Knight S J, Freedman T, Puskas A, Martel P A, O'Donnell C M
TOXI-LAB, Inc, Irvine, CA 92718.
J Occup Med. 1990 Aug;32(8):715-21.
Reference laboratories are an integral component of any industrial employee drug screening program. We evaluated the performance and accuracy of the reference laboratories used by Rockwell International, a large industrial employer with an active drug screening program. A total of 829 commercially prepared control samples containing drugs of abuse, over-the-counter drugs, and analytes found in the normal employee population were disguised as routine submissions to reference laboratories used by nine Rockwell facilities. Analyses of results included compilation of false-positive and false-negative errors per drug per laboratory and a correlation of errors with preliminary and confirmation methods. Error rates of 2% false-positive results and 20% false-negative results were found for all laboratories. The errors were not limited to any one technique nor to any particular drug. The error rates observed suggest the need for routine blind testing programs and increased interaction between industry and laboratories.
参考实验室是任何工业企业员工药物筛查项目不可或缺的组成部分。我们评估了罗克韦尔国际公司（一家积极开展药物筛查项目的大型工业雇主）所使用的参考实验室的性能和准确性。总共829个商业制备的对照样本，其中包含滥用药物、非处方药以及正常员工群体中发现的分析物，被伪装成常规样本提交给罗克韦尔九个工厂所使用的参考实验室。结果分析包括统计每个实验室每种药物的假阳性和假阴性错误，并将错误与初步检测方法和确证方法进行关联。所有实验室的假阳性结果错误率为2%，假阴性结果错误率为20%。这些错误并不局限于任何一种技术或任何一种特定药物。观察到的错误率表明需要开展常规的盲测项目，并加强企业与实验室之间的互动。
