Blood alcohol : forensic analysis by GC-MS and investigation of some pre-analytical factors that may influence the result

Abstract

Forensic laboratories are frequently called upon to determine blood alcohol concentrations, especially as a result of roadside testing. The reliability of the experimentally determined concentrations is being called into question due to the need to transport and store blood specimens – sometimes for months – before analysis can occur. Of specific concern are the length of time the specimens are stored and the temperature at which they are kept. Recently, the possible presence of certain micro-organisms has also been used as a defence as to why a blood alcohol concentration could be found to be elevated above the South African legal limit of 0.05 g/100 mL. Micro-organisms, such as Candida albicans, are known to ferment glucose to ethanol, and thus could potentially artificially increase the ethanol concentrations in contaminated blood specimens. In an effort to prevent this, blood specimens in South Africa are required to be stored with 1 g/100 mL sodium fluoride, and the laboratory processing the sample is required to show that this is the case. In this study, a novel Isotope Dilution Gas Chromatography-Mass Spectrometry method for the determination of ethanol concentration in human blood was developed. Ethanol-d6 was used as internal standard, the analytes derivatized with pentafluorobenzoyl chloride, and the resulting esters detected in single ion monitoring mode. The method was validated based on figures of merit and the expanded measurement uncertainty was determined. Following this, a method for the detection and quantitation of free fluoride in blood by means of fluoride ion selective electrode was developed. It was found that diluting specimens 20 times with Total Ionic Strength Adjustment Buffer II and deionized water sufficiently minimised matrix effects to allow aqueous calibration from 0.25 g/100 mL sodium fluoride to 3.00 g/100 mL sodium fluoride. The method was validated based on figures of merit and the expanded measurement uncertainty was determined. Using this method, the complexation effects of Fe3+, Mg2+ and Ca2+ ions were investigated, and the effect of temperature on complexation of fluoride with Fe3+ was explored. Throughout the study, the fluoride concentrations of specimens were ascertained utilising the in-house developed fluoride ion selective electrode method. Blood specimens spiked at 0.02 g/100 mL, 0.05 g/100 mL and 0.3 g/100 mL ethanol were placed in evacuated tubes containing fluoride – at least 100 mg – as preservative, and stored at room temperature (12 ± 6 °C) and under refrigeration (4 ± 3 °C) for 29 weeks. Using the novel Gas Chromatography – Mass Spectrometry method, the ethanol concentration was monitored on a weekly basis to investigate the stability thereof, as well as its dependence on temperature. The expanded measurement uncertainty was used to gauge the significance of any changes observed. The ethanol concentration of those specimens stored under refrigeration showed no significant deviation from the initial spiked concentration value for the 29 weeks; however, a decreasing trend was observed from week 25 for all three levels. The specimens stored at room temperature exhibited a significant decrease in ethanol concentrations. This was particularly evident in the specimens initially spiked at 0.02 g/100 mL. At all three levels, the ethanol concentration was statistically stable for nine weeks. Upon termination of the study, the concentrations of the medium and higher levels showed a continuing decreasing trend, while the lower level was completely depleted. Fresh blood specimens were then prepared at 0.02 g/100 mL and 0.05 g/100 mL ethanol, divided into evacuated tubes with and without fluoride, and inoculated at five levels of Candida albicans. These were then stored at room temperature and under refrigeration as before and the ethanol concentration as well as the Candida albicans cell count were monitored weekly for up to nine weeks. Specimens stored under refrigeration exhibited statistically stable ethanol concentrations, with no significant deviation from the initial spiked values, while those stored at room temperature in the absence of fluoride showed a marked decrease in ethanol concentration. It was noted that the temperature at which specimens are stored has a greater impact on the ethanol concentration stability than the presence or absence of fluoride, although specimens containing fluoride exhibited greater stability than those that did not. This study showed that the ethanol concentration of stored blood specimens does not increase over time as is currently being claimed, but rather decreases. This is to the advantage of the defendant. It is nevertheless recommended that all blood specimens for the determination of blood alcohol concentrations should be stored below 6 °C in the presence of at least 1 g/100 mL fluoride.