A recent meta-analysis published in Heliyon challenges the reliability of oral fluid (OF) testing for assessing cannabis exposure and impairment. Researchers from the Canadian Institute for Substance Use Research at the University of Victoria in British Columbia evaluated data from seven studies, involving 116 participants, and found that OF tests often yield inconsistent results, even among individuals consuming similar amounts of cannabis.
The findings indicate that there is substantial variability in THC concentrations among subjects given the same dosage, leading to unreliable readings that detract from the effectiveness of OF tests. The researchers noted, “Research shows a high degree of variability in THC concentrations for subjects given the same amount of cannabis.” This variability often results in extreme outliers in THC levels, undermining the credibility of OF tests as a tool for measuring cannabis impairment.
The analysis also highlighted that individuals who smoke cannabis are significantly more likely to test positive on OF tests compared to those consuming cannabis-infused products orally. Furthermore, the study revealed that some individuals may test positive for cannabis exposure for over 24 hours after smoking, which can occur well beyond the timeframe of any perceptible intoxication.
The authors concluded that the validity of OF tests is not adequate for detecting either prior cannabis use or impairment, particularly at the commonly used THC cut-off of 1 ng/mL. They asserted, “OF tests should not be considered a valid indicator of [cannabis-induced] impairment.”
In 2023, U.S. federal transportation regulators introduced new rules allowing commercially licensed drivers to opt for oral fluid testing instead of mandatory urinalysis. However, these rules have not yet been implemented in federal workplaces. Advocates, including NORML, argue that the presence of THC or its metabolites in bodily fluids—whether in blood, urine, or saliva—should not be assumed to indicate impairment, as these markers do not reliably predict performance deficits.
Instead, NORML supports the adoption of performance-based assessments, such as DRUID or Predictive Safety’s AlertMeter, which measure cognitive abilities against either an individual’s previous performance or a general baseline. This approach aims to provide a more accurate representation of a person’s ability to perform tasks while under the influence of cannabis.
The complete study, titled “Concentrations of delta-9-tetrahydrocannabinol in oral fluid at different time points after use: An individual participant meta-analysis,” is available in Heliyon. More information can also be found in the NORML Fact Sheet, ‘Marijuana Legalization and Impact on the Workplace.’
English 
Spanish 
German 
Czech