How managers can influence professional drivers'
A. E. af Wåhlberg
Department of Psychology
P. O. Box 1225
751 42 Uppsala
Tel: +46-18-471 25 90, +46-18-33 90 95
Fax: +46-18-471 21 23
People who drive as part of their work have different accident liabilities. Some of this
variation is due to factors that are within the manager's reach with various methods.
As help for managers (especially of car fleets), a limited review of the scientific
literature was written, concentrating on some variables, which have been shown to
relate to accidents, and various methods for changing these and increasing driving
safety in the company by management. Also, some popular methods for which there
is no evidence of a positive effect were discussed, including driver skills training.
The following variables have been identified as accident risk factors for drivers, and
are also to some degree possible for managers to influence;
Specific methods, which are recommended for managers to increase the safety of
Increase planned driving times
Schedule for rest breaks, which can also be used for office work
Investigate/improve driver health and sleep quality
Discourage use of alcohol the night before driving
Discourage cell phone use while driving
Use low performance vehicles (they also have lower fuel consumption)
Use feedback and reinforcement as closely as possible in time to the behaviour that is
to be changed
Methods that are not recommended;
Extra safety features of cars (if these are known to the drivers)
Transports of every kind are growing every day. Among the many groups of drivers,
those who drive for a living have often been singled out as different from others, due
to the large amount of driving they undertake, their resulting great experience, the
stress and hassles, but also the road accidents that inevitably follow. Several
researchers have reported that professionals are more involved in crashes than private
motorists, even when exposure has been accounted for (e.g. Chapman, Roberts &
Underwood, 2000; Broughton et al, 2003). Also, as they spend so much time on the
road, these accidents are a large part of the total number of work-related deaths in the
world (almost half in Australia; Mitchell, Driscoll & Healey, 2004; see also Boufous
& Williamson, 2006); while in Greece, 25 percent of all accidents involve heavy
trucks; Tzamalouka, Papadakaki & Chliaoutakis, 2005). In Sweden, professional
drivers have the highest numbers of injured and killed per employed person, and the
fatal cases are about one tenth of the total for the country (Bylund, Björnstig &
Larsson, 1997). In terms of costs, Miller and Galbraith (1995) calculated that motor
vehicle crashes caused 16.7 percent of the annual total of 140 billion dollars of worth
for occupational injury in the United States.
Here, it must be remembered that the relation between exposure and accidents is
curvilinear, i.e. at very high levels of exposure, there is hardly any increase in
accidents at all. Also, accidents decrease with experience. As professionals drive
much more than the average driver, this means that the difference is even more
pronounced, especially at low levels of exposure and experience. However, direct
comparisons between drivers with similar values on these two parameters are rare in
the scientific literature.
Individual differences in traffic accident liability have been researched for almost a
hundred years. During this time, it has mainly been found that there is no single
strongly predictive variable, which in many ways make sense, because there are many
ways that accidents can be caused; intentional dangerous driving, drug impairment,
fatigue, sensory weaknesses, inexperience etc. Also, these factors in turn can be
caused by a multitude of distal factors.
Therefore, any single, specific, traffic safety intervention is unlikely to have any
positive effect. For example, driving skills training may improve the actual technical
handling of the vehicle, but this does not necessarily mean that there is a safety effect.
Such knowledge, which is fairly common within the traffic research community, does
not seem to have impacted upon the transport industry, despite the need for effective
safety programs within this area. Given this situation, it was conceived that a guide
for transport managers should be written, summarizing the knowledge about factors,
which can be influenced by the company. It should be noted that this aim is different
from the question of how to select drivers in the first place. Given that the available
predictors are weak and the selection not very strong (i.e. there are not that many
potential drivers to chose from), it is not a strong safety tool. Instead, the aim of this
paper is to describe how managers can control factors of some importance, creating a
safety conscious driving environment. This means that many factors, which have been
shown to have an association with accidents, will be only briefly described, because
they cannot be influenced, but only de-selected by not hiring certain people.
