The Impact of Cannabis on Driving

Abstract

Cannabis is known to have detrimental effects on human performance and may also affect driving adversely. However, studies designed to examine this issue have provided equivocal findings. We set up this study to further determine the effect of cannabis on driving.
We used a cross-sectional, case-control design with drivers aged 20-49 who were involved in a fatal crash in the United States from 1993 to 2003; drivers were included if they had been tested for the presence of cannabis and had a confirmed blood alcohol concentration of zero. Cases were drivers who had at least one potentially unsafe driving action recorded in relation to the crash (e.g., speeding); controls were drivers who had no such driving action recorded. We calculated the crude and adjusted odds ratios (ORs) of any potentially unsafe driving action in drivers who tested positive for cannabis but negative for alcohol consumption. In computing for the adjusted OR, we controlled for age, sex, and prior driving record.
Five percent of drivers tested positive for cannabis. The crude OR of a potentially unsafe action was 1.39 (99% CI = 1.21-1.59) for drivers who tested positive for cannabis. Even after controlling for age, sex, and prior driving record, the presence of cannabis remained associated with a higher risk of a potentially unsafe driving action (1.29, 99% CI = 1.11-1.50).
Cannabis had a negative effect on driving, as would be predicted from human performance studies. This finding supports the need for interventions to decrease the prevalence of driving under the influence of cannabis, and indicates that further studies should be conducted to investigate the dose-response relationship between cannabis and safe driving.

Full text

6 REVUE CANADIENNE DE SANTÉ PUBLIQUE VOLUME 98, NO. 1
The Impact of Cannabis on Driving
Michel Bédard,
PhD
1-3
Sacha Dubois, HBA
3
Bruce Weaver, MSc
1,2
ABSTRACT
Background: Cannabis is known to have detrimental effects on human performance and
may also affect driving adversely. However, studies designed to examine this issue have
provided equivocal findings. We set up this study to further determine the effect of
cannabis on driving.
Methods: We used a cross-sectional, case-control design with drivers aged 20-49 who
were involved in a fatal crash in the United States from 1993 to 2003; drivers were
included if they had been tested for the presence of cannabis and had a confirmed blood
alcohol concentration of zero. Cases were drivers who had at least one potentially unsafe
driving action recorded in relation to the crash (e.g., speeding); controls were drivers who
had no such driving action recorded. We calculated the crude and adjusted odds ratios
(ORs) of any potentially unsafe driving action in drivers who tested positive for cannabis
but negative for alcohol consumption. In computing for the adjusted OR, we controlled for
age, sex, and prior driving record.
Results: Five percent of drivers tested positive for cannabis. The crude OR of a potentially
unsafe action was 1.39 (99% CI = 1.21-1.59) for drivers who tested positive for cannabis.
Even after controlling for age, sex, and prior driving record, the presence of cannabis
remained associated with a higher risk of a potentially unsafe driving action (1.29, 99% CI
= 1.11-1.50).
Conclusion: Cannabis had a negative effect on driving, as would be predicted from human
performance studies. This finding supports the need for interventions to decrease the
prevalence of driving under the influence of cannabis, and indicates that further studies
should be conducted to investigate the dose-response relationship between cannabis and
safe driving.
MeSH terms: Cannabis; accidents, traffic; alcohol drinking; automobiles
A
survey revealed that 7.3% of
Canadians used cannabis (marijua-
na) at least once in the previous
year and that 2.0% used it weekly.
1
Similarly, American (US) surveys revealed
that the prevalence of cannabis use in the
previous year is approximately 4%.
2
However, cannabis use is more prevalent
among young people. Among Canadian
undergraduate students, 18.2% reported
using cannabis during the academic year.
3
In the US, the prevalence of past-year use
among adults aged 18-29 in 2000-01 was
10.5%. In a survey of US college students,
15.7% reported having used cannabis in
the past 30 days.
4
While these trends fuel
many debates about cannabis use, one per-
sistent unresolved issue is the potential
effect of cannabis on driving.
5
The effect of cannabis on the human
system is wide-ranging, combining
“…many of the properties of alcohol, tran-
quilizers, opiates and hallucinogens; it is
anxiolytic, sedative, analgesic, psychedelic;
it stimulates appetite and has many sys-
temic effects.”
6
(p. 103-104) Nonetheless,
while the negative effect of cannabis on
cognitive functions supporting safe driving
was documented by some researchers,
7-9
others
10
have reported that performance is
affected minimally, and often only in con-
junction with alcohol. Epidemiological
studies also provide equivocal findings.
Some have demonstrated a small but statis-
tically significant association between
cannabis use and driving problems.
