BRIEF REPORT
Identifying Older Drivers at Risk of Traffic Violations by
Using a Driving Simulator: A 3-Year Longitudinal Study
Hoe C. Lee
Andy H. Lee
OBJECTIVES. This prospective longitudinal study aims to determine which simulated driving tasks of a personal computer (PC)-based driving simulator can be used to identify problematic older drivers, using their
3-year driver violation points record as the outcome measure.
METHODS. A total of 129 urban community-dwelling older drivers volunteered to participate in the study.
Using a driving simulator, specific driving tasks were devised to test the performance of older drivers. Their
officially recorded driver violation points were retrieved immediately after the simulated driving assessment and
thereafter for the following 2 years. Self-reported driving records were also collected during the same period.
Hierarchical Poisson regression analysis, adjusting for gender, age, and driving exposure (hours of driving per
week), was then undertaken to determine those driving tasks that affected the frequency of traffic violations.
RESULTS. All participants incurred at least one driver violation point during the 3-year period. The simulated driving tasks found to be significantly associated with the incidence of traffic violations were working memory and use of indicator.
CONCLUSIONS. This longitudinal study demonstrated that the driving simulator was able to identify unsafe
older drivers at risk of traffic violations if appropriate simulated driving tasks were used. Such a screening tool
should be adopted prior to administering a more detailed but expensive road test.
Lee, H. C., & Lee, A. C. (2005). Brief Report—Identifying older drivers at risk of traffic violations by using a driving simulator: A 3-year longitudinal study. American Journal of Occupational Therapy, 59, 97–100.
Hoe C. Lee, PhD, is Lecturer, School of Occupational
Therapy, Curtin University of Technology, GPO Box U1987,
Perth, WA 6845, Australia; [email protected]
Andy H. Lee , PhD, is Associate Professor, Department of
Epidemiology and Biostatistics, School of Public Health,
Curtin University of Technology, Australia.
T
he independence of older persons to live
in the community will be severely compromised if they are unable to continue
driving (Andiel & Liu, 1995). Occupational therapists have made significant contributions in helping disabled and older
people to drive safely (Galski, Ehle, &
Willliam, 1997; Lloyd et al., 2001). In
2002, the American Occupational Therapy
Association acknowledged the importance
of driver rehabilitation and nominated it to
be among the top 10 emerging practice
areas for the occupational therapy profession. The challenge is to develop appropriate evaluation methods of identifying those
drivers at high risk of traffic violations and
to provide effective intervention as early as
possible (Lee, Lee, Cameron, & Li, 2003).
Driving performance of the older population has been investigated by simulator
technology (Rizzo, McGehee, Dawson, &
Anderson, 2001). The relatively small size
of the monitor display, together with the
nature of the computer-generated stimuli,
have somewhat limited the simulator in
assessing driving tasks that require the use
of visual details in complex traffic scenarios.
Nevertheless, low-cost personal computer
(PC)-based simulators can effectively measure cognitive and perceptual abilities in
driving (Janke, 2001). Lee, Lee, Cameron,
and Li (in press) showed that it is possible
to differentiate older drivers with various
levels of visual attention skill by using a
driving simulator. Lee, Cameron, & Lee
(2003) validated a driving simulator to
measure on-road driving performance and
reported that observations from simulated
driving assessment could explain over 67%
of on-road driving behavior. A retrospective
study (Lee, Lee et al., 2003) further found
that older drivers at inflated risk of crashes
could be identified through simulated driving criteria. The results confirmed that cognitive skills such as working memory, ability to make rapid decisions, judgment under
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930191/ on 06/15/2017 Terms of Use: http://AOTA.org/terms
97
time pressure, and confidence in driving at
high speed, were associated with the crash
event. The aim of this prospective longitudinal study was to investigate the sensitivity
of a driving simulator in identifying problematic older drivers based on their 3-year
driver violation point records. A longitudinal design was adopted because the study
could be better controlled and hence provide a more reliable source of evidence than
retrospective studies. Although the violation point system varies from state to state
and points penalized do not necessarily
reflect the severity of driving delinquency, a
license will be revoked in this state when 12
violation points are accrued within 2 years.
