Differences between roadside and subsequent evidential breath alcohol results and their forensic significance.

Breath alcohol measurements for forensic purposes are typically not made at the time of a driving incident but at some later time. Therefore, the magnitude of variation in breath alcohol concentration (BrAC) following the time of arrest is of concern. The use of roadside preliminary breath test (PBT) instruments can provide data on BrAC closer to the time of a driving incident and allow for comparison with later evidential analysis. This retrospective study evaluates two distributions (N = 968): differences between PBT results and the first evidential breath test (PBT-BrAC1) and differences between two (duplicate) evidential breath alcohol tests (BrAC1-BrAC2). The two distributions were shown to vary from each other and from the normal with statistical significance (p less than .05). The PBT-BrAC1 distribution had greater variability (SD = .025) than the BrAC1-BrAC2 distribution (SD = .010). An important result was that the PBT was equal to (within duplicate sampling variability) or greater than BrAC1 in approximately 85.5% of the cases. The remaining 14.5% could not be explained by sampling variability within the duplicate test distribution. The variability in both distributions typically exceeds the normally accepted alcohol elimination rates. The conclusion is that differences between roadside and subsequent evidential breath results cannot be attributed solely to absorption or elimination kinetics. Intra-individual breath sample differences can be large and thus obscure the accurate evaluation of absorption and elimination rates. Breath tests conducted within approximately 2 hours of driving will reflect, within experimental uncertainty, the BrAC at the time of driving.