11-13
However, many researchers have not
found an association between cannabis use
and crashes (see refs. 14 and 15 for
reviews).
It is possible that cannabis has minimal
effects because drivers compensate for their
impairment.
5
However, it is also possible
that methodological issues explain discrep-
ancies in these findings. The concentration
of THC (the active ingredient in cannabis)
in drivers is typically not available nor is
the amount of time since it was absorbed.
Recently absorbed cannabis may affect dri-
ving adversely but this effect may wane
once a longer period of time has elapsed
after absorption.
16
Because THC has a
half-life of approximately seven days,
6
many drivers who test positive for cannabis
may not be impaired at the time of testing.
This would lead to an underestimation of
the association between cannabis and poor
driving/crashes and possibly the absence of
La traduction du résumé se trouve à la fin de l’article.
1. Public Health Program, Lakehead University, Thunder Bay, ON
2. Division of Human Sciences, Northern Ontario School of Medicine, Thunder Bay
3. St. Joseph’s Care Group, Thunder Bay
Correspondence and reprint requests: Michel Bédard, Public Health Program, Lakehead University,
955 Oliver Road, Thunder Bay, ON P7B 5E1, Fax: 807-346-7734, E-mail: michel.bedard@lake-
headu.ca
Acknowledgements: The authors thank Julie Riendeau for her assistance with the literature search and
Owen Marks for his assistance with data management. Michel Bédard is a Canada Research Chair in
Aging and Health (www.chairs.gc.ca); he acknowledges the support of the Canada Research Chair
Program. Additional funding for this research was provided through a research grant from AUTO21,
Network of Centres of Excellence. The sponsors were not involved in any aspects of this study.

JANUARY – FEBRUARY 2007 CANADIAN JOURNAL OF PUBLIC HEALTH 7
THE IMPACT OF CANNABIS ON DRIVING
statistical significance. Furthermore,
because the prevalence of cannabis use is
relatively low in the general population,
small sample sizes yield limited numbers of
users involved in crashes.
17
One remedy to
this problem is to use large datasets.
Although these would not eliminate the
underestimation of the association, a
greater number of cases would increase the
statistical power of the analyses.
Another problem is that cannabis use is
often highly correlated with other crash
determinants (e.g., alcohol use, high-risk
behaviour). Alcohol use is especially prob-
lematic given that it is correlated with
crashes, and that it may exacerbate the
effects of cannabis.
16
Other high-risk
behaviours such as speeding, may also con-
found the association between cannabis
and crashes. Drummer and colleagues
18
reported an association between THC and
responsibility for crashes (OR = 2.7, 95%
CI = 1.0-7.0) but could not take into con-
sideration high-risk behaviours (with the
exception of alcohol use). Hence, the pos-
sibility remains that cannabis users are sim-
ply high-risk drivers who are involved in a
disproportionate number of crashes; this
requires that we separate the respective
effects of cannabis and alcohol, and control
for high-risk driving habits.
Another methodological problem is the
reliance on self-reports to identify cannabis
use and crashes.
13
Given the social and
legal implications of cannabis use, self-
reports may also result in the underestima-
tion of the association between cannabis
and crashes. Ideally, the presence of THC
in the body and crash status should be
determined through objective means (e.g.,
blood tests, police reports).
Our goal was to clarify the association
between cannabis use and driver behav-
iour. Specifically, we aimed to determine if
cannabis, in the absence of alcohol, is relat-
ed to driving actions that may have result-
ed in fatal crashes. We used the Fatality
Analysis Reporting System (FARS), an
administrative database where information
on all crashes involving a fatality in the US
is recorded. This database has several
advantages. First, all crashes involving at
least one fatality are included, therefore
reducing selection bias. Second, the infor-
mation is obtained by investigators, there-
fore eliminating biases that may arise from
self-reports. Third, it contains actual blood
alcohol concentration (BAC). Fourth, it
contains information on prior driving inci-
dents (e.g., crashes, violations) to control
for high-risk driving behaviours. Finally, it
is sufficiently large to allow us to focus
only on drivers who were tested for drugs
and alcohol, and to examine the drivers’
actions that may have resulted in the crash.
Based on previous research and the
acknowledged impact of cannabis on cog-
nitive functions, we hypothesized that
cannabis use would be related to poorer
driver behaviour.
METHODS
Data source
The FARS database is one of the most
comprehensive databases on crashes.
19
For
every traffic fatality in the US since 1975,
information is compiled about crash situa-
tions, drivers and passengers, and about
the vehicles involved. The quantity of
information coded in the database, and
number of crashes recorded, allows for the
control of numerous potential con-
founders, and the calculation of crash esti-
mates more easily generalized to all drivers
involved in fatal crashes.
