Comparing roadside with subsequent breath alcohol analyses and their relevance to the issue of retrograde extrapolation.

Driving while intoxicated (DWI) legislation requires proving the critical breath alcohol concentration (BrAC) at the time of driving. With time delayed analysis, retrograde extrapolation is occasionally employed but has several uncertainties associated with it. The present study attempts to address whether subjects actually arrested for DWI are likely to have BrAC values near the time of driving differing largely from those performed at a subsequent time. Selected officers arrested n = 161 subjects where roadside BrAC was determined with Pre-Arrest Breath Test (PBT) devices along with subsequent duplicate evidential analyses followed by an additional PBT analysis. These two sets of duplicates, one with large time interval (mean = 63.5 min.) and one with a 2-3 min difference, were then compared by several statistical methods. The results showing duplicate variability did not differ when the long time interval existed (F = 1.0, P > 0.05). A small but significant decrease in BrAC with respect to time appeared for the duplicate PBT data. Retrograde extrapolation applied to the data employing an assumed 0.015 g/210 l/h yielded a small but significant overestimate of the actual roadside PBT result. Finally, evidentiary analyses performed within 2 h of driving will provide good estimates and certainly not overestimates, of the BrAC existing at the time of driving and it appears that extrapolation may be unwarranted in these cases.