Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis
Smoking cannabis increases your chance of a road traffic incident, although no threshold has been defined
- “Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis” by M Asbridge and colleagues (BMJ 2012;344:e536, doi:10.1136/bmj.e536)
- Objective—To determine whether the acute consumption of cannabis (cannabinoids) by drivers increases the risk of a motor vehicle collision.
- Design—Systematic review of observational studies, with meta-analysis.
- Data sources—We did electronic searches in 19 databases, unrestricted by year or language of publication. We also did manual searches of reference lists, conducted a search for unpublished studies, and reviewed the personal libraries of the research team.
- Review methods—We included observational epidemiology studies of motor vehicle collisions with an appropriate control group, and selected studies that measured recent cannabis use in drivers by toxicological analysis of whole blood or self report. We excluded experimental or simulator studies. Two independent reviewers assessed risk of bias in each selected study, with consensus, using the Newcastle-Ottawa scale. Risk estimates were combined using random effects models.
- Results—We selected nine studies in the review and meta-analysis. Driving under the influence of cannabis was associated with a significantly increased risk of motor vehicle collisions compared with unimpaired driving (odds ratio 1.92 (95% confidence interval 1.35 to 2.73); P=0.0003); we noted heterogeneity among the individual study effects (I=81). Collision risk estimates were higher in case-control studies (2.79 (1.23 to 6.33); P=0.01) and studies of fatal collisions (2.10 (1.31 to 3.36); P=0.002) than in culpability studies (1.65 (1.11 to 2.46); P=0.07) and studies of non-fatal collisions (1.74 (0.88 to 3.46); P=0.11).
- Conclusions—Acute cannabis consumption is associated with an increased risk of a motor vehicle crash, especially for fatal collisions. This information could be used as the basis for campaigns against drug impaired driving, developing regional or national policies to control acute drug use while driving, and raising public awareness.
Why do the study?
For several reasons it is important to know whether driving within several hours of smoking cannabis increases the risk of a motor vehicle collision. Firstly, many young people in Europe, North America, and Australia have used cannabis at some time, a substantial minority do so regularly, and a substantial proportion of regular users drive within several hours of doing so. Secondly, young people are at an increased risk of motor vehicle collision because of their relative inexperience and higher levels of risk taking. Thirdly, after alcohol, cannabis is the drug that has most often been found in the blood in post-mortem studies of drivers killed in incidents in these countries.
Two types of study have assessed whether there is a causal relationship between cannabis use and risk of an incident. In laboratory studies researchers randomly assign people to receive cannabis or not and then compare their performance in a driving simulator.
Occasionally, investigators have assessed cannabis’s effects on driving around courses, and in one Dutch study on the road. The strength of such experimental studies is that random assignment of participants to cannabis use or not controls for the effects of any pre-existing differences in behaviour that might affect driving performance. These studies generally suggest that cannabis adversely affects performance in ways likely to impair driving—for example, by slowing reaction time and impairing hand-eye coordination and judgment.  Their weakness is uncertainty about whether their results translate to risk of incident on the road.
In the second type of study, observational epidemiological studies, researchers observe relationships between cannabis use and incidents in drivers who have used cannabis or not (for example, as indicated by the presence or absence of cannabis metabolites in the blood of people injured or killed in crashes). The strength of these studies is that they directly assess the risk of an incident; their weakness is that we cannot be sure that cannabis use is the only difference between drivers who use the drug and drivers who do not. The relationship between cannabis and risk of an incident would be confounded, for example, if drivers who took more risks were more likely to use cannabis. Observational studies typically use statistical methods to control for the effects of pre-existing differences between cannabis users and non-users (for example, in propensity to take risks or use of alcohol and other drugs) when assessing the association between cannabis use and risk of an incident.
What did the authors do?
The authors undertook a systematic review and meta-analysis of the observational studies to assess whether there was a consistent association between recent cannabis use and risk of a motor vehicle collision. Meta-analysis is a statistical method for assessing the consistency of study findings and assessing the degree to which differing study results reflect the role of chance or differences in study size and statistical power to detect relationships. If the literature is inconsistent, meta-analysis allows us to examine what characteristics of studies are related to their findings—for example, study designs, sample size, the way that recent cannabis use was measured, and the measures used for risk of incidents such as deaths or injuries.
