Norwegian School of Veterinary Science, Department of Production Animal Clinical Science, P.O. Box 8146 Dep., NO-0033 Oslo, Norway.
Prev Vet Med. 2012 Jun 1;105(1-2):25-37. doi: 10.1016/j.prevetmed.2012.02.011. Epub 2012 Mar 3.
The four Nordic countries Denmark (DK), Finland (FI), Norway (NO) and Sweden (SE) all have national databases where diagnostic events in dairy cows are recorded. Comparing and looking at differences in disease occurrence between countries may give information on factors that influence disease occurrence, optimal diseases control and treatment strategies. For such comparisons to be valid, the data in these databases should be standardised and of good quality. The objective of the study presented here was to assess the quality of metabolic disease recordings, primarily milk fever and ketosis, in four Nordic national databases. Completeness of recording figures of database registrations at two different levels was chosen as a measure of data quality. Firstly, completeness of recording of all disease events on a farm regardless of veterinary involvement, called 'Farmer observed completeness', was determined. Secondly, completeness of recording of veterinary treated disease events only, called 'Veterinary treated completeness', was determined. To collect data for calculating these completeness levels a simple random sample of herds was obtained in each country. Farmers who were willing to participate, recorded for 4 months in 2008, on a purpose made registration form, any observed illness in cows, regardless of veterinary involvement. The number of participating herds was 105, 167, 179 and 129 in DK, FI, NO and SE respectively. In total these herds registered 247, 248, 177 and 218 metabolic events for analysis in DK, FI, NO and SE, respectively. Data from national databases were subsequently extracted, and the two sources of data were matched to find the proportion, or completeness, of diagnostic events registered by farmers that also existed in national databases. Matching was done using a common diagnostic code system and allowed for a discrepancy of 7 days for registered date of the event. For milk fever, the Farmer observed completeness was 77%, 67%, 79% and 79%, and for Veterinary treated completeness 88%, 71%, 80% and 82%, in DK, FI, NO and SE, respectively. For ketosis, the Farmer observed completeness was 77%, 55%, 70% and 46%, and for Veterinary treated completeness 84%, 75%, 79% and 56%, in DK, FI, NO and SE, respectively. The differences observed in some of these completeness figures indicate that metabolic disease frequency measures when calculated from these four national databases do not accurately reflect the true difference in metabolic disease occurrence in dairy cows between countries. Further standardisation of disease registrations is necessary for valid between-country assessments.
丹麦（DK）、芬兰（FI）、挪威（NO）和瑞典（SE）这四个北欧国家都有奶牛诊断事件记录的国家数据库。比较和研究国家之间疾病发生的差异，可以提供影响疾病发生的因素、最佳疾病控制和治疗策略的信息。为了使这些比较有效，这些数据库中的数据应该标准化且质量良好。本研究的目的是评估四个北欧国家数据库中代谢疾病记录的质量，主要是产后瘫痪和酮病。选择数据库记录的两个不同水平的记录完整性作为数据质量的衡量标准。首先，确定无论兽医是否参与，都记录农场中所有疾病事件的记录完整性，称为“农民观察到的完整性”。其次，确定仅记录兽医治疗的疾病事件的记录完整性，称为“兽医治疗的完整性”。为了计算这些完整性水平，在每个国家都获得了一个简单的随机抽样的牛群。愿意参与的农民在 2008 年的 4 个月内，使用专门制作的登记表，记录了牛只的任何观察到的疾病，无论兽医是否参与。DK、FI、NO 和 SE 分别有 105、167、179 和 129 个参与的牛群。这些牛群共在 DK、FI、NO 和 SE 分别登记了 247、248、177 和 218 起代谢事件进行分析。随后从国家数据库中提取数据，并使用共同的诊断代码系统将两个数据源进行匹配，以找到农民登记的诊断事件中存在于国家数据库中的比例或完整性。匹配时允许登记事件的日期有 7 天的差异。对于产后瘫痪，农民观察到的完整性分别为 77%、67%、79%和 79%，兽医治疗的完整性分别为 88%、71%、80%和 82%，DK、FI、NO 和 SE。对于酮病，农民观察到的完整性分别为 77%、55%、70%和 46%，兽医治疗的完整性分别为 84%、75%、79%和 56%，DK、FI、NO 和 SE。这些完整性数据的差异表明，从这四个国家数据库计算的代谢疾病频率测量值不能准确反映奶牛产后瘫痪和酮病发生的真实差异。需要进一步规范疾病登记，以进行有效的国家间评估。
