Age and driving

1
Age and driving
2
www.automoblog.net an example of the prejudice
against elderly drivers that is largely
unsupported by scientific research...
3
"Egon", at Automoblog, goes on to
say "...banning old people from driving
completely seems a little bit extreme. But what
about people who cant react in a timely manner,
cant make reasonable decisions, and cant see or
hear as well as other drivers? Allowing them on
the road as freely as anybody else is as
ridiculous as allowing drunk drivers on the
road... ...I propose that beginning at age 65
(maybe 70), the driver takes a strict driving
test every year. Obviously this will keep the
dangerous ones off the road, and the competent
ones can still drive".
4
The case of George Weller
5
How risky are elderly drivers? What kinds of
problem do elderly drivers have? What kinds of
accidents do they have? How can risky elderly
drivers be identified?
6
How risky are elderly drivers? Preusser et al
(1998) FARS data. 65-69 year-olds are 1.45 times
more likely to have a fatal accident than 40-49
year-olds. Evans (1988) data from FARS and
Nationwide Transportation Study used to calculate
fatal crashes per unit distance travelled.
Compared to 40 year-olds, 65 year-old male
drivers had 33 more fatal crashes, 65 year-old
female drivers had 77 more.
7
Complications in assessing elderly drivers' level
of risk Frailty Older drivers are more easily
killed in an accident (e.g. Sjöjren et al 1996,
Li et al 2003). Exposure to risk No good
detailed statistics on different age-groups'
driving behaviour. Poor definitions of
"elderly" e.g. Shanmugaratnam et al (2010)
average age of "elderly" group 65, but ranged
from 44 to 82 !
8
Problem of assessing exposure to risk Elderly
drivers drive less, prefer familiar roads and low
speeds, avoid night driving, heavy traffic and
difficult situations. Therefore might be even
riskier than statistics suggest.
9
Hakamies-Blomqvist, Raitenan and O'Neill
(2002) Apparent increased accident rate for
elderly drivers is an artefact. Calculating
accident rates per unit distance travelled
ignores the fact that high-mileage drivers do
most of their mileage on high-speed roads (which
are safer). Therefore elderly drivers might be
less risky than the statistics suggest!
10
Typical accident patterns of elderly drivers
Hakamies-Blomqvist (1994, 1995) More likely to
be involved in collisions (ROW violations). Less
likely to have single-vehicle accidents. Less
likely to speed, drink-drive, commit major
violations.
Keltner and Johnson (1987) ROW violations
increase from age 50. Turn violations increase
from age 65. Stop sign violations show U-shaped
function - common in young and 60. Speeding
decreases linearly with age.
11
Verhaegen (1995) data on 2,400 two-car
collisions in Belgium. For each age-group,
obtained "over-involvement ratio" - number of
at-fault drivers/ number of blameless
drivers. 50 year-old drivers more likely to be
responsible for their collision than 23-49
year-olds. For 9/10 accident types, over 50's
more likely to be at fault than to be the victim.
12
Preusser et al (1998) Compared to 40-49 year
olds, 65-69 year-olds were over twice as likely
to be involved in a crash at an intersection,
1.29 times more at risk in all other
situations. 85 were 10.62 and 3.74 times as
likely. Most common error was "ran traffic
control" - failing to stop or give way. Accounted
for over half of all fatal accidents for drivers
aged 74. Compared to 40-49 year-olds, 65-69 year
olds were twice as likely to be involved in ROW
violation at an intersection. 85 year olds were
34 times as likely.
13
Daigneault, Joly and Frigon (2002) Accident and
conviction records of 426,000 Quebec drivers aged
60-80. Crashes due to violations decrease with
age, crashes due to errors increase. Elderly
drivers take less risks (speeding, overtaking)
but have more problems at junctions.
14
Healthy and unhealthy elderly drivers Hakamies-Bl
omqvist (1996) Normal healthy ageing is not
associated with a significant decrease in driving
ability. Ageing increases the risk of health
problems (visual pathology, dementia, etc.). Ball
et al (1998) A subset of impaired drivers, at
greater risk of an accident, inflate the crash
risk for the entire age-group. Evans (1993)
Need to distinguish between age and cohort
effects.
15
Why are elderly drivers more at risk? Visual
problems? Both natural and pathological eyesight
deterioration occurs with age (e.g. presbyopia,
glaucoma, hemianopia).
Davison and Irving (1980) 1400 drivers mean
acuity at 20 yrs 6/4.2, at 70 6/7.3, still
good enough for driving (legal limit is roughly
6/9 in U.K. Drasdo and Haggerty 1981).
Percentage of drivers in each age-group failing
to meet various levels of Snellen acuity
16
Why are elderly drivers more at
risk? Attentional problems? Useful Field of View
test (e.g. Owsley et al 1991 Ball et al 1993).