The authors identified potential studies by conducting a literature search using standard search terms in the major databases of published studies such as PubMed. They also searched the papers cited in the studies identified by this search, which produced a small number of additional studies. They included only studies that used biochemical markers of recent cannabis use, such as levels of tetrahydrocannabinol (THC), the active ingredient in cannabis, or its psychoactive metabolites in blood.
The studies included cross-sectional studies and case-control studies, including a variant of the case-control study design known as a culpability study. Cross-sectional studies involved analyses of THC in the blood of persons killed or injured in road traffic incidents. Case-control studies compared rates of recent cannabis use in people killed or injured in motor vehicle collisions with control drivers who had not been in incidents.
In culpability studies researchers make a judgment about whether a driver involved in an incident was culpable or not—that is, contributed to the incident by the way that he or she drove. Researchers who make this judgment do not know whether the driver had used any drugs. Researchers then compare rates of cannabis (and other drugs) use in drivers who are culpable and those who are not. If cannabis use increases the risk of incidents then rates of its use will be higher among culpable than non-culpable drivers.
Two reviewers independently assessed the quality of each study. The authors then assessed the consistency of the strength of association between cannabis use and risk of motor vehicle collision across different study designs and design quality (medium v high).
What did the study find?
The authors identified 2975 papers, reviewed the abstracts of 222, and did a full review of 23 papers. They identified 10 studies that met criteria but excluded one that duplicated data from a previous paper. Their formal analysis was done on nine studies (figure). 1
The authors found that pooled or overall risk of a motor vehicle collision while driving under the influence of cannabis in all nine studies (which included 49 411 participants) was nearly twice the risk of driving while not impaired (odds ratio 1.92; 95% confidence interval 1.35 to 2.73).
The size of the risk marginally varied between study designs. The high quality studies had a larger pooled OR than the medium quality ones, although both showed a significant association. Case-control studies (2.79; 1.23 to 6.33) found a higher risk than the culpability studies (1.65; 1.11 to 2.46). Studies of fatal collisions (2.10; 1.31 to 3.36) had a higher pooled odds ratio than studies of non-fatal collisions (1.74; 0.88 to 3.46), which did not show significant results. In all studies that assessed the effects of cannabis use in combination with levels of alcohol indicative of intoxication, the risk of using both drugs was higher than that for cannabis use alone.
Strengths and weaknesses
The study had several limitations. Firstly, only nine studies were analysed. Secondly, by their nature, observational studies do not completely exclude the possibility that the association between cannabis use and motor vehicle collision is caused by unmeasured pre-existing differences in risk of motor vehicle collision. For example, most studies in the meta-analysis have not been able to statistically adjust the history of incidents in drivers who do use cannabis and those who do not use cannabis.
Across studies, the main possible source of bias in exposure measurement was the concurrent use of alcohol or other drugs as well as cannabis. Although these data were available for alcohol, there was insufficient information available to assess the effects of other illicit drugs. This bias would overestimate the effect of cannabis on collision risk so the authors calculated odds ratios using “cannabis only” cases and controls—that is, only motorists whose system contained cannabis without any other drugs or alcohol present. They found an increased risk of incident in those using only cannabis compared with controls.
Thirdly, there were variations in findings between case-control studies and culpability studies, and in studies of fatalities and injuries, although these were as expected. For example, the lower risk in culpability studies reflects the fact that culpability is high in drivers who did not use alcohol or any other drugs. This limits the size of the increase in risk that can be detected in drivers who used cannabis or other drugs. Similarly, the weaker relationship found in studies of injuries other than fatalities was probably explained by the longer time, on average, between testing for THC in the blood of persons injured than in those killed in motor vehicle collisions. It is standard practice in many countries to take a blood sample from persons killed in incidents. Consent has to be obtained from people who have been injured and this could be delayed if they are being treated.