20
We used data from 1993 to 2003 (inclu-
sive) because drug information has been
collected more comprehensively since
1993. The database also contains informa-
tion to identify risk factors pertinent to
crash initiation and not only crash involve-
ment. This information is contained in the
“driver-related factors” (DRF). Briefly, for
every driver, up to three (four since 1997)
driver-related factors were coded according
to police reports. Driver-related codes from
20 to 59 (inclusive) were considered
actions that may have contributed to the
crash (e.g., driving too fast for the condi-
tions; these codes are presented in the
Appendix). Drivers for whom no DRFs
were specified were assumed to not have
contributed to crash initiation. This
approach is not a perfect substitute for the
assessment of “responsibility” but has been
used successfully by other researchers.
21-23
Our aim was to identify predictors of any
DRF.
We used other data from FARS, includ-
ing: age, sex, drug test results (blood or
urine), alcohol tests results (blood), drivers’
past driving record, and the type of vehicle
driven. For each driver, a maximum of
three drug results were provided (in no
particular order). The cannabinoid drug
compounds (FARS 600 series) were
included: Delta 9 (600), Hashish Oil
(601), Hashish (602), Marijuana (603),
Marinol (604), Tetrahydrocannabinoid
(605), THC (606), Cannabinoid, Type
Unknown (695); the concentration of the
drug is not available. For alcohol, the actu-
al BAC is available. This allowed us to
identify drivers who were tested for alcohol
use, to verify the validity of our approach
with alcohol data, and to examine the con-
tribution of cannabis in drivers free of
alcohol. We elected to use only drivers for
whom both alcohol and drug tests were
performed and reported because one
TABLE I
Descriptive Statistics for Drivers Involved in Fatal Automobile Collision
Demographics (N=314,636)
Age, mean (SD) 32.48 (8.59)
Male, No. (%) 222,671 (71)
Driving Record – Any in the Past Three Years (N=314,636) No. (%)
Crashes 50,430 (16)
DWI 15,873 (05)
Other conviction 59,983 (19)
Speeding 79,000 (26)
License suspension/revocation 52,171 (17)
Any of the above 153,240 (49)
TABLE II
Odds Ratios and Confidence Intervals of a DRF by BAC
Model 1 – Unadjusted Model 2 – Adjusted
BAC Level Odds Ratio (95% CI; 99% CI) Odds Ratio (95% CI; 99% CI)
0 Reference category Reference category
0.05 2.20 (2.08-2.32; 2.05-2.36) 2.01 (1.90-2.13; 1.87-2.16)
0.10 3.37 (3.21-3.54; 3.17-3.59) 3.06 (2.91-3.22; 2.87-3.27)
0.15 4.73 (4.52-4.94; 4.46-5.01) 4.40 (4.20-4.61; 4.14-4.68)
0.20 5.74 (5.50-5.98; 5.43-6.06) 5.61 (5.36-5.86; 5.29-5.95)
0.30 6.00 (5.69-6.33; 5.59-6.44) 6.16 (5.82-6.52; 5.71-6.64)

THE IMPACT OF CANNABIS ON DRIVING
8 REVUE CANADIENNE DE SANTÉ PUBLIQUE VOLUME 98, NO. 1
potential bias is that only drivers suspected
of impairment may be tested.
We also used data on drivers’ past three
years’ driving record, including: number of
accidents (crashes), number of recorded
convictions for driving while impaired
(DWI; includes both alcohol and drug),
speeding convictions (going too fast or too
slow), other harmful moving violation con-
victions, and licence suspensions and revo-
cations. Because controlling for driving
habits through these variables is important
to rule out the confounding effect of high-
risk driving habits, we excluded drivers
aged less than 20 because they may not
have had sufficient opportunity (years) to
incur such records. We excluded drivers
aged 50 and over because cannabis use
becomes much less frequent with advanc-
ing age. Finally, we limited our analyses to
drivers of passenger vehicles, sport-utility
vehicles and light trucks (pick-up trucks)
only. The focus on these drivers will facili-
tate the interpretation of the findings.
Analytical plan
We first sought to confirm known findings
regarding alcohol use and crashes by
demonstrating a dose-response relationship
between BAC and DRF. To this effect, we
used two logistic regressions using the
report of any DRF as the dependent vari-
able. In the first model, the only predictor
variable was BAC coded as the following
categories: 0 (0-0.02; the reference catego-
ry), 0.05 (0.03-0.07), 0.10 (0.08-0.12),
0.15 (0.13-0.17), 0.20 (0.18-0.24), 0.30
or more (0.25-0.94). In the second model,
we added the following list of potential
confounding variables: age, sex, and the
past three years’ driving record.
