Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht University, Utrecht, The Netherlands.
PLoS One. 2012;7(9):e44831. doi: 10.1371/journal.pone.0044831. Epub 2012 Sep 7.
This study was performed to determine a sampling strategy to quantify the prevalence of antimicrobial resistance on veal calf farms, based on the variation in antimicrobial resistance within and between calves on five farms. Faecal samples from 50 healthy calves (10 calves/farm) were collected. From each individual sample and one pooled faecal sample per farm, 90 selected Escherichia coli isolates were tested for their resistance against 25 mg/L amoxicillin, 25 mg/L tetracycline, 0.5 mg/L cefotaxime, 0.125 mg/L ciprofloxacin and 8/152 mg/L trimethoprim/sulfamethoxazole (tmp/s) by replica plating. From each faecal sample another 10 selected E. coli isolates were tested for their resistance by broth microdilution as a reference. Logistic regression analysis was performed to compare the odds of testing an isolate resistant between both test methods (replica plating vs. broth microdilution) and to evaluate the effect of pooling faecal samples. Bootstrap analysis was used to investigate the precision of the estimated prevalence of resistance to each antimicrobial obtained by several simulated sampling strategies. Replica plating showed similar odds of E. coli isolates tested resistant compared to broth microdilution, except for ciprofloxacin (OR 0.29, p ≤ 0.05). Pooled samples showed in general lower odds of an isolate being resistant compared to individual samples, although these differences were not significant. Bootstrap analysis showed that within each antimicrobial the various compositions of a pooled sample provided consistent estimates for the mean proportion of resistant isolates. Sampling strategies should be based on the variation in resistance among isolates within faecal samples and between faecal samples, which may vary by antimicrobial. In our study, the optimal sampling strategy from the perspective of precision of the estimated levels of resistance and practicality consists of a pooled faecal sample from 20 individual animals, of which 90 isolates are tested for their susceptibility by replica plating.
本研究旨在确定一种采样策略，以基于五家农场中犊牛个体间和个体内的抗生素耐药性变化来量化犊牛养殖场中抗生素耐药性的流行率。采集了 50 头健康犊牛（每农场 10 头犊牛）的粪便样本。从每个个体样本和每个农场的一个混合粪便样本中，共检测了 90 株选定的大肠杆菌分离株对 25mg/L 阿莫西林、25mg/L 四环素、0.5mg/L 头孢噻肟、0.125mg/L 环丙沙星和 8/152mg/L 甲氧苄啶/磺胺甲噁唑（tmp/s）的耐药性，采用复制平板法进行检测。从每个粪便样本中，又通过肉汤微量稀释法检测了另外 10 株选定的大肠杆菌分离株的耐药性，作为参考。采用逻辑回归分析比较两种检测方法（复制平板法与肉汤微量稀释法）检测分离株耐药的可能性，并评估混合粪便样本的效果。采用自举法分析来研究通过几种模拟采样策略获得的每个抗生素耐药性估计流行率的精度。与肉汤微量稀释法相比，复制平板法检测到的大肠杆菌分离株耐药的可能性相似，除了环丙沙星（OR 0.29，p≤0.05）。一般来说，与个体样本相比，混合样本中分离株的耐药可能性较低，尽管这些差异并不显著。自举法分析表明，在每种抗生素中，混合样本的各种组成都为耐药分离株的平均比例提供了一致的估计值。采样策略应基于粪便样本中分离株间和粪便样本间的耐药性变化，这可能因抗生素而异。在我们的研究中，从估计耐药水平的精度和实用性的角度来看，最佳的采样策略是从 20 头个体动物中采集一个混合粪便样本，其中 90 个分离株通过复制平板法进行药敏性检测。