October 2002. Self-reported recall of driver
violation points and average weekly driving
exposure were also obtained via telephone
interview at the same time points.
Simulator Driving Tasks
The STISIM Driving Simulator (Allen,
Stein, Aponso, Rosenthal, & Hogue, 1990;
Lee, Lee, & Cameron, 2003) was used to
study the behavior of the participants.
Based on the task analysis process described
by Levine and Brayley (1991), a computer
program was developed and previously test-
Table 1. Simulated Driving Tasks To Assess the Driving Skills of Participants
Simulated driving tasks to be
accomplished during the assessment
Methods
Participants and Procedure
Ethics approval from the researchers’ institution was obtained prior to commencement of the study. During October 2000,
129 older drivers residing in Perth, Western
Australia, and ages 60 years or older, volunteered to take part in the study. The participants were required to hold a valid driving
license. Each assessment included a 30minute initial interview, a 45-minute simulated driving session, and a 20-minute
postassessment feedback session. A laboratory technician, who was blinded to the
driving history of the participants, monitored the simulated driving assessment. In
the event that the performance of a participant indicated unsafe driving practice as
judged by the principal investigator, free
counseling and advice were available. With
the permission of the individual concerned,
a follow-up referral involving an occupational therapist would be arranged to
address the issue.
At the initial interview, the frequency
of motor vehicle crashes and traffic violations were collected from each participant.
Confidentiality on the information provided was assured. Written authorization to
access official driving records was sought
from all participants. Immediately after the
simulated driving session, their driver violation points for the current year were
retrieved. Driver violation points incurred
during the following 2 years were subsequently retrieved in October 2001 and
98
ed (Lee, Cameron et al., 2003) to simulate
and assess 10 driving tasks specifically for
testing older adults. A consensus panel,
consisting of occupational therapists, geriatricians, clinical psychologists and experienced driving instructors, was consulted on
the appropriateness of the driving tasks
being applied to the older population. After
the tasks were developed and trialled with
older drivers, the panel was reconvened to
fine-tune the scoring mechanisms. Table 1
provides a brief description of the driving
tasks and associated scoring mechanisms.
Measure of the participants’
performance in the simulated driving task
with maximum possible scores in brackets
Rule Compliance: Lane changing in double-laned
road, where participant’s car was in the right lane.
KEEP RIGHT signs displayed every 55 yards to
prompt participants to go back to the inner lane
Follow “keep to right” rule; voluntarily (2) or with
visual prompt (1); Check traffic by head turn (1) with
rear mirror (1); proper use of indicators (up to
maximum of four points)
Traffic Sign Compliance: Drive through “STOP,”
“GIVE WAY” and pedestrian crossings safely
Approach slowly “STOP,” “GIVE WAY” and pedestrian
crossings (1); Stop in right place (1); Give way as
required (1); Proceed when opportunity comes (1);
Correct use of indicators (1) Check mirror before
proceed (1)
Driving Speed: Drive 1.5 miles along the road
according to the designated speed of double-laned
straight road (40 miles/hr speed limit)
Speed (1.25 mile/run time of the distance)
Use of Indicator: Drive around “Road Work”
obstacles blocking the road and return to the
inner lane as soon as possible
Signal to the right and left to change lane (2, one
each); Check traffic (2); Voluntarily return to inner
lane (2)
Road Use Obligation: Observe traffic conditions,
make decision while manoeuvring through
T-junctions leading to main road with STOP
signs safely
Approach “Road work” scenario twice with caution
and slow down (1); Indicate right or left turn in
driving pass (1); Proceed at appropriate opportunity
(1); Check traffic with head turn (1) or rear mirror
(1); and Correct use of indicators (1)
Decision and Judgement: Avoid colliding with
pedestrians running across the road hastily,
vehicle parked on the road side moving out
without signalling and vehicle in front suddenly
slowing down
One mark for each success in avoiding crash when
confronted with simulated dangerous driving
scenarios
Working Memory: Remember five street names
and five manoeuvres (turn left or right) marked
on a route on a road map to a fictitious park in
5 minutes and recall them after 10 minutes’
simulated driving
Names and maneuvers recalled (1 for each correct
