Presi P, Knopf L, Regula G, Pacciarelli B, Frey J, Stärk K D C
Federal Veterinary Office, Schwarzenburgstrasse 155, CH-3003 Bern, Switzerland.
Food Addit Contam. 2007 Jun;24(6):590-7. doi: 10.1080/02652030601137346.
This paper evaluates whether the Swiss monitoring programme for foreign substances in animal products fulfils basic epidemiological quality requirements, and identifies possible sources of bias in the selection of samples. The sampling was analysed over a 4-year period (2002-05). The sampling frame in 37 participating abattoirs covered 51% of all slaughtered pigs, 73% of calves, 68% of beef and 36% of cows. The analysis revealed that some sub-populations as defined by the region of origin were statistically over-represented while others were under-represented. The programme that is in accordance with European Union requirements contained some relevant bias. Patterns of under-sampled regions characterized by management type differences were identified. This could lead to an underestimate of the number of contaminated animals within the programme. Although the current sampling was stratified and partially risk-based, its efficiency could be improved by adopting a more targeted approach.
本文评估了瑞士动物产品中外来物质监测计划是否满足基本的流行病学质量要求，并确定了样本选择中可能存在的偏差来源。对四年期间（2002 - 2005年）的抽样情况进行了分析。37个参与屠宰场的抽样框架涵盖了所有屠宰猪的51%、犊牛的73%、牛肉的68%和奶牛的36%。分析表明，按原产地区域定义的一些亚群体在统计上代表性过高，而其他亚群体则代表性不足。符合欧盟要求的该计划存在一些相关偏差。确定了以管理类型差异为特征的抽样不足地区模式。这可能导致该计划中受污染动物数量被低估。尽管当前抽样是分层的且部分基于风险，但通过采用更具针对性的方法可以提高其效率。