Fourthly, in most of these studies, researchers were unable to assess whether the risk of a motor vehicle collision increased with the amount of cannabis that a driver had used. This reflects technical challenges in quantifying THC and its metabolites, and the absence of a clear relationship between blood THC and impairments, as exists between blood alcohol level and driving impairment. Nonetheless, in the three studies that did assess the risk, a dose response was found.   And laboratory studies have shown a dose-related impairment in performance on tasks related to driving, including performance in a driving simulator. 
What does the study mean?
The authors found that in controlled observational studies of the effects of acute cannabis use on risk of motor vehicle collision, there was a near doubling of risk of a driver being in a motor vehicle collision resulting in serious injury or death. The increased risk was most evident for high quality studies, case-control studies, and studies of fatal collisions. The effect of acute cannabis use on minor motor vehicle collision risk was less clear.
If these studies comprised the only evidence for a relationship between cannabis use and risk of incident some people might want to reserve judgment until better studies had been done. It is difficult to do better studies, however, and the results are consistent with the results of more rigorous laboratory and simulator studies.
The evidence collectively shows the need for policies to reduce cannabis impaired driving, possibly including roadside drug testing, like those introduced to reduce drunk driving. The major challenge in roadside drug testing is in specifying a level of cannabis in blood that provides strong evidence of intoxication. That is, a level that would be similar to the one used in defining alcohol-impaired driving—for example, a blood alcohol concentration of less than 0.05% or 0.08%. Several Australian states, some US states, and several Scandinavian countries have nonetheless introduced roadside drug testing. These countries have defined as impairment any detectable level of cannabis that indicates recent use, rather than attempting to define a level that is associated with impairment. It remains to be seen what impact this type of roadside drug testing has had on motor vehicle collisions.
- EDITORIAL, p 10
1University of Queensland Centre for Clinical Research, University of Queensland, Royal Brisbane and Women’s Hospital, Herston, QLD 4029, Australia
Correspondence to: firstname.lastname@example.org
Competing interests: WH has received support from an Australia Fellowship with the National Health and Medical Research Council.
Provenance and peer review: Commissioned; not externally peer reviewed.
- Hall W, Degenhardt, L. The adverse health effects of nonmedical cannabis use. Lancet 2009;374:1383-91.
- Russell KF, Vandermeer B, Hartling L. Graduated driver licensing for reducing motor vehicle crashes among young drivers. Cochrane Database Syst Rev 2011;(10):CD003300.
- Kelly E, Darke S, Ross J. A review of drug use and driving: epidemiology, impairment, risk factors and risk perceptions. Drug Alcohol Rev 2004;23:319-44.
- Robbe HWJ. Influence of marijuana on driving. Institute for Human Psychopharmacology, University of Limberg, 1994.
- Grotenhermen F, Leson G, Berghaus G, Drummer OH, Kruger HP, Longo M, et al. Developing limits for driving under cannabis. Addiction 2007;102:1910-7.
- Ramaekers JG, Berghaus G, van Laar M, Drummer OH. Dose related risk of motor vehicle crashes after cannabis use. Drug Alcohol Depend 2004;73:109-19.
- Longo MC, Hunter CE, Lokan RJ, White JM, White MA. Role of alcohol, cannabinoids, benzodiazepines and stimulants in road crashes. Alcohol, Drugs and Traffic Safety—T2000: Proceedings of the 15th International Conference on Alcohol, Drugs and Traffic Safety; May 22-6 2000. International Council for Alcohol, Drugs & Traffic Safety, 2000:363-74.
- Laumon B, Gadegbeku B, Martin J-L, Biecheler M-B, SAM Group. Cannabis intoxication and fatal road crashes in France: population based case-control study. BMJ 2005;331:1371-7.
- Drummer OH, Gerostamoulos J, Batziris H, Chu M, Caplehorn J, Robertson MD, et al. The involvement of drugs in drivers of motor vehicles killed in Australian road traffic crashes. Accid Anal Prev 2004;36:239-48.
Cite this as: Student BMJ 2012;20:e1735