Storage stability of simulator ethanol solutions for vapor-alcohol control tests in breath-alcohol analysis.

摘要

We conducted a one-year stability study on aqueous alcohol simulator solution, stored in sealed polyethylene bottles at 4 degrees C and at room temperature. Thirty-nine aliquots of simulator solution with a VAC target value of 0.079 +/- 0.010 g/210 L were stored and analyzed independently monthly at two locations: The University of Oklahoma HSC Toxicology Laboratories (OUHSC) and Oklahoma Department of Public Safety Alcohol and Drugs Countermeasures Unit (DPS/ADCU). Paired Intoxilyzer 5000 Evidential Breath-Alcohol Analyzers + Guth Breath Test Simulators were used to measure 20 consecutive VACs generated at 34 degrees C with individual simulator solution aliquots, followed by VAC control tests in triplicate. Control tests at OUHSC were carried out with a NIST-traceable compressed ethanolitrogen gas mixture with a validated label ethanol concentration of 0.085 g/210 L +/- 2% at 760 torr. Control tests at DPS/ADCU were carried out with simulator-generated vapor-alcohol samples with a validated VAC of 0.10 +/- 0.01 g/210 L. The measurement results were analyzed by standard statistical methods with a STATGRAPHICS Plus for Windows Version 6.0 microcomputer statistics program. No significant changes occurred during the one-year study period in the concentration of the aqueous simulator alcohol solution, as reflected by close agreement of the mean results of each set of 20 consecutive VAC analyses. The summary descriptive statistics for the VAC measurements were OUHSC (4 degrees C) mean +/- SD = 0.077 +/- 0.0004 g/210 L, median = 0.077, mode = 0.077, span = 0.075-0.079; DPS/ADCU (4 degrees C) mean +/- SD = 0.079 +/- 0.0013 g/210 L, median 0.079, mode = 0.078, span = 0.077-0.081; DPS/ADCU (Room Temp.) mean +/- SD = 0.078 +/- 0.0011 g/210 L, median = 0.078, mode = 0.078, span = 0.076-0.080. All OUHSC and DPS/ADCU VAC control tests were within acceptable target value ranges. The least squares linear regression model equations for y (= VAC, g/210) upon x (= time in months) were OUHSC (4 degrees C) y = 0.077 - 0.000021x; DPS/ADCU (4 degrees C) y = 0.079 + 0.000032x; and DPS/ADCU (Room Temp.) y = 0.078 + 0.000038x. The essentially zero slopes for all 3 models signify that no significant change occurred in the alcohol concentrations of all 39 simulator solutions aliquots measured over the 1-year study period, for either 4 degrees C or room temperature storage. The Pearson correlation coefficients for the above three regression models were R = -0.20, 0.10, and 0.13, respectively; each value is close to zero, confirming the absence of significant statistical relationship between VACs and passage of time. The coefficients of determination for the above three regression models were 4.3%, 1.0%, and 1.7%, respectively. These statistics indicate that the fitted models explain only 1 to 4% of the small total variability in the VAC as a function of time. ANOVA statistics for each of the three sets of VAC test results yield a P-value greater than 0.10, indicating that there is not a statistically significant relationship between VAC and the passage of time, at a 90% or higher confidence level. We conclude that the ethanol concentration of simulator alcohol solutions stored in sealed polyethylene bottles, at either 4 degrees C or normal room temperature, does not change significantly for at least one year after preparation, and that a correct initial VAC target value will remain valid during that time period.