UFOV score correlates better with "at fault"
accident rates, than do other measures of visual
performance.
Suggests elderly drivers may be less able to
distribute their attention to different parts of
the visual field.
17
Why are elderly drivers more at risk? Information
processing problems (cognitive slowing)? Cerella
(1985), Salthouse (1991, 1996) Age leads to
slower processing speed, which leads to impaired
performance (in both speed and accuracy).
Differences are small in absolute terms. Wood and
Troutbeck (1995) Effects of simulated visual
impairment on driving, in young and old
drivers. Goggles simulated effects of cataracts,
visual field restriction, monocular loss. Older
drivers affected more than younger drivers -
implying former's deficits are not solely due to
visual impairment, but lie elsewhere (cognitive
slowing?)
18
Hole (2007) cognitive slowing as an explanation
of junction accidents Driving is a hierarchy of
operations (Rasmussen 1983). Operational and
strategic levels unaffected in elderly. Cognitive
slowing mainly affects tactical level - but since
driving is largely self-paced, elderly can cope
by driving slower and more cautiously. However,
junctions demand effective attention and rapid
information-processing.
The "window of opportunity" for emerging at a
junction
19
Hakamies-Blomqvist (1996) intersections and
ageing
Functional domain Task demands Age-related change
Motor performance Do complex vehicle-handling movement sequences swiftly, while attending to traffic. Motor slowing. Seriality in movement organisation.
Perception Detect objects in peripheral vision. Perceive movement. Estimate own and other's speed. Visual field shrinking, Decreased dynamic acuity. Speed estimates for self-performed actions may be outdated.
Attention Divide attention between environment and car handling, focus on task-relevant aspects of traffic Difficulties in divided attention tasks, selective attention tasks, ignoring irrelevant information.
Interaction with other road users Predict other's behaviour and behave predictably. Slowness in approaching intersections misinterpreted as reflecting intention to give way.
20
How can unfit elderly drivers be identified? Many
elderly drivers have insight into their problems
and restrict their driving to conditions they can
cope with. Up to 45 of dementia patients
continue to drive (e.g. O'Neill et al 1992,
Lucas-Blaustein et al 1988). Dobbs, Heller and
Schopflocher (1998) Driving test focusing on
lane changes and traffic merging. Failed by 3 of
young normal, 25 of elderly normal and 68 of
elderly early dementia drivers. Demented group
made many more hazardous errors 50 occurred
while changing lanes, merging with traffic or
approaching an intersection 21 during left
turns 15 due to failing to stop at an
intersection.
21
Marottoli and Richardson (1988) Relationship
between self-ratings of confidence and driving
performance (accidents and violations) in 126
drivers aged 77. 50 of these drivers also had a
driving test. None rated themselves as worse
than their peers. 32 rated themselves "the same
as" their peers. 44 rated themselves "a little
bit better than" their peers. 24 rated
themselves as "much better than" their
peers. Actually no relationship between accident
/violations record and self-rated driving ability.
22
Marottoli and Richardson (1988) "Despite the
fact that 40 of the cohort reported a history of
adverse events, and 27 of those participating in
the driving performance sub-study were rated...as
exhibiting moderate or major difficulties, all
drivers rated themselves as being average to
above average drivers...Thus, objective evidence
of driving ability... did not appear to impact on
a person's confidence or self-rating of
abilities..." Freund et al (2005) 47 elderly
drivers referred for testing. 38 were found to
be unsafe all of these rated their expected
simulator driving performance to be the same or
better than other drivers of their age.
23
Conclusions The "elderly driver problem" is
overstated as a group, have fewer accidents and
commit fewer violations than younger drivers
(Evans 1988). Age per se is a poor indicator of
driving performance (Hakamies-Blomqvist
1996). Bimodal distribution elderly drivers are
a mix of healthy/safe and unhealthy/unsafe
drivers. Some evidence that elderly drivers are
more at risk of an accident at intersections. Rest
ricted UFOV and dementia are health problems that
particularly increase risk of an accident.
24
Conclusions Shanmugaratnam et al (2010) elderly
drivers' overall performance is a mixture of
beneficial effects of judgement (drive slower,
avoid difficult conditions) and adverse effects
of impaired ability (cognitive and psychomotor
slowing). Problem of reliably discriminating
between safe and unsafe elderly drivers
large-scale screening is unreliable and not
cost-effective. A growing problem in all
developed countries, as drivers live longer and
are less prepared to quit driving. Sivak et al
(1995) 1995, 1 in 7 U.S. drivers were 65. 2030
will be 1 in 5. Retchin and Arapolle (1993) by
2020, will be 33 million U.S. drivers aged 65.