In the present paper, a review was undertaken of available evidence concerning
individual differences in accident record and some factors that are considered to cause
accidents. The research included has mainly used accidents as unit of analysis, or as
outcome. This means that research using parameters that are assumed to be related to
crashes as dependent variables, like near misses, traffic conflicts, errors etc, have been
excluded. The difference in validity as evidence between these two types of study is
unfortunately seldom recognized.
Are professional drivers different in any way from the general population, i.e. are they
selected in some way and/or do they change their behaviour due to their work? No
such evidence has been presented, and the question not even asked, so the only thing
that can be really ascertained is that these drivers are much more experienced than the
private drivers. From the perspective of the driver, it is therefore not today possible to
explain the high rate of accidents of professional drivers, and we therefore turn to the
driving environment, meaning mainly the circumstances under which the driver is
working, and the consequences this has upon driver states (fatigue, for example).
In general, it has been found that professional drivers have very tough work
environments, and that the result is very negative for their general well being. Here,
bus drivers tend to stand out from other categories of drivers and workers; they have
high rates of cardiovascular disease, gastrointestinal problems, musculoskeletal
disorders, psychoticism, post-traumatic stress syndrome, alcohol use, absenteeism,
labour turnover and possibly a number of other negative outcomes which have been
less thoroughly researched (see review by Tse, Flin & Mearns, 2006). These effects
are thought to be due to different stressors (in a wide sense), like time pressure, bad
ergonomics, low control of work situation, conflicting demands (safety and service
versus time table), fear of assault etc. Several of these factors are probably defining
features of other drivers' work situation too, but to a lesser degree. Unfortunately,
other professional drivers are not as well studied as bus operators, but similar
although milder effects have been found for truck drivers (Meuleners, Lee, Legge &
Cercarelli, 2005), and can be suspected to exist for those who drive as part of their
work. Driving is simply a stressful experience.
Reviews of factors predictive of individual accident
For those interested in factors not covered in this study, several reviews are available.
For general overviews of several groups of predictors, see Donovan, Marlatt and
Salzberg (1983), Lester (1991) and af Wåhlberg (2003). However, as with other
research, one should not uncritically accept the conclusions of reviews. Here, it is
important to see whether the authors discuss the methodology of the studies included.
If not, extra caution is needed on behalf of the reader.
Personality as an accident predictor has been extensively researched, and reviews
have been published on the general concept as well as more limited parts of it. For
example, Jonah (1997) summarized the sensation seeking trait literature, and Clarke
and Robertson (2005) meta-analysed the Big Five personality reports. Low values on
the conscientiousness and agreeableness personality traits were found to be weakly
related to accident involvement in occupational settings.
In California, the Department of Motor Vehicles has for several decades been
investigating the type of variables which are gathered for the state records, like traffic
violations, but also demographics. Much of this research was summarized by Peck
(1993). Many of these reports are available from the department's homepage. In
general, these associations have been found to be very weak.
Information processing capacity as accident predictor was rather popular in the 70'ies
and 80'ies, but the interest seem to have waned, despite some rather strong effects
found (Arthur, Barrett & Alexander, 1991). Some of these studies were reviewed in
Shinar (1993), which also described the general hypotheses and background of this
research. A similar approach was that of field dependence, reviewed by Goodenough
Some accident factors that are possible for managers to
Sleep and fatigue
Many researchers claim that fatigue is a major reason for many traffic accidents;
actually, figures up to 40 percent have been stated. However, this statement can be
questioned, mainly due to the methods used for estimating fatigue and its contribution
to crashes. Also, the estimates differ rather wildly between sources.
Here, it is useful to distinguish between two or three ways of doing research in this
area, which yield slightly different answers, because the specific question asked
differs, as do the methods. First, there is the overall question of how many accidents
are caused by fatigue (falling asleep at the wheel, low level of attention, slow
reactions), where the common method is crash investigation, trying to determine in
each case whether fatigue was implicated as a cause, or possibly present. The second
is a more specific question; are people who are generally fatigued more crash
involved than those who are not? The third is an in-between variant of these that ask
whether crashes occur when it can be assumed that fatigue is high, like at the end of
long shifts of driving.
