Comparison between self-report of cannabis use and toxicological detection of THC/THCCOOH in blood and THC in oral fluid in drivers in a roadside survey.

The objective of this study was to compare the number of drivers who self-reported cannabis use by questionnaires to the results of toxicological analysis. During roadside surveys, 2957 respondents driving a personal car or van completed a questionnaire to report their use of drugs and medicines during the previous two weeks and to indicate the time of their last intake. Cannabis was analyzed in oral fluid by ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), in blood by gas chromatography-mass spectrometry (GC-MS). Frequencies in the time categories were calculated and compared with toxicological results. Diagnostic values were calculated for the time categories in which positive findings were to be expected (<4 h and <2 4h, respectively for tetrahydrocannabinol (THC) and delta9-tetrahydrocannabinol (THCCOOH) in blood, <12 h for THC in oral fluid). Most self-reported cannabis use was more than 12 h before driving. The sensitivity of the questionnaire was low, while the specificity and accuracy were high. Kappa statistics revealed a fair agreement between self-report and positive findings for THC in oral fluid and blood and moderate agreement with THCCOOH in blood. Self-report largely underestimates driving under the influence of cannabis, particularly recent cannabis use; therefore analysis of biological samples is necessary.