Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, Agassiz, British Columbia, Canada, V0M 1A0.
J Dairy Sci. 2013 Jul;96(7):4725-37. doi: 10.3168/jds.2012-6359. Epub 2013 May 6.
A body condition score (BCS) in dairy cattle is a subjective assessment of the proportion of body fat that she possesses and is a common measure used in animal welfare assessment. The objectives of our study were to develop and implement a training program to produce highly repeatable BCS by many assessors as part of a cross-Canada epidemiological study on dairy cow comfort and welfare. In preliminary studies, we established that without any proper standard operating procedures (SOP) to describe the practical steps of the process and good standard reference for each score, assessors provided with a BCS chart scored with each other only with substantial agreement within 0.5 points and moderate agreement on exact score (mean weighted kappa coefficient=0.79 and 0.46, respectively). Detailed SOP were developed to assess BCS in 4 locations on a dairy farm. Assessing BCS presented more challenges in some locations (when cows exited the milking parlor, when the assessor was located outside the freestall pen) than others (when cows were headlocked at the feed bunk, when assessor was located inside the freestall pen). Additionally, training material and a training procedure were developed to ensure that future assessors would achieve almost perfect repeatability with the trainer within 0.5 points (weighted kappa coefficient >0.80). Twelve trainees followed this training and their repeatability was assessed using photographs in classroom sessions and live observations on farm over a 1-wk period. Repeatability was maintained above target agreement at periodic checks over the 6 mo of on-farm data collection. Two trainers were used as a reference standard to which all trainees were compared. This study demonstrates that to obtain reliable measures, a training program must include validated procedures to help assessors cope with a variety of farm setups. Regular repeatability checks are essential to ensure that the reference standard is maintained over time and to secure high data quality. This method to develop a training program as well as the training program implemented can be used as a model to successfully train on-farm assessors.
奶牛的身体状况评分（BCS）是对其体脂比例的主观评估，是动物福利评估中常用的一种衡量标准。我们的研究目的是制定并实施一项培训计划，以在加拿大奶牛舒适度和福利的跨区域流行病学研究中，由众多评估员对奶牛的 BCS 进行高度可重复的评估。在初步研究中，我们发现，如果没有适当的标准操作程序（SOP）来描述该过程的实际步骤，并且没有每个评分的良好标准参考，那么评估员仅使用 BCS 图表进行评分时，彼此之间的评分仅在 0.5 分以内具有高度一致性，在确切评分上具有中度一致性（平均加权 Kappa 系数分别为 0.79 和 0.46）。我们制定了详细的 SOP 来评估奶牛场的 4 个位置的 BCS。在某些位置（奶牛离开挤奶厅时，评估员位于畜栏外时）评估 BCS 比其他位置（奶牛在饲料槽前被锁头时，评估员位于畜栏内时）更具挑战性。此外，还开发了培训材料和培训程序，以确保未来的评估员能够在 0.5 分以内与培训师实现几乎完美的可重复性（加权 Kappa 系数>0.80）。12 名学员参加了该培训，他们在为期一周的课堂和现场农场观察中，通过照片和现场观察来评估他们的可重复性。在农场数据收集的 6 个月期间，定期检查时的可重复性仍保持在目标一致性以上。使用两位培训师作为参考标准，将所有学员与之进行比较。这项研究表明，为了获得可靠的测量结果，培训计划必须包括经过验证的程序，以帮助评估员应对各种农场设置。定期的可重复性检查对于确保随着时间的推移保持参考标准以及确保高质量的数据至关重要。这种制定培训计划的方法以及实施的培训计划，可以作为成功培训农场评估员的模式。