This means that the crash investigations will probably yield the highest estimate for
accident causation, because they will canvass all fatigue situations, regardless of
reason, while on the other hand they will only sample the most serious incidents. The
other methods are more inclusive concerning seriousness of accident, but exclude
those that are due to random fatigue states. It should also be remembered that the
definition of fatigue differs a lot between studies, and that this research area therefore
has very fuzzy borders.
So, what are the figures concerning fatigue and crashes reported in the literature?
McCartt, Ribner, Pack and Hammer (1996) asked drivers various questions about
drowsy driving, and more than half said that they had been driving while drowsy the
last year. Several percent admitted to crashing in such a state or while asleep. This can
be compared to the percentage of accidents which is attributed to sleepiness; 16-23
percent in the UK (Horne & Reyner, 1995). Comparing drivers in accidents with a
random sample of the population, Connor et al (2002) found that the crashed drivers
were much more likely to have had less than five hours of sleep in the previous 24
Other studies that have found effects of sleep variables include driver groups such as
bus and truck drivers in Greece (Tzamalouka, Papadakaki & Chliaoutakis, 2005) and
car drivers in Spain (Masa, Rubio & Findley, 2000). However, it should be
remembered that not all studies have actually found an effect of habitual fatigue on
accidents (e.g. Morrow & Crum, 2004).
The studies using the individual's habitual fatigue level as predictor have often
worked with sleep disorder patients, mainly those with Obstructive Sleep Apnoea
Syndrome (OSAS), meaning that the effects found in these are really only applicable
for this special group. Also, the number of methodologically sound studies is low
(Connor, Whitlock, Norton & Jackson, 2001). However, the evidence to date would
seem to indicate that sleep apnoeas have at least two or three times as many accidents
as the general population (Findley, Unverzagt & Suratt, 1988; Gonzalez-Rothi,
Foresman & Block, 1988; Wu & Yan-Go, 1996; Barbé et al, 1998), and that snorers
come a good second (Maycock, 1997; for reviews, see Connor, Whitlock, Norton &
Jackson, 2001; Philip & Åkerstedt, 2006), although the evidence here is even weaker.
However, similar results have been found for occupational injuries of all types
(Ulfberg, Carter & Edling, 2000). Other sleep problems, like insomnia and narcolepsy
have not really been investigated at all (Philip & Åkerstedt, 2006).
Under the heading of fatigue can also be described the effects of time schedules on
accident risk. Salminen, Perttula and Merjama (2005) found a very small effect of the
freedom of drivers of choosing the time of their rest break (self-reported data), while
Stutts, Wilkins, Scott Osberg and Vaughn (2003) reported that drivers in sleep-related
crashes were more likely than other crashers to have multiple jobs and unusual work
schedules (night shifts etc). Similar effects for rotating shifts have been found for
nurses; the number of errors of all types doubles (Gold et al, 1992).
Finally, it can be observed that it is well established that drowsiness and the resulting
accidents are fairly strongly influenced by circadian rhythms, which means that
people tend to be sleepier and have more accidents at a few specific times as
compared to the rest of the day. Exactly when these peaks happen differ somewhat
between studies, which is probably due to differences between countries in general
habits, but also random effects of differences between people. Anyway, Horne and
Reyner (1995) found that there were three peaks in UK data; at 0200, 0600 and 1600.
Although the first two were of much larger magnitude, the last is of more interest in
the present work. It can be assumed that sometime during the afternoon, drivers will
be at an increased risk for accidents, due to a slump in their alertness.
