An On-Road Cannabis Trial’s Later Null Result Was Not a Four-Hour Driving Clearance
An on-road trial found increased lane weaving soon after THC-containing cannabis, but no statistically significant group difference during a later driving period. The later result does not establish that every person was safe to drive four hours after use.
Two driving periods answered a bounded question
Twenty-six healthy occasional cannabis users were randomized in a crossover trial involving vaporized THC-dominant, CBD-dominant, THC/CBD-equivalent and placebo cannabis. The active target doses were 13.75 mg of the relevant cannabinoids.
Participants drove an instrumented vehicle on a 100-kilometer highway route with an instructor and dual controls. The two test periods were 40–100 minutes and 240–300 minutes after administration. The main outcome was variability in lateral position, a measure of lane weaving.
Significant early differences, uncertain later differences
Both THC-containing conditions increased lane weaving relative to placebo during the earlier period. Neither produced a statistically significant difference from placebo during the later period. CBD-dominant cannabis did not significantly differ from placebo in either period.
A nonsignificant comparison means the study did not establish a difference under that analysis. It is not proof of identical performance, and the confidence intervals leave uncertainty about the size of effects. It also does not identify the precise recovery time between the two sampled periods.
Interrupted drives belong in the interpretation
Twenty-two participants completed all eight planned drives. Sixteen of the 188 test drives were terminated for safety concerns. The safety procedures were essential, but they also emphasize why this was a supervised research protocol rather than unrestricted ordinary driving.
The investigators measured plasma cannabinoids, not interchangeable whole-blood concentrations. They also measured cognitive and subjective outcomes at their own scheduled times. Those measurements should not be collapsed into one simultaneous impairment assessment.
What the trial cannot establish
The sample was young, healthy and small. The doses, route and highway task do not represent all edibles, patients, frequent users or driving demands. A clinical or forensic opinion should preserve those limits and the significant early findings while resisting an individualized clearance rule based on the later group average.
For the separate forensic question, see our discussion of cannabinoid concentrations and evidence of impairment.
Primary research
Effect of Cannabidiol and Δ9-Tetrahydrocannabinol on Driving Performance. JAMA (2020). DOI: 10.1001/jama.2020.21218.
