Reliability of roadside oral fluid testing devices for ∆ 9 -tetrahydrocannabinol (∆ 9 -THC) detection.

Driving under the influence of cannabis (DUIC) is increasing worldwide, and cannabis is the most prevalent drug after alcohol in impaired driving cases, emphasizing the need for a reliable traffic enforcement strategy. ∆ 9 -tetrahydrocannabinol (THC) detection in oral fluid has great potential for identifying recent cannabis use; however, additional data are needed on the sensitivities, specificities, and efficiencies of different oral fluid devices for detecting cannabinoids at the roadside by police during routine traffic safety enforcement efforts. At the roadside, 8945 oral fluid THC screening tests were performed with four devices: AquilaScan®, Dräger DrugTest®, WipeAlyser Reader®, and Druglizer®. A total of 530 samples screened positive for THC (5.9%) and were analyzed by liquid chromatography-tandem mass spectrometry at multiple cutoff concentrations (2 ng/mL, 10 ng/mL, and manufacturers' recommended device cutoffs) to investigate device performance. Results varied substantially, with sensitivities of 0%-96.8%, specificities of 89.8%-98.5%, and efficiencies of 84.3%-97.8%. The Dräger DrugTest® outperformed the other devices with a 96.8% sensitivity, 97.1% specificity, and 97.0% efficiency at a 5-ng/mL LC-MS/MS confirmation cutoff. The WipeAlyser Reader® had good performance with a 91.4% sensitivity, 97.2% specificity, and 96.4% efficiency. AquilaScan® and Druglizer® had unacceptable performance for cannabinoid detection, highlighted by sensitivity <13%. The choice of roadside oral fluid testing device must offer good analytical performance for cannabinoids because of its high prevalence of use and impact on road safety.