The reliability of the webster method for measurement of TNT

Kgasoane, Joyce
Journal Title
Journal ISSN
Volume Title
Background: Organic nitro aromatic compounds are the most frequently used constituents of explosives. The most widely distributed and best-known nitro aromatic compound used in explosives is probably 2,4,6-trinitrotoluene. Though it is relatively safe to handle during the manufacturing process, its toxicity has been widely known for many years. Numerous studies have indicated that chronic exposure to TNT can lead to chronic toxicity in humans. Given the numerous toxic effects of TNT it is important to monitor exposure, by means of biological monitoring, since it is a more accurate estimation of systemic absorption then air monitoring (occupational hygiene surveys). The Webster urine test and GC-MS analytical technique for TNT metabolites were used in the munitions facility to quantify worker exposure to TNT. The Webster urine test in general presented negative results, suggesting that control measures were in place and were effective against TNT, during biological monitoring of exposed workers. However, test conducted using the GC-MS method indicated the presence of TNT metabolites in the urine samples of some of the exposed workers. Given that the two methods yielded different results, the purpose of the study was to determine, whether the Webster method was a reliable method to determine TNT exposure in workers. Objectives: This research report focused on the reliability of the Webster method to determine TNT exposure in workers. The objectives of the study were as follows: • To describe the concentration of TNT in the urine of munitions workers exposed to TNT by using the Webster method for the years 2001 to 2007. • To compare TNT concentration in the Webster method with the concentration of TNT metabolites in the GC-MS analytical technique. Methods: A retrospective descriptive study was conducted, using historical data collected between the years 2001 to 2007. For the first objective historical data from the years 2001 to 2007 were entered and tabulated on a Microsoft Excel spread sheet, to demonstrate the distribution of positive Webster urine results by means of a table and figure. For the second objective the GC-MS method for TNT urine metabolite analysis was considered to be the “gold standard” and the following steps were conducted. First, urine sample data were presented in a XY scatter diagram. Second, paired urine sample data were presented in a box and whisker plots. Third, paired urine samples were presented in a Bland-Altman scatter plot. Fourth, the association between the Webster test and the GC-MS analytical technique for TNT metabolites was examined by using Chi-square statistics. Fifth, the Webster test and the GC-MS analytical technique were measured in terms of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) Results: The results for the first objective indicate that only 1.9% of the 426 samples collected for the years 2001 to 2007 tested positive ((≥2.0mg/ℓ) i.e. indicated significant exposure. Urine samples that tested negative over this period were 98.1% i.e. all readings between 0mg/ℓ and 1.5mg/ℓ. The results of the statistical analysis for the second objective applying sensitivity and specificity indicated that the Webster test is more specific and not sensitive. The Webster test had a sensitivity of 27.2%; this is poor, indicating that a large proportion of over-exposed subjects were not identified by the Webster test. The specificity for the Webster test was 97.8%. A chi-square test showed that there were statistically significant differences between the Webster and the GC-MS methods: P value = 0.0000. The PPV was 94% for the Webster test. This indicated that 94% of workers who tested positive with Webster were truly positive. The NPV was 50% for the Webster test, thus indicating that 50% of the Webster test that were negative was truly negative. The area under the ROC (AUC) equalled to 0.80, this basically stated that the Webster test performed well in general. However it should be noted that only a small proportion of values was positive in this population and that false negatives were particularly problematic, so the AUC of 80% is a bit misleading. Discussion and conclusion: The Webster test failed to identify workers that were over-exposed even when it was zero. The sensitivity of the test was very low (27%), this is too low for a test that has serious health implications if over-exposure is missed. This study indicates that the Webster method cannot be used as applied in the munitions plant to determine TNT exposure in workers due to the toxicity of TNT exposure.
A dissertation submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfillment of requirements for the degree of Masters in Public Health – Occupational Hygiene Johannesburg, September 2018