Computer-aided headspace gas chromatography applied to blood-alcohol analysis: importance of online process control.

This paper describes the analysis of ethanol in blood specimens from suspect drunk drivers and the associated quality assurance procedures currently used in Sweden for legal purposes. Aliquots of whole blood from two separate Vacutainer tubes are diluted with 1-propanol as internal standard before analysis by headspace gas chromatography (HS-GC) with three different stationary phases: Carbopak B, Carbopak C, and 15% Carbowax 20 M. The actual HS-GC analysis, the integration of chromatographic peaks, the collection and processing of results, as well as the quality control tests involve the use of computer-aided techniques. The standard deviation of analysis(y) increased with concentration of ethanol in the blood specimen(x), and above 0.50 mg/g the regression equation was y = 0.0033 + 0.0153x. The prosecution blood-alcohol concentration (BAC) is the mean of three separate determinations made by different laboratory technicians working independently with different sets of equipment. A deduction is made from the mean analytical result to compensate for random and systematic errors inherent in the method. At BACs of 0.5 and 1.5 mg/g, which are the statutory limits in Sweden, the allowances currently made are 0.06 and 0.09 mg/g, respectively. Accordingly, the reduced prosecution BAC is less than the actual BAC with a statistical confidence of 99.9%.