Our second set of analyses focused solely
on cannabis. To achieve this goal, we
looked exclusively at drivers who had a
confirmed BAC of zero. We present demo-
graphic characteristics and the 10 most fre-
quent DRFs for these drivers, and used
independent t-tests and Pearson’s Chi-
square tests to compare drivers who tested
positive for cannabis and those who did
not. We then present regression models
(crude and adjusted) to determine the con-
tribution of cannabis to DRFs. We report
95% and 99% confidence intervals (CI).
RESULTS
Alcohol data
Briefly, approximately two thirds of the
sample were male, the mean age was 32.5
(SD = 8.6), and one of every two drivers in
the sample had a driving record (see Table
I). Table II and Figure 1 present the OR
and CI of any DRF by BAC category. In
Model 1, we present the crude association
between DRFs and BAC. In Model 2, we
present the OR adjusted for potential con-
founders (age, sex, prior driving record).
We found a clear relationship between
increasing BAC and the OR of any DRF.
These results replicate the known dose-
response relationship between BAC and
crash risk.
Cannabis data
Of the 32,543 drivers tested, 1,647 (5%)
tested positive for cannabis. Drivers who
tested positive were generally younger,
male, and had a poorer driving record in
the past three years (see Table III; all dri-
vers had a confirmed BAC of zero). The 10
most frequently reported DRFs are pre-
sented in Table IV. A greater proportion of
drivers who tested positive for cannabis
had a DRF related to speeding or
erratic/reckless driving.
The crude OR between DRFs and
cannabis was 1.39 (99% CI = 1.21-1.59).
After adjusting this association for age, sex,
and driving record (see Table V) the OR
for THC was 1.29 (99% CI = 1.11-1.50).
Younger age and poorer driving records,
but not sex, were also associated with a
higher risk of a reported DRF.
DISCUSSION
The findings point to cannabis as a poten-
tial risk factor in fatal crashes. Individuals
Figure 1. OR and 99% CI for any DRF by BAC (categorical) level
Panel A: Unadjusted Odds Ratios Panel B: Adjusted Odds Ratios
Alcohol Level (BAC = 0 to 0.02 is the reference) Alcohol Level (BAC = 0 to 0.02 is the reference)

who tested positive for cannabis but nega-
tive for alcohol had 29% excess risk (99%
CI = 11-50) of having driven in a fashion
that may have contributed to the crash,
compared to drivers who tested negative
for cannabis. This association was found
after controlling for age, sex, and prior dri-
ving record. However, our findings likely
reflect an underestimation of the actual
effect of cannabis on driving. Given the
long half-life of THC, it is possible that
many drivers tested positive for cannabis
without being impaired at the time of the
crash. Furthermore, it is difficult to make
the distinction between the presence of the
active component of cannabis (THC), and
its metabolites, which have no effects on
the brain.
24
We found that 5% of the drivers tested
positive for cannabis, but this number is
not based on a representative sample. The
authors of a study based on a representa-
tive sample of adults living in Ontario
reported that only 1.9% of adults drove
under the influence of cannabis in the pre-
vious year.
25
However, this number was
19.7% in a survey of Ontario high school
students who had a driver’s licence.
26
In a
study of impairment in reckless drivers,
more than half of those tested but not
impaired by alcohol were impaired by
some drug (cannabis was the most fre-
quent).
27
These numbers may be related to
young people’s perception of the risks asso-
ciated with cannabis use. Contrary to what
many of them would report regarding alco-
hol, several consider cannabis to have a
negligible effect on driving, and some even
believe that cannabis may enhance
driving,
15,28,29
which suggests that we need
to deal with these perceptions and atti-
tudes.
30
We also replicated findings regarding
alcohol.
21
While this is not new, it increas-
es our confidence in our results regarding
cannabis and also allows us to put these
findings in perspective. The frequency of
drinking and driving and the severe impact
of alcohol on driving abilities are well
beyond what has been shown with
cannabis, even if we consider that we may
be underestimating the association. We
also found, once more, that young males,
and especially those with a bad driving
record, were at greater risk of driving in an
unsafe fashion, and controlled for these
variables. This was important given that
others have reported statistical associations
between cannabis use and traffic
violations,
31
and in one study the excess
risk posed by cannabis was eliminated once
drinking and driving behaviour and sex
were considered.
13
Hence, we are confident
that the higher risk found in drivers who
tested positive for cannabis cannot be
explained by aggressive driving patterns or
alcohol consumption.
One unanswered question is whether we
could identify dosages at which cannabis
may start to pose a crash risk, as has been
done with alcohol.
