Risk Factors for Fluoroquinolone Resistance in Nosocomial Urinary Tract Infections Caused by Gram-Negative Bacilli

Background: The prevalence of urinary tract infections (UTIs) caused by fluoroquinolone-resistant gram-negative bacilli (FQR-GNB) has significantly increased in recent years. Although many studies have explored risk factors for FQ resistance, the majority have focused only on UTIs caused by E. coli and/or failed to distinguish colonization and infection.

Objective: To identify risk factors fluoroquinolone resistance among nosocomial UTIs caused by GNB.

Methods: We conducted a case-control study of hospitalized patients with nosocomial GNB UTIs at 2 medical centers within the University of Pennsylvania Health System. Of subjects in whom GNB were isolated on urine culture, only those who met CDC criteria for nosocomial UTI were eligible for the study. Of these eligible subjects, all patients with urinary pathogens demonstrating a levofloxacin-MIC of ≥ 8 ug/mL were selected as cases. Among all subjects with FQ-susceptible gram-negative bacilli (FQS-GNB), controls were frequency matched to cases by month of isolation and species of infecting organism. Cases and controls were compared with regard to demographics, comorbid conditions, and use of medications (particularly antibiotics) within the 30 days prior to the UTI.

Results: A total of 247 cases with FQR-GNB and 255 controls with FQS-GNB UTIs were included from January 1, 2003 – March 31, 2005. Among 502 patients, major causative pathogens were E. coli (52.0%), P. aeruginosa (21.9%), K. pneumoniae (7.8%), Enterobacter species (6.8%) and P. mirabilis (4.6%). A total of 33.7% of patients were males and the median age was 68 (range, 22 to 94) years. Independent risk factors for fluoroquinolone resistance identified via multiple logistic regression analyses are noted in the table.

Conclusions: Recent use of FQs, cotrimoxazole and metronidazole were independent risk factors for FQ resistance. In the other hand, recent use of cefazolin (perhaps for surgical prophylaxis) which has selective activity against gram-positive pathogens was protective. These results have enlightened us that not only FQ use, but also cotrimoxazole and metronidazole should be optimized in effort to reduce the emergence of FQ resistance among GNB uropathogens. Furthermore, residence in a long term care facility was also noted to be the risk factor, suggesting that FQ resistance may be spreading across different types of healthcare facilities. Additional studies should be done to explore the extent of this problem in the long term care facilities setting.