answer, up to maximum of 8) Sequence of maneuvers
(3: in perfect order; 2: 2-3 correct; 1: 1 correct and 0:
none correct)
Multi-Tasks: Starting with a nominal total of
100, take away “5” every time the “SUBTRACT”
billboard comes out. Fifteen billboards with
“SUBTRACT” sign were posted along the road
Correct answer to the visual stimulus of “SUBTRACT”
sign (1)
Speed Compliance: Observe and maintain a
speed close to the posted speed limits (40, 45,
and 70 miles/hr), which vary according to traffic
conditions
Number of tokens received when the driving speed is
close to the designated speed (± 3 miles/hr)
Visual Attention Tasks: Signal the traffic indicator
when the “diamond” shapes on the monitor
screen change to “triangle” shapes incidentally
for 15 seconds
Responses to the change of Visual Attention Symbol
to “triangle” shapes
January/February 2005, Volume 59, Number 1
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930191/ on 06/15/2017 Terms of Use: http://AOTA.org/terms
The driving task measures of driving speed,
use of indicator, decision and judgment, speed
compliance, and visual attention tasks were
automatically recorded by the simulator
computer whereas the remaining five
measures were collected by the laboratory
technician.
Statistical Analysis
All data were coded and analyzed using
STATA version 8 (Stata Corporation,
2003). A threefold analytic approach of
descriptive statistics, reliability check, and
inferential analysis was adopted. Reliability
of the driving task measure scales was
assessed by Cronbach alpha coefficient. The
repeated measure of driver violation points
was the main outcome variable of interest.
In view of the dependency of the observations, hierarchical Poisson regression analysis based on generalized estimating equations (Hardin & Hilbe, 2002) was
undertaken to determine a subset of the
simulated driving tasks affecting the frequency of traffic violations over the 3-year
period. The method is an extension of standard Poisson regression to handle the hierarchical data (repeated outcomes from the
same subject). It accommodates the inherent correlation of the repeated observations
clustered within a subject and provides
robust standard errors for the Poisson
regression coefficients, so that correct inferences can be made. A statistically significant
relationship would provide evidence of
validity of the simulator to successfully
identify older drivers at inflated risk of traffic violations.
Results
truck drivers prior to retirement, which
required driving a vehicle regularly.
According to the state records, all participants incurred at least one driver violation
point within the past 3 years, whereas the
maximum yearly point count was five. The
most common reason of traffic violation
reported by the participants was speeding.
Despite 9% of participants developing and
reporting some degree of simulator sickness
such as mild dizziness during the simulated
driving session, the symptoms only lasted
for a short time and did not prohibit them
from undertaking the assessment. None of
the participants requested or was required a
follow-up referral or counseling.
Reliability Check
The measurement properties of the 10 simulated driving tasks were next examined.
The Cronbach alpha coefficient was found
to be 0.742, confirming the internal consistency of the scales that measured performance in the simulated driving tasks. As
expected, there was substantial negative correlation between each measure and the age
of participants (correlation ranged from
–0.303 to –0.596), suggesting that simulated driving performance could deteriorate
with increasing age.
Inferential Analysis
The objective was to identify a subset of
simulated driving tasks related to the official and self-reported records of driver violation points. Six of the 129 participants
could not be contacted at second or third
year, or both, resulting in N = 380 observa-
tions and a loss-to-follow up rate of 4.7%.
The agreement between the self-reported
and state records was significantly strong
(Kappa statistic = 0.72, p = 0.001),
although 48 out of the 380 self-reports
(12.63%) yielded lower count than the corresponding official records. Results of the
hierarchical Poisson regression analysis of
state data are presented in Table 2.
Adjusting for age, gender, and individual
driving exposure, performance in driving
tasks involving working memory and use of
indicator was found to be significantly associated with the driver violation points.
Goodness of fit of the hierarchical Poisson
regression model was satisfactory (Chisquare = 362.19, p = 0.561). The 11 participants suffering simulator sickness were
next excluded, but neither the model fit nor
the significance of the variables could be
affected by eliminating these observations.