32
This question points
to a lack of knowledge regarding the dose-
response relationship between cannabis
and driving, knowledge which is important
to better educate drivers and support policy-
making. For example, new procedures for
roadside screening of cannabis use (using
saliva) are highly predictive of actual use.
33
However, the successful implementation of
such roadside testing may require the
determination of the minimum blood con-
centration at which one does become
“impaired” to safely drive an automobile.
The issue of safe driving and cannabis
use is not restricted to recreational users;
cannabis is increasingly used for medicinal
purposes,
34
and long-term use of cannabis
may create residual cognitive impair-
ment.
35
Moreover, many conditions for
which cannabis may be useful (e.g., glauco-
ma, cancer) are seen in older people; hence
the effect of cannabis may be exacerbated
by age-related changes in pharmacokinetics
and/or the presence of other medications.
Further studies are required to better
understand the impact of cannabis on dri-
vers who are using it as a medication.
We see our findings as additional evi-
dence suggesting that cannabis may
adversely affect drivers. The results are
consistent with our knowledge of cannabis’
effect on human performance and all
mechanisms known to support safe dri-
ving.
24
Furthermore, the sample size
allowed us to obtain narrow confidence
JANUARY – FEBRUARY 2007 CANADIAN JOURNAL OF PUBLIC HEALTH 9
THE IMPACT OF CANNABIS ON DRIVING
TABLE III
Descriptive Statistics for Drivers Tested for THC and a BAC of Zero
Demographics Cannabis Cannabis
χχ
2
/t* p-value
Absent Present
(N=30,896) (N=1,647)
Age, mean (SD) 32.94 (8.88) 30.2 (8.44) 12.23 <0.001
Male, No. (%) 19,791 (64) 1296 (79) 146.72 <0.001
Driving Record – Any in the
past three years No. (%) No. (%)
Crashes 5174 (17) 295 (18) 1.52 0.22
DWI 767 (2) 74 (4) 25.12 <0.001
Other conviction 5831 (19) 502 (30) 134.40 <0.001
Speeding 7801 (26) 535 (33) 43.76 <0.001
License suspension/revocation 4229 (14) 419 (26) 177.86 <0.001
Any of the above 14,690 (48) 1013 (62) 122.02 <0.001
* Chi square values are presented for all variables with the exception of Age where the t-statistic is
presented.
TABLE IV
The Top 10 Reported DRF from 1993 to 2003
Driver-related Factor Cannabis Cannabis
χχ
2
p-value
Absent Present
(N=30,896) (N=1,647)
No. (%) No. (%)
Failure to keep in proper lane 9074 (29.4) 494 (30.0) 0.29 0.59
Driving too fast for conditions or
in excess of posted maximum 6072 (19.7) 428 (26.0) 39.24 0.001
Failure to yield right of way, obey
signs or other safety zone traffic
laws 3613 (11.7) 189 (11.5) 0.07 0.79
Erratic, reckless, careless or
negligent vehicle operation 1715 (5.6) 146 (8.9) 31.85 <0.001
Making improper turn 1718 (5.6) 89 (5.4) 0.07 0.79
Overcorrecting 1270 (4.1) 57 (3.5) 1.69 0.19
Driving on wrong side of road
(intentional or unintentional) 736 (2.4) 28 (1.7) 3.17 0.08
Passing with insufficient distance,
or visibility, or failing to yield
to overtaking vehicle 400 (1.3) 25 (1.5) 0.61 0.44
Improper or erratic lane changing 381 (1.2) 21 (1.3) 0.02 0.88
Following improperly 291 (0.9) 19 (1.2) 0.74 0.39
Any 18,405 (59.6) 1106 (67.2) 37.44 <0.001

intervals around our point estimates and
allowed for the control of important con-
founders; others have reported similar risk
estimates but could not rule out the play of
chance because of small sample sizes.
36,37
Yet, these estimates appear small compared
to alcohol and some prescription medica-
tions (e.g., long-acting benzodiazepines).
Nonetheless, we remain cautious because
we lacked knowledge about drug concen-
trations and we could not fully ascertain
responsibility for crashes. Possibly the best
approach to resolve the issue will be to
determine the dose-response relationship
between THC and driving performance.
REFERENCES
1. Ogborne AC, Smart RG. Cannabis users in the
general Canadian population. Subst Use Misuse
2000;35:301-11.
2. Compton W, Grant B, Colliver J, Glantz M,
Stinson F. Prevalence of marijuana use disorders
in the United States. JAMA 2004;291:2114-21.
3. Adlaf EM, Gliksman L, Demers A, Newton-
Taylor B. Illicit drug use among Canadian uni-
versity undergraduates. Can J Nurs Res
2003;35:24-43.
