Étude comparative des alcoolémies obtenues par la méthode officielle par chromatographie en phase gazeuse sur sang total versus méthode enzymatique-ADH automatisée Integra 800 sur plasma

We report results about a comparative study involving 3 ethanol assay used in practice clinical laboratory conditions: Gas chromatographic method (GC) on whole blood (GCWB) and on plasma (GCPL) and Roche Integra ADH-enzymatic (Integra) ethanol assay on plasma. A correlation study of blood ethanol determinations in duplicate, was conducted on 110 patients admitted at the Emergency Departement. Repeatability tests on serum control level = 0,93 g/l, provided coefficients of variation: CVs GC = 0,66% and CVs Integra = 0,70%. Reproducibility tests on the same control serum resulted in CVs GC = 1,90% and CVs Integra = 2,17%. Correlation coefficient between GCWB versus Integra and GCWB versus GCPL and GCPL versus Integra are the same: r = 0,998. Passing-Bablok equations are respectively: Y(Integra) = 1,067 × (GCWB) + 0,015 and Y (GCPL) = 1,067 × (GCWB) + 0,006 and Y (Integra) = 1,006 X (GCPL) + 0,008. In the bias corrected Bland-Altman difference plot for Integra recalculated results versus CPGWB only 7 results fell outside. Statistical results on the panel of 110 patients gave: age average = 43,7 years, 25% aged 50-60 years, average alcoholemia rate = 1,82 g/l (45% between 1,00- 2,00 g/l), 80% are males. There is no correlation between ages versus blood alcohol level: r = 0,0044. During this study GC method confirmed his good reproducibility and accuracy. Integra ADH automatic method demonstrates also his good precision. However the enzymatic method gave a positive bias of 8% ± 4 when compared to GCWB method. The bias is the same with the GCPL method. Evidence is given that this bias depends only from the types of specimens analysed: whole blood or plasma and not from the method, GC or ADH-enzymatic assay. Analytical qualities revealed by the Roche Integra ADH assay and the possibility to correlate the method with GCWB method needs to be confirmed by some others teams before to be a candidate as a third official method for blood alcohol determination in legal medicine.