Analysis of the self-reported data also produced similar results and was omitted from
the reported results for brevity.
Discussion
In this study, appropriate driving tasks were
chosen to test older drivers using a PCbased driving simulator. Ten internally consistent measures of driving task were used to
assess each participant’s performance in
these tasks. The negative correlation
between individual assessment score and
age confirmed that driving skills generally
decline with age (Brayne et al., 2000). The
overall agreement between the self-reported
and officially recorded driver violation
Table 2. Hierarchical Poisson Regression Results for State Recorded Driver Violation Points,
Adjusting for Clustering and Individual Driving Exposure (Hours Per Week) (N = 380)
Descriptive Analysis
The age of the 129 participants ranged
between 60 and 88 years (mean 72.9, SD
7.1), and 22% of the drivers were female.
The self-reported estimated driving hours
per week for each individual did not vary
significantly throughout the 3-year period
(overall mean 11.22; SD 8.56), according
to repeated measures analysis of variance
(ANOVA) (p = 0.455). Shopping and social
activities were the most common reasons
that participants drove. Around 12% of
them were either courier workers, taxi or
95% Confidence interval
Variable
Rule Compliance
Traffic Sign Compliance
Driving Speed
Use of Indicator
Road Use Obligation
Decision and Judgement
Working Memory
Multi-Tasks
Speed Compliance
Visual Attention Tasks
Age
Gender
Incidence rate ratio
1.09
1.17
0.99
0.77
1.00
0.99
1.16
0.96
1.05
0.98
1.00
0.82
p
0.271
0.109
0.063
0.014**
0.991
0.749
0.015**
0.281
0.141
0.316
0.913
0.378
Lower
Upper
0.93
0.96
0.98
0.62
0.90
0.89
1.03
0.89
0.98
0.93
0.97
0.53
1.27
1.40
1.00
0.94
1.10
1.08
1.31
1.03
1.11
1.02
1.03
1.27
** p < 0.05
The American Journal of Occupational Therapy
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930191/ on 06/15/2017 Terms of Use: http://AOTA.org/terms
99
points was significantly strong, although
the self-reported data yielded fewer counts
than the corresponding state records. While
the volunteered participants should have no
apparent reason to provide untrue information, their recall of driving history cannot
be guaranteed (Voelker, 1999).
This longitudinal study further investigated the identification of older drivers at
inflated risk of adverse traffic events. The
hierarchical Poisson regression analysis
showed that appropriate use of indicator
would reduce the incidence rate of traffic
violations. The ability to use the vehicle
indicator properly requires well-timed reaction (Ball & Rebok, 1994) and dexterous
motor coordination (Bourwer & Ponds,
1994), which may deteriorate with physiological aging and affect driving skills
(Graca, 1986; Sims, McGwin, Allman,
Ball, & Owsley, 2000). The positive association between working memory and driver
violation points is also logical, because
drivers with good working memory and confidence tend to drive above the speed limit
(Lee, Lee et al., 2003). The results are also
consistent with previous findings that driving simulators can reliably reflect on-road
driving behaviors and functions (Lee,
Cameron et al., 2003). Our findings suggest that practitioners should target working memory and correct use of indicator
when assessing older adults for safe driving.
It should be remarked that the driver
violation points retrieved did not reflect the
severity of the traffic offence(s) committed
(e.g., driving through a red light versus several minor speeding violations). Nevertheless, the official point record system does
provide a reliable source of driver violations
over a specific time period.
Limitations of the study included
unequal gender distribution. Therefore,
gender was controlled for in the hierarchical Poisson regression analysis. The participants who volunteered for this study cannot be taken as representative of the
population of older drivers, because the
sample was not randomly selected but only
covered some sectors of the community.
Selection bias was therefore unavoidable in
the recruitment of the participants. However, random sampling was neither possible
100
nor practical in this type of study.
At the present time, PC-based simulators can generate sufficiently complex traffic scenarios comparable to the on-road
environment and conditions. This prospective longitudinal study confirms that a relatively low-cost driving simulator can be
used by occupational therapists as an offroad screening tool to identify older drivers
at risk of traffic violations. With further
research, the authors believe such simulator
technology can also be employed for training incompetent drivers.▲
Acknowledgments
This study was supported by a research
grant from Curtin University of Technology. The authors are indebted to the
Royal Automobile Club of Western
Australia and the Council on the Ageing for
assisting the recruitment of participants.