4. Gledhill-Hoyt J, Lee H, Strote J, Wechsler H.
Increased use of marijuana and other illicit drugs
at US colleges in the 1990s: Results of three
national surveys. Addiction 2000;95:1655-67.
5. Strang J, Witton J, Hall W. Improving the quali-
ty of the cannabis debate: Defining the different
domains. BMJ 2000;320:108-10.
6. Ashton H. Pharmacology and effects of cannabis:
A brief review. Br J Psychiatry 2001;178:101-6.
7. Kurzthaler I, Hummer M, Miller C, Sperner-
Unterweger B, Gunther V, Wechdorn H, et al.
Effect of cannabis use on cognitive functions and
driving ability. J Clin Psychiatry 1999;60:395-99.
8. Rizzo M, Lamers CTJ, Sauer CG, Ramaekers JG,
Bechara A, Andersen GJ. Impaired perception of
self-motion (heading) in abstinent ecstacy and
marijuana users. Psychopharmacology
2005;179:559-66.
9. Ramaekers JG, Robbe H, O’Hanlon JF.
Marijuana, alcohol and actual driving perfor-
mance. Human Psychopharmacology Clinical and
Experimental 2000;15:551-58.
10. Lamers CTJ, Ramaekers JG. Visual search and
urban city driving under the influence of mari-
juana and alcohol. Human Psychopharmacology
Clinical and Experimental 2001;16:393-401.
11. Macdonald S, Mann RE, Chipman M, Anglin-
Bodrug K. Collisions and traffic violations of
alcohol, cannabis and cocaine abuse clients before
and after treatment. Accid Anal Prev
2004;36:795-800.
12. Chipman ML, Macdonald S, Mann RE. Being
“at fault” in traffic crashes: Does alcohol,
cannabis, cocaine, or polydrug abuse make a dif-
ference? Inj Prev 2003;9:343-48.
13. Fergusson DM, Horwood LJ. Cannabis use and
traffic accidents in a birth cohort of young adults.
Accid Anal Prev 2001;33:703-11.
14. Macdonald S, Anglin-Bodrug K, Mann RE,
Erickson P, Hathaway A, Chipman M, et al.
Injury risk associated with cannabis and cocaine
use. Drug Alcohol Depend 2003;72:99-115.
15. Kelly E, Darke S, Ross J. A review of drug use
and driving: Epidemiology, impairment, risk fac-
tors and risk perceptions. Drug Alcohol Rev
2004;23:319-44.
16. Ramaekers JG, Berghaus G, van Laar M,
Drummer OH. Dose-related risk of motor vehi-
cle crashes after cannabis use. Drug Alcohol
Depend 2004;73:109-19.
17. Walsh JM, DeGier JJ, Christopherson AS,
Verstraete AG. Drugs and driving. Traffic Inj
Prev 2004;5:241-53.
18. Drummer O, Gerostamoulos J, Batziris H, Chu
M, Caplehorn J, Robertson M, et al. The
involvement of drugs in drivers of motor vehicles
killed in Australian road traffic crashes. Accid
Anal Prev 2004;36:239-48.
19. Fife D. Matching fatal accident reporting system
cases with National Center for Health Statistics
motor vehicle deaths. Accid Anal Prev
1989;21:79-83.
