Problems of Forensic Sciences 2000 Vol. 42 (XLII) 209-214
STATISTICAL ESTIMATION OF HIGH ALCOHOL CONCENTRATIONS DETERMINED IN BREATH AND IN BLOOD
Dariusz ZUBA1, Wojciech GUBAŁA2, Jerzy ŁABĘDŹ2
1Faculty of Chemistry, Jagiellonian University, Cracow, Poland
2Institute of Forensic Research, Cracow, Poland
Streszczenie
In the study an attempt was made to compare the blood alcohol concentrations obtained from the analysis of breath and from direct blood investigation. Blood samples were taken from those patients of the Sobering Chamber in Cracow in whom blood alcohol concentration determined by the analysis of breath exceeded 3‰. 56 results of simultaneous determinations of alcohol in blood samples and in breath were taken for the comparison. Breath alcohol concentrations were determined using Alcomat V5, which utilises selective absorption in the IR region. Blood alcohol investigations were carried out by means of gas chromatography with the use of headspace technique. The results of breath-alcohol analysis differed from -0.83‰ to 1.37‰ in comparison with the results of direct blood investigation. The relative difference between these two variables ranged from 18.6% to 54.8% for particular patients. The correlation between the results of breath- and blood-alcohol analyses were evaluated. The correlation coefficient amounted to 0.242 and was statistically insignificant (p > 0.05). Furthermore, the slope and the intercept of the straight line adjusted to the experimental data were significantly different from the expected values (tb0 = 11.35, tb1 = 11.16, whereas t0.05,54 = 1.98). This might indicate a very low precision of the results of breath analyses at high alcohol concentrations. Moreover, the blood/breath ratios were calculated for each patient. The mean value of this coefficient for the investigated group amounted to 2063:1 ? 230 (1 standard deviation), and the coefficient of variation amounted to 11.2%. Although the mean value of the blood/breath ratio is nearly the same as the one used in Alcomat, it varies depending on the patient from 1357:1 to 2580:1. The results of the study show that there is no justification for continuation of the current practice which consists in treating alcohol blood concentrations obtained by the analysis of breath as equivalent to alcohol concentrations determined in the direct analysis of blood samples.
Słowa kluczowe
Ethanol; Breath analysis; Blood analysis.