For the general population, fairly strong effects of alcohol impairment on accident
risk have been shown (reviews by Mayhew, Donelson, Beirness & Simpson, 1986;
Hedlund, 1994), although the official claim of about 50 percent of all US highway
deaths being due to alcohol has been challenged as strongly overstated (Zylman,
1974). In short, it is very difficult to ascertain how many drunk drivers there are on
the road, if drivers in accidents were impaired at the time of the crash, and whether
the impairment was actually part of the cause of the incident. There are also other
data-gathering problems, like the decision by the police to test or not to test for
presence of alcohol in drivers in crashes. However, although the size of the problem is
in dispute, even the smallest estimates would seem to warrant some caution by the
manager, as the effects of reducing drunk driving would seem to be fairly strong
Cell phone use
The use of cell phones has been investigated for some decades, and the results so far
would seem to be rather similar; telephoning while driving is indeed a risk factor,
although not very strong. In the probably largest study to date, Laberge-Nadeau et al
(2003) ascertained a relative risk of 1.1-1.2 for cell phone users, when many other
variables had been accounted for. It can be noted that these authors used police-
recorded accidents as dependent variable, meaning that the effect could be stronger
for minor incidents (which are more numerous, but seldom find their way into police
records). One interesting feature of this study was that the number of calls (retrieved
from the phone operator companies) was predictive of the degree of extra risk; the
highest users had risks above 2. However, this can be compared to the relative risks of
age groups in this study. In fact, the heavy cell phone users had accident risks fairly
similar to 16-24 year olds.
Stronger effects were reported by Redelmeier and Tibshirani (1997); about four times
higher risk. No difference was found between hand-held and hands-free phones. Even
higher risks were indicated by Violanti and Marshall (1996) in a very small sample,
while Violanti (1997) found that drivers who owned cell phones and crashed had
some features of these crashes which were different from other drivers'; most of them
can be interpreted as lack of attention problems. Also, these drivers were at an
increased risk of being killed in their accidents.
As is common in research, estimates of effects differ, sometimes strongly, between
studies. This can be partly explained by random factors, but it should also be
remembered that the specific method used will also have an impact, because different
ways of conceptualising the problem will be used, as well as different statistical
methods. Given the research on cell phones so far, it would seem safe to say that there
is indeed an increased risk with their use while driving. How large this risk is,
however, depends somewhat upon how you phrase the question.
Given that speed is almost unanimously seen by most safety workers as the main or
part of the reason for the majority of accidents, this variable should really not be left
to the drivers' choice. Here, it will only be pointed out that the speed factor comes out
as a predictor of accidents regardless of how you measure, where, or in what
population. For example, speed has been found to be important at the level of crash
investigations, roads, and individuals, using self-reports, police and company data
(Aarts & van Schagen, 2006; af Wåhlberg, 2006). It can also be pointed out that the
increase in risk with increased speed is not linear, but rather curvilinear, i.e. the
increase in risk is larger at higher speeds.
One of the, for the general public, most unknown psychological effects within
transportation is that of behavioural adaptation. This term denotes changes in
behaviour triggered by safety features, but usually not those intended by the makers of
the feature. For example, there are more accidents on zebra crossings than in other
places, per person crossing (Herms, 1972), probably due to less caution on behalf of
the pedestrians. The general principle would seem to be some kind of risk
consistency, where people who have the experience of increased safety tend to change
their behaviour to reap some other benefit instead. However, it should be pointed out
that scientists do not agree how strong this effect is, or even if it exists (for a review,
see Grayson, 1996).
The same kind of unwanted effect has been apparent with some safety devices in cars,
notably ABS brakes. Here, results sometimes indicate a change in driving style, like
keeping shorter headways (Sagberg, Fosser & Sätermo, 1997), and the net result
seems to be a zero safety effect (Farmer, 2001). In some early studies (e.g. Evans &
Gerrish, 1996; Farmer, Lund, Trempel & Braver, 1997) there was actually an
increased risk for some crash types, which disappeared after a few years (Farmer,
2001), possibly because many drivers were initially not using the system correctly
(Williams & Wells, 1994).
