70 (95% CI, 0.62–0.79; P < .001) showed statistical significance.
Discussion
Although the unmasking of FQ susceptibility seemed to drive a rise in inappropriate culture-directed prescribing, the number of culture-directed prescriptions did not drive up overall FQ usage and overall inpatient FQ use actually decreased over the study period. Possibly, the updated black box warning issued by the US Food and Drug Administration5 in July 2016 played a role in tempering the overall usage of FQs.
Selective reporting of antibiotic susceptibility results is one of the recommended strategies in the Infectious Diseases Society of America’s guidelines for implementing an ASP.Reference Barlam, Cosgrove and Abbo6 In our study, the masking of FQ susceptibilities may have limited their use in culture-directed therapy, but a dramatic rise in other unintended consequences was not observed when this practice was halted. No discernible changes in ciprofloxacin resistance among E. coli or P. aeruginosa isolates was detected, overall FQ usage remained largely unchanged, and C. difficile rates were not adversely affected. Ciprofloxacin resistance among K. pneumoniae increased slightly, but similar trends have been noted elsewhere.7,Reference Sanchez, Master and Clark8
Because culture-directed use was a minor component of usage (20% of inpatient FQ prescriptions) in our setting, the benefits of masking FQ sensitivities were obscured by the large number of FQs prescribed for empiric therapy. The most significant gains from an ASP perspective are likely related to the control of empiric use by other strategies, such as audits with feedback,Reference Elligsen, Walker and Pinto9 improved uptake of empiric guidelines, education of prescribers, and/or prior authorization.Reference Lewis, Fang and Gooch10
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2020.74
Financial support
No financial support was provided relevant to this article.
Conflicts of interest
All authors report no conflicts of interest relevant to this article.
