Gibbons J F, Boland F, Egan J, Fanning S, Markey B K, Leonard F C
School of Veterinary Medicine, University College Dublin, Dublin, Ireland.
School of Mathematical Sciences, University College Dublin, Dublin, Ireland.
Zoonoses Public Health. 2016 May;63(3):241-50. doi: 10.1111/zph.12225. Epub 2015 Sep 10.
Antimicrobial use and resistance in animal and food production are of concern to public health. The primary aims of this study were to determine the frequency of resistance to 12 antimicrobials in Escherichia coli isolates from 39 pig farms and to identify patterns of antimicrobial use on these farms. Further aims were to determine whether a categorization of farms based on the duration of in-feed antimicrobial use (long-term versus short-term) could predict the occurrence of resistance on these farms and to identify the usage of specific antimicrobial drugs associated with the occurrence of resistance. Escherichia coli were isolated from all production stages on these farms; susceptibility testing was carried out against a panel of antimicrobials. Antimicrobial prescribing data were collected, and farms were categorized as long term or short term based on these. Resistance frequencies and antimicrobial use were tabulated. Logistic regression models of resistance to each antimicrobial were constructed with stage of production, duration of antimicrobial use and the use of 5 antimicrobial classes included as explanatory variables in each model. The greatest frequencies of resistance were observed to tetracycline, trimethoprim/sulphamethoxazole and streptomycin with the highest levels of resistance observed in isolates from first-stage weaned pigs. Differences in the types of antimicrobial drugs used were noted between long-term and short-term use farms. Categorization of farms as long- or short-term use was sufficient to predict the likely occurrence of resistance to 3 antimicrobial classes and could provide an aid in the control of resistance in the food chain. Stage of production was a significant predictor variable in all models of resistance constructed and did not solely reflect antimicrobial use at each stage. Cross-selection and co-selection for resistance was evident in the models constructed, and the use of trimethoprim/sulphonamide drugs in particular was associated with the occurrence of resistance to other antimicrobials.
动物及食品生产中的抗菌药物使用与耐药性问题关乎公众健康。本研究的主要目的是确定来自39个养猪场的大肠杆菌分离株对12种抗菌药物的耐药频率，并识别这些猪场的抗菌药物使用模式。进一步的目的是确定基于饲料中抗菌药物使用时长（长期与短期）对猪场进行分类是否能够预测这些猪场耐药性的发生，并识别与耐药性发生相关的特定抗菌药物的使用情况。从这些猪场的所有生产阶段分离出大肠杆菌；针对一组抗菌药物进行药敏试验。收集抗菌药物处方数据，并据此将猪场分为长期或短期使用两类。列出耐药频率和抗菌药物使用情况。构建每种抗菌药物耐药性的逻辑回归模型，将生产阶段、抗菌药物使用时长以及5类抗菌药物的使用情况作为每个模型中的解释变量。观察到对四环素、甲氧苄啶/磺胺甲恶唑和链霉素的耐药频率最高，在第一阶段断奶仔猪的分离株中观察到的耐药水平最高。长期和短期使用猪场之间在使用的抗菌药物类型上存在差异。将猪场分为长期或短期使用足以预测对3类抗菌药物耐药性的可能发生情况，并有助于控制食物链中的耐药性。在构建的所有耐药模型中，生产阶段都是一个显著的预测变量，且并不完全反映每个阶段的抗菌药物使用情况。在构建的模型中，耐药性的交叉选择和共同选择很明显，特别是甲氧苄啶/磺胺类药物的使用与对其他抗菌药物的耐药性发生有关。
