Risk Factor for Prolonged Outpatient Fecal Colonization with Fluoroquinolone-Resistant Escherichia coli

Background: The prevalence of FQ-resistant E. coli (FQREC) has increased significantly in recent years. Few studies have focused on gastrointestinal tract colonization with FQREC. The duration and epidemiology of FQREC colonization in outpatients has not been investigated previously.

Objective:   1) To calculate the duration of fecal FQREC colonization among patients recently discharged from the hospital; and 2) to identify risk factors for prolonged FQREC colonization.

Methods:   From 2004-2007 we collected rectal swabs from all consenting patients who had been hospitalized at 2 academic centers at least 48 hours. We used a culture medium designed to identify a lower threshold of FQ resistance (MIC ≥0.125μg/mL to levofloxacin). All subjects colonized with FQREC were followed prospectively after hospital discharge with fecal samples obtained every 2 weeks for one year.  For FQREC isolates, we identified mutations in gyrA and parC and assessed organic solvent tolerance (OST) as a marker for efflux overexpression. Data ascertained on hospital discharge included demographics, comorbidities, and medication use (particularly antibiotic use) in the past 30 days. During the year of outpatient follow up, we also obtained data on new antibiotic use, use of indwelling devices, and changes in number of household members.

Results: Of the 353 total subjects enrolled, the mean (95%CI) age was 56 (54-57) and 166 (47%) were female. The mean (95%CI) duration of FQREC colonization was 212 days (140, 297). No baseline demographic data or comorbidities were associated with prolonged FQREC colonization. Use of any antibiotic during the preceding hospitalization was also not associated with prolonged FQREC colonization [HR (95%CI) = 1.02 (0.75, 1.4); p=0.9]. Among specific classes of antibiotics, only prior use of carbapenems was associated with prolonged colonization [HR (95%CI) = 0.24 (0.06, 0.98); p=0.04]. Prior use of FQs was not associated with duration of FQREC colonization [HR (95%CI) = 0.75 (0.47, 1.17); p=0.20]. Antibiotic use during follow up was also not associated with duration of FQREC colonization. Presence of a gyrA mutation [HR (95%CI): 0.17 (0.12, 0.24); p<0.001] and presence of a parC mutation [HR (95%CI): 0.537 (0.39, 0.73); p<0.001] were both significantly associated with a prolonged duration of FQREC colonization. On the other hand, presence of OST was associated with shorter duration of FQREC colonization [HR (95%CI): 1.39 (1.03, 1.85): p=0.03].

Conclusions: FQREC colonization persists for prolonged periods of time (mean of approximately 7 months in this study). The underlying resistance mechanism(s) is strongly associated with the duration of colonization while use of antibiotics is not. Efforts to predict duration of colonization (and identify likely long term carriers) should focus on further elucidating the relationship between the underlying resistance mechanism(s) on colonization.