Variability of the blood/breath alcohol ratio in drinking drivers

The ratio of blood-alcohol concentration (BAC) to breath-alcohol concentration (BrAC) was determined for 799 individuals apprehended for driving under the influence of alcohol (DUI) in Sweden. The BrAC was determined with an infrared analyzer (Intoxilyzer 5000S) and venous BAC was measured by headspace gas chromatography. The blood samples were always taken after the breath tests were made and the average time delay was 30 +/- 12 min (+/- SD), spanning from 6 to 60 min. The blood/breath ratios of alcohol decreased as the time between sampling blood and breath increased (F = 15.4, p < 0.001), being 2337 +/- 183 (6 to 15 min), 2302 +/- 202 (16 to 30 min), 2226 +/- 229 (31 to 45 min), and 2170 +/- 225 (46 to 60 min). When the BAC was corrected for the metabolism of alcohol at a rate of 0.019 g%/h, the mean blood/breath ratios were 2395 +/- 193 (6 to 15 min), 2416 +/- 211 (16 to 30 min), 2406 +/- 223 (31 to 45 min), and 2407 +/- 210 (45 to 60 min); no significant differences (F = 0.197, p > 0.05). The overall mean time-adjusted blood/breath ratio (+/- SD) was 2407 +/- 213 and the 95% limits of agreement (LOA) were 1981 and 2833. During 1992, 1993, and 1994, the mean blood/breath ratios of alcohol were remarkably constant, being 2409 +/- 288, 2407 +/- 206, and 2421 +/- 235, respectively, and the values were not significantly influenced by the person's age, gender, or blood-alcohol content. In 34 individuals (4.3%), the blood/breath ratio was less than 2100 after compensating for metabolism of alcohol between the times of sampling blood and breath. This compares with 156 individuals (19.6%) having a blood/breath ratio less than 2100:1 without making any correction for the metabolism of alcohol.