The PAS III passive alcohol sensor was evaluated to determine its effectiveness as a screening device to assist law enforcement officers in identifying alcohol-impaired drivers. Under laboratory conditions, the PAS III achieved its best discriminations of drinking subjects with high blood alcohol concentrations (BACs) of 0.10 percent and low BACs of 0.02 percent when the sensor was held at a distance of 5 inches from the subject. At this distance, the sensor would be expected to correctly identify 95 percent of drinking subjects with BACs of 0.10 percent, and incorrectly identify 1 percent of subjects with BACs of 0.02 percent as having a 0.10 percent BAC. The new design of the PAS III results in improved performance at further test distances compared with an earlier design. The PAS III is expected to correctly identify more subjects having a 0.10 percent BAC at a 10 inch test distance than the previous design did at 5 inches, with a reduction in the percentage of lower BAC subjects misidentified as having a high BAC. Because the BAC threshold varies by jurisdiction and class driver (i.e., adult drivers, young drivers, and commercial drivers), expected detection rates for the PAS III were also calculated for BAC's of 0.08, 0.05, and 0.02 percent. Passive alcohol sensors may be particularly important in identifying drivers at these lower BACs because behavioral evidence of impairment may not be present. A second study indicated that these results can be obtained under laboratory conditions even with relatively inexperienced sensor operators.
Abstract