Abstract
The European Union aggregates surveillance data for antimicrobial resistance (AMR) in livestock at the member state level. However, data evaluated nationally do not allow regionalized risk assessment and reactions. In an attempt to estimate the magnitude and granularity of regional differences, we collected AMR data for Escherichia coli by sampling fecal matter using boot swabs from vehicles carrying fattening pigs to an abattoir with a cross-regional catchment area. One E. coli strain was isolated per animal transport and its susceptibility to 14 antimicrobials was determined by broth microdilution. The minimum inhibitory concentrations obtained served to group the isolates into wild type (without a resistance phenotype) and non-wild type (with a resistance phenotype) using epidemiological cut-off values. Isolates were classified based on the origin of the respective transport into different levels of regionalization, using administrative as well as artificial borders, to obtain subnational AMR prevalence estimates for comparisons. The resistance prevalence of the total data set (n = 992) matched well with German data from national AMR surveillance. At all regional levels assessed, we detected significant differences for different compounds, including one highly significant difference for the veterinary antibiotic enrofloxacin. The antibiotics sulfamethoxazole and tetracycline, with resistance prevalence spanning 27.6%-52.4% and 18.8%-43.2%, respectively, most frequently showed significant differences in resistance prevalence. Our findings demonstrate that rapid sampling of animal transporters with boot swabs at a central node in the production chain and subsequent assignment of the regional source of the animals can serve as an approach to unveil existing subnational AMR prevalence differences with reasonable effort. IMPORTANCE: Surveillance is a key element of the World Health Organization's Global Action Plan on Antimicrobial Resistance (AMR). With livestock representing a potential source for AMR transmission, its inclusion in surveillance programs is indispensable. Governmental surveillance of livestock in the European Union only depicts nationally aggregated AMR data. However, obtaining AMR data at subnational levels, although possibly laborious, is essential to gain insights into local situations. These may differ depending on the region surveyed, but their knowledge is essential for targeted risk assessment and interventions. Our study aimed at collecting such data while simultaneously testing a time-saving sampling approach. Hence, we sampled animal transporters carrying fattening pigs to an abattoir with boot swabs and determined AMR prevalence for Escherichia coli strains isolated thereof. The significant subnational differences in AMR prevalence detected indicate that regionalized surveillance data improve and hone the AMR knowledge base. Furthermore, our sampling approach was applicable with reasonable effort.