Thanks are also due to the editor and two
anonymous reviewers for their constructive
comments and suggestions. The authors
also wish to thank Occupational Therapists’
Registration Board (WA) for financial support of this research.
References
Allen, R. W., Stein, C. A., Aponso, B. L,
Rosenthal, T. J., & Hogue, J. R. (1990).
A low cost, part task driving simulator based
on microcomputer technology. Paper presented at the 69th Annual Meeting of the
Transportation Research Board, Washington, DC.
Andiel, C., & Liu, L. (1995). Working memory
and older drivers: Implications for occupational therapy. American Journal of Occupational Therapy, 49, 681–686.
Ball, K., & Rebok, G. (1994). Evaluating the
driving ability of older drivers. Journal of
Applied Gerontology, 13(1), 20–38.
Bourwer, W., & Ponds, R. (1994). Driving competence in older persons. Disability and
Rehabilitation, 16(3), 149–161.
Brayne, C., Dufouil, C., Ahmed, A., Dening, T.
R., Chi, L. Y., McGee, M., et al. (2000).
Very old drivers: Findings from a population cohort of people aged 84 and over.
International Journal of Epidemiology, 29,
704–707.
Galski, T., Ehle, H. T., & Willliam, J. B. (1997).
Off-road driving evaluations for persons
with cerebral injury: a factor analytic study
of pre-driver and simulator testing.
American Journal of Occupational Therapy,
51, 352–359.
Graca, J. L. (1986). Driving and aging. Clinics
in Geriatric Medicine, 2(3), 577–589.
Hardin, J. W., & Hilbe, J. M. (2002). Generalized estimating equations. Boca Raton, FL:
Chapman & Hall/CRC.
Janke, M. K. (2001). Assessing older drivers.
Two studies. Journal of Safety Research, 32,
43–74.
Lee, H. C., Cameron, D., & Lee, A. H. (2003).
Assessing the driving performance of older
adult drivers: On-road versus simulated
driving. Accident Analysis and Prevention,
35(5), 797–803.
Lee, H. C., Lee, A. H., & Cameron, D. (2003).
Validation of a driving simulator by measuring the visual attention skill of older
adult drivers. American Journal of Occupational Therapy, 57, 324–328.
Lee, H. C., Lee, A. H., Cameron, D., & Li, C.
(2003). Using a driving simulator to identify older drivers at inflated risk of motor
vehicle crashes. Journal of Safety Research,
34, 453–459.
Levine, R. E., & Brayley, C.R. (1991).
Occupation as a therapeutic medium—A
contextual approach to performance intervention. In C. Christianse & C. Baum
(Eds.), Occupational therapy—overcoming
human performance deficits (pp. 592–631).
Thorofare, NJ: Slack.
Lloyd, S., Cormack, C. N., Blais, K., Messeri,
G., McCallum, M. A., Spicer, K., et al.
(2001). Driving and dementia: A review of
the literature. Canadian Journal of Occupational Therapy, 68(3), 149–156.
Rizzo, M., McGehee, D. V., Dawson, J. D., &
Anderson, S. N. (2001). Simulated car
crashes at intersections in drivers with
Alzheimer disease. Alzheimer Disease and
Associated Disorders, 15(1), 10–20.
Sims, R.V., McGwin, G., Jr., Allman, R. M.,
Ball, K., & Owsley, C. (2000). Exploratory
study of incident vehicle crashes among
older drivers. Journal of Gerontology A:
Biological Sciences and Medical Sciences, 55,
22–27.
Stata Corporation (2003). STATA Release 8.
College Station, TX.
Voelker, R. (1999). Crash risk among older
drivers studies. JAMA, 282, 1610–1611.
January/February 2005, Volume 59, Number 1
Downloaded From: http://ajot.aota.org/pdfaccess.ashx?url=/data/journals/ajot/930191/ on 06/15/2017 Terms of Use: http://AOTA.org/terms