A summary index for antimicrobial resistance in food animals in the Netherlands

The Dutch government has set targets for reduction of antimicrobial usage in food animals, stipulating a 50% reduction in usage (on a weight basis) in 2013 as compared to 2009 and a 70% decrease in 2015. A monitoring program has been instituted to evaluate the impact on antimicrobial resistance (AMR). The Dutch Ministry of Public Health Welfare and Sports has expressed the need for a summary index to present the results of the monitoring data concisely to policy makers.

We use data on AMR in bacteria from randomly collected samples from broiler chickens, fattening pigs, veal calves and dairy cows. Escherichia coli was selected for resistance monitoring because they are intrinsically susceptible to the antibiotics included in the test panel (ciprofloxacin, cefotaxime, tetracycline and ampicillin) and they are present in all samples, which facilitates proper randomization and trend analysis. The AMR summary index was calculated for each animal species as a weighted average over the four antibiotics, taking into account their clinical relevance. Weights were obtained by conjoint analysis, a pairwise comparison study involving infectious diseases professionals with clinical and public health backgrounds, with data analysis by conditional logistic regression. The AMR summary index was then computed by Monte Carlo simulation, accounting for sampling and regression uncertainty.

The highest weights (0.35) were given to ciprofloxacin and cefotaxime followed by ampicillin (0.23) and tetracycline (0.07). Throughout the years, the AMR index was highest in broiler chickens, followed by pigs and veal calves, while the lowest values were consistently recorded in dairy cows. In all animal species, the index in 2014 was significantly lower than in 2009.

We demonstrate that high-dimensional data on surveillance of antimicrobial resistance can be summarized in an index for evaluating trends between and within food animal species by a process involving decision makers and scientists to select and weight the most relevant antibiotics.