One notable exception to the general finding of behavioural adaptation to safety
devices seem to be Electronic Stability Control (ESC) in cars, where rather strong
safety effects (20-60 percent) have been reported (Farmer, 2004; Lie, Tingvall, Krafft
& Kullgren, 2006). Why this is so, e.g. why behavioural adaptation has not happened,
does not seem to have been investigated. It could be suspected that the effect of the
system is less salient to drivers than for example ABS, and has therefore not been
experienced as an increase in safety, in line with the results found for airbags
(Sagberg, Fosser & Sätermo, 1997). On the other hand, it could be that an even larger
safety effect has been offset by some behaviour change.
Methods for counteracting road traffic accidents in
companies; targeting specific variables
In this section, a number of methods for safety work available to the driver manager
are described. These range from rather specific ones to general principles, and all are
not applicable in all instances. However, there is no doubt that there are always
possibilities available for increasing driver safety in companies.
The selection of drivers on attributes considered to denote safe driving has a long
history. However, equally long is the misunderstanding of the low power of this
method (as pointed out by Johnson, 1946). There are two reasons why this is so; first,
no strong correlates of accident liability have been found (see the various reviews
referenced in the second section). Second, selection only works well if there is a fairly
large pool of applicants to choose from (this is inversely related to the power of the
Therefore, the use of selection of drivers should not be expected to strongly increase
safety within a company, unless there is at least twice as many prospective drivers as
jobs, and a number of fairly complicated tests are undertaken. Here it can also be
cautioned that various selection services sold by specialized companies are usually
not worth the cost. If a company has decided to outsource such a function, the
methods used by the bidders should be surveyed by an independent specialist.
The description above may seem to indicate that transportation companies should hire
anyone and not bother with selection procedures. This is not so. What is argued is
instead that very simple procedures should be used, as they are probably cost-
effective. This includes looking at a number of common-sense variables, which have
been shown to be associated with accidents.
Age and experience have many times been shown to correlate at least weakly with
accident involvement (Cornwall, 1962; Evans & Courtney, 1985; Blom, Pokorny &
van Leeuwen, 1987; af Wåhlberg, 2005; Maycock, Lockwood & Lester, 1991;
Maycock, Lester & Lockwood, 1996). More specifically, experience has the largest
impact in the first few years of driving (Sagberg, 1998), where after the effect is
rather small. Age has a somewhat more linear effect, until rather old ages are reached.
For the manager, this can be boiled down to preferring to hire drivers with at least ten
years experience of driving such vehicles as is used by the company. If the applicant
is older than thirty, this is good.
If accident records are available from previous employers, insurance companies and
national sources, this should be used. Although the number of accidents for a driver
can differ a lot between years, in the long term (10+ years), previous accidents are
highly predictive of later ones.
Criminal record is a fairly good predictor of traffic accidents, and is of course of
general interest to an employer. As a special case of criminal acts can be mentioned
traffic violations, i.e. having been fined by the police for breaches of the traffic rules.
The power of this variable is weak, but its there.
Similar to criminal activity in its general interest to companies is drug use. Although
alco locks are available on the market, it is better not to hire a person with these
problems, as they also indicate other unwanted characteristics that make these persons
worse drivers even when they are sober.
Several of these factors are probably already used by most transport companies, more
or less explicitly. The claim here is that some of these variables, taken together,
probably have more predictive power than any specialized selection tool that is
offered by recruitment and fleet management companies.
Today, the most popular type of driver training would seem to be the skill-based sort
which stresses the technical handling of the vehicle, traffic rules and attention,
although other subjects are also covered. Unfortunately, the effects of such training
have not been shown to be positive (e.g. Downing, 1988). Also, Gregersen (1994)
found that being trained by a professional instructor in addition to private education
did not have any impact on crash risk for novice drivers.
In northern countries, so-called skid training is often a part of the standard driver
training. The driver is taught, on a special skid range, to handle a skidding car, with
the aim of improving their chances of being able to handle such occasions on the road.
However, little, if any, positive effect seems to ensue. It should be pointed out that in
the one study where an effect was found (Katila, Keskinen, Hatakka & Laapotti,
2004), the skid training used did not have an emphasis on skills in handling skids, but
was more about avoiding them.
