AAA Finds No Basis for Equating THC Blood Levels With Driver Impairment

Traffic safety trends in Washington after marijuana legalization are ambiguous.


AAA Foundation for Traffic Safety

According to a new report from the AAA Foundation for Traffic Safety, the percentage of drivers in fatal collisions who tested positive for THC doubled in Washington after that state legalized marijuana. But other reports published the same day by the same organization cast doubt on the significance of that finding, underlining the perils of equating THC in the blood with impairment.

In 2014, the first year that marijuana was legally sold for recreational use in Washington, 17 percent of drivers in fatal crashes tested positive for THC, up from 8.3 percent in 2013, when recreational marijuana was legal to possess but not to grow or sell. "The proportion of drivers positive for THC was generally flat before and immediately after Initiative 502 [Washington's legalization measure] took effect," the AAA report says, "but began increasing significantly…approximately 9 months after Initiative 502 took effect. It was not clear whether this increasing trend was attributable to Initiative 502 or to other factors that were beyond the scope of the study."

The report also notes that "results of this study do not indicate that drivers with detectable THC in their blood at the time of the crash were necessarily impaired by THC or that they were at fault for the crash," since "the data available cannot be used to assess whether a given driver was actually impaired, and examination of fault in individual crashes was beyond the scope of this study." The increase in drivers testing positive for THC may reflect an increase in marijuana use, but it does not necessarily indicate an increase in the number of dangerously stoned drivers on the road.

Another AAA study further muddies the picture. It found that the share of drivers involved in accidents (both fatal and nonfatal) or arrested on suspicion of driving under the influence (DUI) who tested positive for THC rose from 20 percent in 2005 to 30 percent in 2014. But this upward trend, which may be related to an increase in the number of Washington State Patrol troopers trained to recognize drug-impaired drivers, slowed after passage of I-502, contrary to what you would expect if legalization led to more stoned driving. In both this study and the one focusing on fatal accidents, drivers who tested positive for THC typically also tested positive for alcohol or other drugs. That was true for 66 percent of the THC-positive drivers in fatal crashes and 73 percent of the THC-positive drivers who were arrested or involved in any sort of collision.

The report notes that the median time between a DUI arrest or collision and obtaining a blood sample was more than two and a half hours. That lag, which in some cases might have been long enough for THC to fall below detectable levels, suggests that the prevalence of "THC-involved driving" may be underestimated. The researchers say "evaluating the impact of protracted time until blood testing is complicated by the lack of available standardized law enforcement data on the time of testing."

A third AAA study focuses on the distinction between "THC-involved driving" and THC-impaired driving, finding no clear relationship between THC blood levels in DUI arrestees and performance on roadside sobriety tests (the walk-and-turn test, one-leg-stand test, and finger-to-nose test). THC-positive drivers were much more likely to fail the tests than a group of drug-free controls (although even the latter group had substantial failure rates, ranging from 33 percent to 51 percent, which makes you wonder how accurate these tests are as measures of impairment). But the amount of THC in drivers' blood was not correlated with their test performance. "There was no correlation between blood THC concentration and scores on the individual indicators," the report says, "and performance on the indicators could not reliably assign a subject to the high or low blood THC categories." In short, "there is no evidence from the data collected…that any objective threshold exists that established impairment."

The implications for states contemplating a per se DUI standard similar to Washington's (which equates a THC blood concentration of five nanograms per milliliter with impairment) are clear: "Based on this analysis, a quantitative threshold for per se laws for THC following cannabis use cannot be scientifically supported." That conclusion is similar to the position taken by the National Highway Traffic Safety Administration (NHTSA). "Whereas the impairment effects for various concentration levels of alcohol in the blood or breath are well understood," NHTSA says, "there is little evidence available to link concentrations of other drugs to driver performance."

The lack of a scientific foundation for defining impairment based on THC blood levels has not stopped legislators from adopting or considering such standards as a way to create the impression that they are doing something about the threat posed by stoned driving (which is real but pales beside the threat posed by drunk driving). Such rules are both underinclusive, since some people may be dangerously impaired at THC levels below the cutoff, and overinclusive, since some regular cannabis consumers are perfectly capable of driving safely at THC levels above the cutoff.