Decreasing prevalence of contamination with extended-spectrum beta-lactamaseproducing Enterobacteriaceae (ESBL-E) in retail chicken meat in the Netherlands

Pepijn Huizinga, Marjolein Kluytmans Van Den Bergh, John W. Rossen, Ina Willemsen, Carlo Verhulst, Paul H.M. Savelkoul, Alexander W. Friedrich, Silvia García-Cobos, Jan Kluytmans

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Retail chicken meat is a potential source of extended-spectrum beta-lactamase-producing Enterobacteriaceae (ESBL-E). In the past decade, vast national efforts were undertaken to decrease the antibiotic use in the veterinary sector, resulting in a 58% decrease in antibiotic sales in the sector between 2009 and 2014. This decrease in antibiotic use was followed by a decrease in ESBL-E prevalence in broilers. The current study investigates the prevalence of contamination with ESBL-E in retail chicken meat purchased in the Netherlands between December 2013 and August 2015. It looks at associations between the prevalence of contamination with ESBL-E and sample characteristics such as method of farming (free-range or conventional), supermarket chain of purchase and year of purchase. In the current study, 352 chicken meat samples were investigated for the presence of ESBL-E using selective culture methods. Six samples were excluded due to missing isolates or problems obtaining a good quality sequence leaving 346 samples for further analyses. Of these 346 samples, 188 (54.3%) were positive for ESBL-E, yielding 216 ESBL-E isolates (Escherichia coli (n = 204), Klebsiella pneumoniae (n = 11) and Escherichia fergusonii (n = 1)). All ESBL-E isolates were analysed using whole-genome sequencing. The prevalence of contamination with ESBL-E in retail chicken meat decreased from 68.3% in 2014 to 44.6% in 2015, absolute risk difference 23.7% (95% confidence interval (CI): 12.6% - 34.1%). The ESBL-E prevalence was lower in free-range chicken meat (36.4%) compared with conventional chicken meat (61.5%), absolute risk difference 25.2% (95% CI: 12.9% - 36.5%). The prevalence of contamination with ESBL-E varied between supermarket chains, the highest prevalence of contamination was found in supermarket chain 4 (76.5%) and the lowest in supermarket chain 1 (37.8%). Pairwise isolate comparisons using whole-genome multilocus sequence typing (wgMLST) showed that clustering of isolates occurs more frequently within supermarket chains than between supermarket chains. In conclusion, the prevalence of contamination with ESBL-E in retail chicken in the Netherlands decreased over time; nevertheless, it remains substantial and as such a potential source for ESBL-E in humans.

Original languageEnglish
Article numbere0226828
JournalPLoS ONE
Volume14
Issue number12
DOIs
Publication statusPublished - 1 Jan 2019

Cite this