Department of Clinical Chemistry, Microbiology and Immunology, Ghent University, De Pintelaan 185, Ghent, Belgium.
Clin Chim Acta. 2010 Aug 5;411(15-16):1041-5. doi: 10.1016/j.cca.2010.03.031. Epub 2010 Mar 30.
A proficiency testing scheme was set up for the DRUID (Driving under the influence of Drugs, Alcohol and Medicines) research project, funded by the European Commission, in which oral fluid is analysed by eleven laboratories. A common collection and analysis methodology is used: Statsure Saliva Sampler is used for collection and LC-MS/MS or GC-MS confirmation analysis of 22 substances is performed on all samples. Despite internal validation and quality control samples, external quality assessment is still necessary to further increase comparability of results. Four rounds of proficiency testing (PT) were organized between March 2008 and September 2009.
Qualitative results were evaluated using sensitivity and specificity. Quantitative results were evaluated using z-scores and the standard deviation of Horwitz.
Specificity was above 99% in each round, sensitivity per analyte varied between 81.7 and 100%, and 20 out of 22 analytes had a sensitivity above 90%. The percentage of satisfactory z-scores increased from 79.4% to 89.2%. This trend was seen for all drug classes, except zopiclone. Results were discussed with participating laboratories and problems were addressed.
Because of these corrective actions, DRUID laboratories have a lower variation in results than previously published PT schemes in oral fluid.
在欧盟委员会资助的 DRUID（药物、酒精和药物影响下驾驶）研究项目中，设立了一个能力验证计划，由十一个实验室对口服液进行分析。采用了共同的采集和分析方法：Statsure 唾液采样器用于采集，对所有样品进行 22 种物质的 LC-MS/MS 或 GC-MS 确证分析。尽管进行了内部验证和质量控制样品，但仍需要进行外部质量评估，以进一步提高结果的可比性。在 2008 年 3 月至 2009 年 9 月期间组织了四轮能力验证（PT）。
使用灵敏度和特异性评估定性结果。使用 z 分数和 Horwitz 标准差评估定量结果。
在每一轮中，特异性均高于 99%，每种分析物的敏感性在 81.7%至 100%之间，22 种分析物中有 20 种的敏感性高于 90%。满意的 z 分数百分比从 79.4%增加到 89.2%。这种趋势见于所有药物类别，除了佐匹克隆。与参与实验室讨论了结果，并解决了问题。
由于这些纠正措施，DRUID 实验室的结果变异比以前发表的口服液能力验证计划要小。
