An Urgent Scenario Drew Different Answers About Driving After Cannabis
A statement that someone would drive after cannabis can reflect the reason offered for the trip, as well as how that person feels. In a small human pilot study, participants were much more willing to drive for an imagined emergency than for less compelling reasons. The experiment measured stated willingness, not whether anyone could drive safely.
The researchers asked three different questions
The pilot in Robbe and O’Hanlon’s 1993 report enrolled 24 cannabis-experienced adults. One participant was excluded because plasma analyses did not show the expected pattern of cannabinoid exposure, leaving 23 for analysis. These were selected volunteers, and all had previously acknowledged driving after cannabis use.
Participants smoked in their own fashion, for up to fifteen minutes, until reaching their desired effect. They could use portions of up to three cigarettes containing 2.57% THC. This was not a fixed absorbed dose: cigarette-weight calculations included THC lost during burning and in smoke that did not enter the participant.
Before and repeatedly after smoking, participants answered whether they would drive for three reasons: taking a friend to another party; taking a mildly sick friend home when a taxi was available; or taking a severely sick infant to a hospital. Those scenarios differed in urgency and in the alternatives described.
Urgency mattered while participants still felt high
The results section described approximately one in five participants as willing to drive for the less compelling scenarios near the peak subjective response, compared with approximately four in five for the urgent scenario. These are approximate descriptions from a small selected sample, not population estimates.
Willingness subsequently increased across the session. The overall tests of change over time were significant for all three scenarios. However, the separate comparisons with the pre-smoking baseline did not find a significant change at any individual post-smoking assessment for the urgent scenario. The less compelling scenarios showed significant reductions from baseline at some assessments.
Those findings can coexist because the overall time test and each baseline comparison ask different statistical questions. Neither analysis establishes that participants’ willingness accurately identified their driving ability.
The pilot did not measure a driving deficit
Participants performed laboratory tracking and hand-steadiness tasks, but they did not drive during this pilot. Critical tracking did not show a significant overall time effect. Hand steadiness worsened significantly at the first post-smoking test, while later comparisons were nonsignificant.
The pilot had no placebo comparison. The authors explicitly cautioned that practice, habituation or procedural issues could complicate interpretation and that the design did not support conclusions about marijuana’s effects on performance. The later laboratory null results therefore cannot establish a recovery time or validate a participant’s decision to drive.
A hypothetical answer is not an observed emergency response
The investigators did not randomly assign real emergencies or observe whether participants actually drove in those circumstances. The scenarios also changed more than urgency alone. The findings therefore do not isolate a pure causal effect of urgency, establish an emergency-driving failure rate or provide a legal justification for a particular trip.
They do show why the exact question matters. “Would you drive?” is incomplete evidence unless the reviewer knows the circumstances described, when the answer was obtained and whether functional performance was measured separately.
Keep exposure, choice and function separate
THC and its carboxy metabolite were measured in plasma separated from blood samples, not in interchangeable whole blood. Those exposure measurements were separate from subjective intoxication, hypothetical choices and laboratory performance. None supplied an actual driving result for this pilot.
The small sample, older cigarette formulation, selected prior users and lack of a control condition limit generalization. These participants also took part in the report’s subsequent restricted-highway experiment, so the two phases should not be counted as independent samples.
A separate, later randomized simulator trial addressed a different question: how willingness changed while measured driving performance recovered. See cannabis users’ driving confidence and simulator performance. The pilot discussed here concerns the scenario attached to the willingness question.
Primary research
Robbe HWJ, O’Hanlon JF. Marijuana and Actual Driving Performance. National Highway Traffic Safety Administration, November 1993. Report DOT HS 808 078, Chapter 3. doi:10.21949/1525346.
