Prevalence of fecal carriage of ESBL-producing Enterobacteriaceae in hospitalized and ambulatory patients during two non-outbreak periods

Abstract

Beta-lactam resistance in Enterobacteriaceae associated with plasmid-mediated extended-spectrum beta-lactamase (ESBL) has become a worldwide problem [1]. ESBLs were initially associated with nosocomial outbreaks caused by strains producing single enzymes, but recent studies have revealed a more complex situation, with a significant increase among community isolates [2]. The rate of fecal carriage of ESBL-producing bacterial isolates in nosocomial outbreaks has been studied previously [3, 4], but few prospective studies investigating the prevalence of fecal carriage during non-outbreak periods have been conducted [5, 6]. In an effort to fill this knowledge gap, we performed a study of hospitalized patients and outpatients to determine the prevalence of fecal carriage of ESBL-producing Enterobacteriaceae in stool samples submitted to our laboratory for culture during two non-outbreak periods separated by 17 months (April 2002–February 2003 and August 2004– July 2005). A total of 7,065 and 7,686 fecal samples from 4,676 and 4,708 patients were analyzed in each period, respectively. Ambulatory patients accounted for 75.5 and 70% of the total in each respective period. The samples were cultured using standard methods for Salmonella, Shigella, Campylobacter, Yersinia, Aeromonas and Plesiomonas. In 13.8% of all stool cultures collected during both periods, other nonCampylobacter strains belonging to the family Enterobacteriaceae were isolated. These grew in modified charcoal-cefazolin-deoxycholate agar containing cefoperazone 32 mg/l and amphotericin B supplement (Oxoid, Basingstoke, UK) incubated for 48 h under microaerobic conditions—a method used routinely to isolate Campylobacter spp in our laboratory. All of the non-Campylobacter strains that grew in the medium containing cefoperazone were identified according to conventional methods [7] and screened for production of ESBL by identification of the resistance phenotype and using the double-disk synergy test with cefotaxime, ceftazidime, cefepime, aztreonam and amoxicillin-clavulanic acid [8]. Results were confirmed by determining the MICs using the E-test (AB Biodisk, Solna, Sweden) with cefotaxime/cefotaxime-clavulanic acid, ceftazidime/ceftazidime-clavulanic acid and cefepime/cefepime-clavulanic acid. Strains producing ESBL were defined as strains showing synergism between clavulanic acid and cefotaxime, ceftazidime, cefepime and/or aztreonam [9]. All strains suspected of having a resistance pattern compatible with hyperproduction of the chromosomal enzymes as well as resistant strains without synergy were excluded from the study. Of the enterobacteria isolated in the modified charcoalcefazolin-deoxycholate agar, 16.3 and 26.76% (105 of 643 and 349 of 1,304 in each period, respectively) were ESBLproducing strains. Statistical significance was calculated for comparison of proportions using the chi-square test, and a p value of <0.05 was considered statistically significant (SPSS V 12.0; SPSS, Chicago, IL, USA). The rates of fecal carriage of ESBL-producing strains belonging to the Enterobacteriaceae family increased dramatically in our area from 2.3% (105 of 4,676 patients) in 2002 to 7.4% (349 of 4,708 patients) in 2004 (p<0.001). Eur J Clin Microbiol Infect Dis (2007) 26:77–78 DOI 10.1007/s10096-006-0242-8