摘要
目的考察我国食品安全检测机构脱氧雪腐镰刀菌烯醇(DON)的实验室检测能力,分析统计方法对实验室间能力验证结果评定的影响,探讨不同实验方法的可比性。方法制备了天然污染小麦粉的高、中、低3个浓度水平的DON质量控制盲样,对全国47个省级疾病预防控制中心和粮食监测中心进行含量测定考核。根据考核结果用稳健统计算法A进行指定值、标准差和z比分数的计算;并用z比分数对参加考核实验室检测结果进行分析和评价。结果共43家检测机构按时提交了检测结果,A、B、C 3水平合格率分别是63.2%、93.8%和83.5%,考核总体合格率为79.1%。结论通过考核进行实验室间比对,了解我国DON检测水平的现状、各机构的检测水平差异和检测方法间存在的差异,探讨了不同统计学方法评价测试结果的适用性和可比性。
Objective To find out the proficiency of Chinese food safety risk surveillance agencies on determination of deoxynivalenol （DON）, evaluate different statistical methods and discuss the comparability of analytical methods used by participants, this assessment program was established. Methods Quality control samples with three different levels （A, B and C） of DON were prepared from naturally contaminated wheat flour and were dispatched to 47 agencies at random. The homogeneity and stability of the samples were tested and confirmed. The assigned values of the samples were derived by algorithm A of robust analysis. The performance of each agency was evaluated by a z - score. Results According to the results collected from 43 agencies, the pass rates at level A, B and C were 63.2 % ,93.8 % and 83.5 %, respectively, and the total pass rate was 79. 1%. Conclusion From inter - laboratory comparison, the proficiency disparity among par- ticipants and the significant difference between the analytical methods used in proficiency test can be recognized. To ensure the accuracy and reliability of the evaluation of proficiency test results, appropriate statistical methods should be applied.
出处
《中国卫生工程学》
CAS
2015年第2期97-100,107,共5页
Chinese Journal of Public Health Engineering
基金
国家国际科技合作专项(2011DFA31770)
