Colistin exposure as a risk factor for infections caused by inherently colistin resistant Enterobacteriaceae—a case–control study

Objectives: Epidemiological studies have suggested an association between increased colistin use and selection for inherently colistin-resistant Enterobacteriaceae (ICRE). We aimed to evaluate whether colistin exposure is a risk factor for ICRE infection. Methods: A matched 1:1 case–control study including patients recently hospitalized for ≥14 days with ICRE infection as cases matched with similar patients with a clinical isolate of a colistin-susceptible Enterobacteriaceae as controls was performed. Univariate analysis using McNemar test and multivariate analysis were conducted to explore risk factors for ICRE isolation, including colistin exposure 90 days before the positive culture. Results: We included 446 patients, 223 cases and 223 controls matched for gender, age, department, year, source of culture and duration of hospitalization before positive culture. Colistin exposure was significantly associated with ICRE isolation in both univariate (14/223, 6.3% of cases versus 4/223, 1.8% of controls, p 0.031) and multivariate analyses (odds ratio 4.415, 95% CI 1.078–18.082). Curtailed functional capacity was a significant risk factor for ICRE as well. Exposure to other broad-spectrum antibiotics was associated with isolation of a colistin-susceptible pathogen. Conclusions: Exposure to colistin is associated with an increased risk of isolating an inherently colistin-resistant Enterobacteriaceae in patients with prolonged hospitalization. This should be taken into account while considering empirical therapy for such patients. Use of colistin should be judicious. The correlation between duration and magnitude of exposure and ICRE infection should be investigated in further studies.