Dubowski K M, Essary N A
Department of Medicine, University of Oklahoma Health Sciences Center, Oklahoma City 73190.
J Anal Toxicol. 1992 Sep-Oct;16(5):325-7. doi: 10.1093/jat/16.5.325.
The between-run accuracy and reproducibility of vapor-alcohol control tests associated with quantitative evidential breath-alcohol testing in the field were evaluated. Control samples were generated at six separate sites with 34 degrees C TOXITEST II breath-alcohol simulators and analyzed by infrared spectrometry with Model 5000-D intoxilyzers in the recirculation mode. Control test results at target alcohol concentrations of 0.060, 0.080, 0.090, 0.100, 0.110, and 0.120 g/210 L (n = 779) were combined and analyzed by standard statistical methods. Results were correlated with target values, and the signed and absolute differences were calculated and analyzed. Data treatment included ANOVA, linear regression analysis, t statistics, and relative and cumulative frequency distributions of the differences. We found the performance of these current generation simulators in the field to be similarly satisfactory to that obtained in our laboratory evaluation. We further found that vapor-alcohol control samples generated with these devices conformed to established formal requirements and that they can serve as an effective quality assurance measure in evidential breath-alcohol analysis.