THE IMPACT OF CANNABIS ON DRIVING
10 REVUE CANADIENNE DE SANTÉ PUBLIQUE VOLUME 98, NO. 1
TABLE V
Odds Ratios with 95% and 99% CI for Models Predicting DRF
Main Variables Odds Ratio (95% CI; 99% CI)
THC – Adjusted Model 1.29 (1.15, 1.45; 1.11, 1.50)
Confounding Variables
Age 0.98 (0.98, 0.98; 0.98, 0.98)
Male 1.01 (0.96, 1.06; 0.94, 1.07)
Previous Accident
None (reference)
One in past three years 1.11 (1.04, 1.19; 1.02, 1.22)
Two or more in past three years 1.32 (1.16, 1.51; 1.11, 1.57)
Previous DWI
None (reference)
One in past three years 1.02 (0.86, 1.22; 0.81, 1.29)
Two or more in past three years 1.03 (0.67, 1.58; 0.59, 1.81)
Previous Other Moving Violation
None (reference)
One in past three years 1.12 (1.04, 1.21; 1.02, 1.23)
Two or more in past three years 1.25 (1.12, 1.40; 1.08, 1.45)
Previous Speeding Conviction
None (reference)
One in past three years 1.11 (1.04, 1.18; 1.02, 1.20)
Two or more in past three years 1.11 (1.01, 1.21; 0.99, 1.25)
Previous Licence Suspension/Revocation
None (reference)
One in past three years 1.36 (1.23, 1.50; 1.20, 1.54)
Two or more in past three years 1.67 (1.49, 1.87; 1.44, 1.93)
Appendix
Driver-related Factors
20 Leaving Vehicle Unattended in Roadway
21 Overloading or Improper Loading of Vehicle with Passengers or Cargo
22 Towing or Pushing Vehicle Improperly
23 Failing to [Dim Lights or, Since 1995] Have Lights on When Required
24 Operating without Required Equipment
25 Creating Unlawful Noise or Using Equipment Prohibited by Law
26 Following Improperly
27 Improper or Erratic Lane Changing
28 Failure to Keep in Proper Lane or Running off Road
29 Illegal Driving on Road Shoulder, in Ditch, on Sidewalk, on Median
30 Making Improper Entry to or Exit from Trafficway
33 Passing where Prohibited by Posted Signs, Pavement Markings, Hill or Curve, or School Bus
Displaying Warning not to Pass
34 Passing on Wrong Side
35 Passing with Insufficient Distance or Inadequate Visibility or Failing to Yield to Overtaking
Vehicle
36 Operating the Vehicle in Other Erratic, Reckless, Careless or Negligent Manner [or
Operating at Erratic or Suddenly Changing Speeds, Since 1995]
37 Traveling on Prohibited Trafficway (Since 1995)
38 Failure to Yield Right of Way
39 Failure to Obey Traffic Signs, Traffic Control Devices or Traffic Officers, Failure to Observe
Safety Zone Traffic Laws
40 Passing Through or Around Barrier Positioned to Prohibit or Channel Traffic
41 Failure to Observe Warnings or Instructions on Vehicles Displaying Them
42 Failure to Signal Intentions
43 Giving Wrong Signal
44 Driving Too Fast for Conditions or in Excess of Posted Speed Limit
45 Driving Less than Posted Maximum
46 Operating at Erratic or Suddenly Changing Speeds
47 Making Right Turn from Left Turn Lane or Making Left Turn from Right Turn Lane
48 Making Improper Turn
49 Driving Wrong Way on One-Way Trafficway
50 Driving on Wrong Side of Road [(Intentionally or Unintentionally) Since 1995]
51 Operator Inexperience
52 Unfamiliar with Roadway
53 Stopping in Roadway (Vehicle not Abandoned)
54 Underriding a Parked Truck
55 Getting Off/Out of or On/Into Moving Transport Vehicle
56 Getting Off/Out of or On/Into Non-Moving Transport Vehicle
57 Improper Tire Pressure (Since 1995)
58 Locked Wheel (Since 1995)
59 Overcorrecting (Since 1995)

JANUARY – FEBRUARY 2007 CANADIAN JOURNAL OF PUBLIC HEALTH 11
THE IMPACT OF CANNABIS ON DRIVING
20. Dischinger PC, Ho SM, Kufera JA. Medical con-
ditions and car crashes. 44th Annual Proceedings:
Association for the Advancement of Automotive
Medicine 2000;44:335-46.
21. Perneger T, Smith GS. The driver’s role in fatal
two-car crashes: A paired “case control” study.
Am J Epidemiol 1991;134:1138-45.
22. Bédard M, Myers J. The influence of passengers
on older drivers involved in fatal crashes. Exp
Aging Res 2004;30:205-15.
23. Pruesser DF, Williams AF, Ferguson SA, Ulmer
RG, Weinstein HB. Fatal crash risk for older dri-
vers at intersections. Accid Anal Prev
1998;30:151-59.
24. O’Kane CJ, Tutt DC, Bauer LA. Cannabis and
driving: A new perspective. Emerg Med
2002;14:296-303.
25. Walsh G, Mann RE. On the high road: Driving
under the influence of cannabis in Ontario. Can J
Public Health 1999;90(4):260-63.
26. Adlaf EM, Mann RE, Paglia A. Drinking,
cannabis use and driving among Ontario stu-
dents. CMAJ 2003;168:564-66.
27. Brookoff D, Cook CS, Williams C, Mann CS.
Testing reckless drivers for cocaine and marijua-
na. N Engl J Med 1994;331:518-22.
28. Albery IP, Strang J, Gossop M, Griffiths P. Illicit
drugs and driving: Prevalence, beliefs and acci-
dent involvement among a cohort of current out-
of-treatment drug users. Drug Alcohol Depend
2000;58:197-204.
29. Terry P, Wright KA. Self-reported driving behav-
iour and attitudes towards driving under the
influence of cannabis among three different user
groups in England. Addict Behav 2005;30:619-
26.
