Abstract
BACKGROUND: Driving under the influence of marijuana is a serious traffic safety concern in the United States. Delta 9-tetrahydrocannabinol (THC) is the main active compound in marijuana. Although blood THC testing is a more accurate measure of THC-induced impairment, measuring THC in oral fluid is a less intrusive and less costly method of testing. METHODS: We examined whether the oral fluid THC test can be used as a valid alternative to the blood THC test using a sensitivity and specificity analysis and a logistic regression, and estimate the quantitative relationship between oral fluid THC concentration and blood THC concentration using a correlation analysis and a linear regression on the log-transformed THC concentrations. We used data from 4596 drivers who participated in the 2013 National Roadside Survey of Alcohol and Drug Use by Drivers and for whom THC testing results from both oral fluid and whole blood samples were available. RESULTS: Overall, 8.9% and 9.4% of the participants tested positive for THC in oral fluid and whole blood samples, respectively. Using blood test as the reference criterion, oral fluid test for THC positivity showed a sensitivity of 79.4% (95% CI: 75.2%, 83.1%) and a specificity of 98.3% (95% CI: 97.9%, 98.7%). The log-transformed oral fluid THC concentration accounted for about 29% of the variation in the log-transformed blood THC concentration. That is, there is still 71% of the variation in the log-transformed blood THC concentration unexplained by the log-transformed oral fluid THC concentration. Back-transforming to the original scale, we estimated that each 10% increase in the oral fluid THC concentration was associated with a 2.4% (95% CI: 2.1%, 2.8%) increase in the blood THC concentration. CONCLUSIONS: The oral fluid test is a highly valid method for detecting the presence of THC in the blood but cannot be used to accurately measure the blood THC concentration.