Driving under the influence of cocaine: quantitative determination of basic drugs in oral fluid obtained during roadside controls and a controlled study with cocaine users

Using the Belgian Drugs and Driving procedure, 36% of the cocaine-positive oral fluid (OF) screening results were not confirmed in plasma. This study investigates the impact of the choice of screening devices and confirmation matrix on the detection of cocaine use. An ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) method quantifying cocaine, benzoylecgonine (BZE), and other basic drugs in OF was developed and validated. This method monitored OF samples obtained either from a roadside (n = 12) or a double-blind controlled study with cocaine users (n = 10) who were given either a capsule containing 300 mg of cocaine-HCl or a placebo. The OF data were compared to plasma concentrations to obtain concentration-time profiles. In addition, the sensitivity and accuracy of the Drugwipe5S® was assessed. A significant difference between the OF volume collected at baseline/placebo (median 0.93 mL [range 0.43-1.92 mL]) or after cocaine-HCL intake (0.79 mL [0.30-1.21 mL]) was observed. The median OF/Plasma at the 3 collection time points were 10.7, 13.8, 6.7 for cocaine and 0.8, 1.7, 0.8 for BZE, respectively. The Drugwipe5S® detected cocaine use until at least 4 hours after intake. When applying the Belgian legal confirmation decision limit of 10 ng/mL in OF, an accuracy of 75%-98% was observed, depending on the study setting. Cocaine concentrations in OF were much higher and were detected longer as compared to plasma, when applying the same decision limit. From a toxicological viewpoint, the longer detection window with the higher sensitivity of Cocaine and BZE is beneficial to detect drivers in the crash/fatigue phase.