30. Davey JD, Davey T, Obst P. Drug and drink dri-
ving by university students: An exploration of the
influence of attitudes. Traffic Inj Prev 2005;6:44-
52.
31. Begg DJ, Langley JD, Stephenson S. Identifying
factors that predict persistent driving after drink-
ing, unsafe driving after drinking, and driving
after using cannabis among young adults. Accid
Anal Prev 2003;35:669-75.
32. Consensus Development Panel. Drug concentra-
tions and driving impairment. JAMA
1985;254:2618-21.
33. Toennes SW, Steinmeyer S, Maurer HJ, Moeller
MR, Kauert GF. Screening for drugs of abuse in
oral fluid—correlation of analysis results with
serum in forensic cases. J Anal Toxicol
2005;29:22-27.
34. Williamson EM, Evans FJ. Cannabinoids in clin-
ical practice. Drugs 2000;60:1303-14.
35. Solowij N, Stephens RS, Roffman RA, Babor T,
Kadden R, Miller M, et al. Cognitive functioning
of long-term heavy cannabis users seeking treat-
ment. JAMA 2002;287:1123-31.
36. Movig KLL, Mathijssen MPM, Nagel PHA, van
Egmond T, de Gier JJ, Leufkens HGM, et al.
Psychoactive substance use and the risk of motor
vehicle accidents. Accid Anal Prev 2004;36:636.
37. Lowenstein SR, Koziol-McLain J. Drugs and traf-
fic crash responsibility: A study of injured
motorists in Colorado. J Trauma 2001;50:313-20.
Received: November 25, 2005
Accepted: May 25, 2006
RÉSUMÉ
Contexte : On sait que le cannabis a des effets nuisibles sur les performances humaines et qu'il
pourrait aussi nuire à la conduite d'un véhicule. Cependant, les études sur le sujet donnent des
résultats peu probants. Dans notre étude, nous avons cherché à approfondir la question des effets
du cannabis sur la conduite.
Méthode : Nous avons mené une étude cas-témoin transversale auprès de conducteurs de 20 à 49
ans impliqués dans des accidents mortels aux États-Unis entre 1993 et 2003; nous n'avons inclus
que les conducteurs ayant fait l'objet d'un test de dépistage du cannabis, mais dont le taux
d'alcoolémie était nul. Les cas étaient des conducteurs ayant commis au moins un acte de conduite
potentiellement dangereux dans le contexte de l'accident (p. ex., dépasser la limite de vitesse); les
témoins étaient des conducteurs dont la conduite n'avait pas été dangereuse lors de l'accident.
Nous avons calculé les rapports de cotes (RC) bruts et ajustés de tout acte de conduite
potentiellement dangereux chez les conducteurs déclarés positifs pour le cannabis, mais négatifs
pour la consommation d'alcool. Dans notre calcul des RC ajustés, nous avons tenu compte de
l'âge, du sexe et du dossier de conduite antérieur.
Résultats : Cinq p. cent des conducteurs avaient été déclarés positifs pour le cannabis. Le RC brut
d'un acte de conduite potentiellement dangereux était de 1,39 (IC de 99 % = 1,21-1,59) pour les
conducteurs déclarés positifs. Même compte tenu de l'âge, du sexe et du dossier de conduite
antérieur, la présence de cannabis demeurait associée à un risque plus élevé d'avoir eu une
conduite potentiellement dangereuse (1,29, IC de 99 % = 1,11-1,50).
Conclusion : Le cannabis a eu un effet néfaste sur la conduite, comme on pouvait le prédire d'après
les études de performance humaine. Cette constatation confirme la nécessité d'intervenir pour
réduire la prévalence de la conduite avec facultés affaiblies par le cannabis et montre qu'il faudrait
pousser la recherche sur la relation dose-réponse entre le cannabis et la prudence au volant.
Public Health in Canada is at a crossroads…
Since 1910, CPHA has been the only national voice for Public Health in
Canada. The Association:
 encourages citizen involvement in Public Health policy and
programming;
 brings together diverse individuals and organizations, creating a
united voice on Public Health issues in Canada and around the
world; and
 champions universal and equitable access to the basic conditions
necessary to achieve health for all.
CPHA’s strength is its members who give us credibility, direction and
authority. To continue to be the voice of Public Health, CPHA needs your
expertise and support.
Join your voice to ours.
Join CPHA today.
Call us at 613-725-3769 ext. 118, e-mail us at « membership@cpha.ca »
or visit us on-line at www.cpha.ca/membership
Canadian Public
Health Association
The voice of Public Health
in Canada since 1910.