| 10206041 |
Daemen E, van Risseghem M, de Bacquer D, Bulat P, Braeckman L, Vanhoorne M: Preliminary external quality assessment for the biological monitoring of carbon disulfide with urinary 2-thiothiazolidine-4-carboxylic acid. Ann Occup Hyg. 1999 Feb;43(2):125-30.
Four laboratories have participated in an external quality control assessment for the determination of 2-thiothiazolidine-4-carboxylic acid (TTCA). TTCA is used as a biomarker for exposure to CS2. Thirteen different urine samples were analyzed by each laboratory. Ten of these were spiked with known amounts of TTCA, and had either a high or intermediate creatinine content. Two samples without any TTCA were used as controls and one sample was a pool of samples of urine from five employees occupationally exposed to CS2. The latter had unknown TTCA content. For each sample, TTCA and creatinine concentration were determined. The samples were supplied in three consecutive deliveries. Several samples were offered more than once. Thus, within-laboratory variability could be established for creatinine and TTCA determination and accuracy could be determined for TTCA analysis. Within-laboratory variability was low for all laboratories for creatinine, although laboratory D seemed to have a slight downward bias. Accuracy for TTCA was good for all laboratories. No significant mean deviation from the expected TTCA value was encountered. There does not seem to be any clear influence of the TTCA concentration level of the samples on the accuracy and within-laboratory variability. Two of the four laboratories (A and C) showed lower within-laboratory variability than the other two for TTCA, although coefficients of variation between replicated samples are high for these two laboratories as well. The laboratory giving the best accuracy, gave the highest within-laboratory variability. A non-systematic, random error is probably the source of this. The results of this preliminary study indicate that analysis of TTCA, although regarded as an established biomarker, can give biases and thus negatively interfere with inferred dose-effect or dose-response relationships in occupational epidemiology. |
2(0,0,0,2) |