Section of Herd Health and Animal Husbandry, School of Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland, D04 W6F6.
Section of Herd Health and Animal Husbandry, School of Veterinary Medicine, University College Dublin, Belfield, Dublin 4, Ireland, D04 W6F6.
J Dairy Sci. 2017 Nov;100(11):9245-9257. doi: 10.3168/jds.2017-12985. Epub 2017 Sep 6.
Control of paratuberculosis is challenging due to the relatively poor performance of diagnostic tests, a prolonged incubation period, and protracted environmental survival. Prioritization of herd-level interventions is not possible because putative risk factors are often not supported by risk factor studies. The objective for this study was to investigate the relative importance of risk factors for an increased probability of herd paratuberculosis infection. Risk assessment data, comprehensive animal purchase history, and diagnostic test data were available for 936 Irish dairy herds. Both logistic regression and a Bayesian β regression on the outcome of a latent class analysis were conducted. Population attributable fractions and proportional reduction in variance explained were calculated for each variable in the logistic and Bayesian models, respectively. Routine use of the calving area for sick or lame cows was found to be a significant explanatory covariate in both models. Purchasing behavior for the previous 10 yr was not found to be significant. For the logistic model, length of time calves spend in the calving pen (25%) and routine use of the calving pen for sick or lame animals (14%) had the highest attributable fractions. For the Bayesian model, the overall R was 16%. Dry cow cleanliness (7%) and routine use of the calving area for sick or lame cows (6%) and had the highest proportional reduction in variance explained. These findings provide support for several management practices commonly recommended as part of paratuberculosis control programs; however, a large proportion of the observed variation in probability of infection remained unexplained, suggesting other important risks factors may exist.
由于诊断测试的性能相对较差、潜伏期长以及环境中存活时间长，副结核病的控制具有挑战性。由于假定的风险因素通常得不到风险因素研究的支持，因此无法对畜群层面的干预措施进行优先排序。本研究的目的是调查增加畜群副结核病感染概率的风险因素的相对重要性。为 936 个爱尔兰奶牛场提供了风险评估数据、全面的动物购买历史和诊断测试数据。对潜在类别分析的结果进行了逻辑回归和贝叶斯β回归分析。在逻辑和贝叶斯模型中，分别计算了每个变量的人群归因分数和方差解释比例的减少。在这两个模型中，将产房用于治疗病牛或跛牛的常规做法被发现是一个重要的解释性协变量。在过去 10 年中购买动物的行为没有被发现具有重要意义。对于逻辑模型，小牛在产房的停留时间（25%）和产房中常规用于治疗病牛或跛牛（14%）的做法具有最高的归因分数。对于贝叶斯模型，整体 R 值为 16%。干奶牛的清洁度（7%）和产房中常规用于治疗病牛或跛牛（6%）的做法对方差解释比例的减少有最大影响。这些发现为副结核病控制计划中通常推荐的几种管理措施提供了支持；然而，观察到的感染概率的变化仍有很大一部分无法解释，这表明可能存在其他重要的风险因素。