Also, the conclusion by Lund and Williams (1985) from their review of the research
on the US Defensive Driving Course was that it had no measurable safety effect,
although violations did decrease. These results were similar to those of a review on
'driver improvement programs' in the US, where the aim of the intervention was to
change drivers' knowledge and attitudes, especially for drivers with violations
(Struckman-Johnson, Lund, Williams & Osborne, 1989).
What goes wrong in most skills training is probably that the drivers think they have
learned how to drive more safely (Gregersen, 1996), but instead of actually doing this,
they change their behaviour into more dangerous driving.
That other types of training are more beneficial than the basic skills methods have
been shown for example in Denmark, where a change of curriculum towards hazard
perception and defensive driving lead to a reduction in accidents for beginner drivers
(Carstensen, 2002). It can be noted that this was a strong study, where official
statistics and self-reports yielded the same results, and that many error factors were
During the 1990ies, so-called fuel-efficient driver training emerged in several
countries, and it was claimed that, apart from reducing fuel consumption, this type of
skill also had safety effects. The evidence to date, however, is mostly hearsay, and the
only study undertaken found no effect on accidents and a very small impact on
behaviour and fuel consumption (af Wåhlberg, in press). It should be noted that the
lack of efficiency of this type of training does not seem to be due to any error in the
principles conveyed, as very strong differences are regularly found during training (af
Wåhlberg, 2002). Instead, the problem is that the behaviour change does not seem to
transfer very well to the daily driving undertaken by those trained.
Overall, it can therefore be concluded that the traditional type of training offered by
most driving schools is useful only for very limited groups, deleterious for some, and
without effect for most. In general, it is therefore a waste of time and money.
The influence of the characteristics of the vehicle driven upon the behaviour of the
driver is hardly researched at all, while the effects upon the results of an accident have
been extensively covered. It is important to distinguish between these two cases,
which can be made with the example of a new, powerful, car filled with safety
devices, as compared to an older one without these features. The evidence would
seem to indicate that while the modern car is more protective of its occupants, given
that there is an accident, it also has the negative effect of changing the behaviour of
the driver in a riskier direction (mainly higher speeds). Whether the positive or
negative effect is stronger does not seem to have been investigated.
As most so-called 'safety' features of modern cars have unproven, and often doubtful,
effects, these should not be a primary target for safety interventions, apart from the
wearing of seatbelts.
At a very basic level, people do unsafe acts because they are reinforced in some way,
by the behaviour or its consequences. Most often, the reinforcement is immediate,
while punishment for these acts are delayed, weak and/or infrequent. As all species
learn better from immediate consequences of their actions, humans tend to prefer
short-term satisfaction, and are hard pressed to avoid long-term negative outcomes.
Smoking is a prime example of such behaviour, but so is fast driving and similar acts.
As a principle, a manager should therefore look for ways of identifying the reinforcers
of dangerous behaviours and remove them, or put in some kind of punishment
contingent upon taking risks. Much better, however, is to use some kind of positive
reinforcement schedule. The trick is to find something that rewards the driver while
driving, because schemes such as 'best driver of the year' have all the bad features of
the punishments discussed above; delayed, weak and infrequent. Few people will be
reinforced in their everyday driving by the faint possibility of being rewarded in the
future with some not very attractive prize.
Similarly, feedback of all types needs to be immediate, or people have trouble
connecting their own behaviour to what happens later. For example, most drivers are
unaware of the large influence (10-20%) their behaviour has on fuel consumption.
Although they might grumble at the fuel pump, they do not feel the fuel being wasted
in the same way as they feel the acceleration, or the time pressure of work. Therefore,
it should be possible to increase safety by giving feedback on the negative, immediate
consequences of risky behaviour.
Technical surveillance/feedback systems
One further feature of cars, which may be an important safety tool, is electronic
surveillance equipment. In this rapidly developing and expanding area, there are many
possibilities for future safety work. Today, the main aim of most surveillance systems
seem to be to monitor the technical aspects of the vehicle, i.e. the engine and
driveline, and also its physical whereabouts by GPS. This makes it possible to fine-
tune service and delivery times etc, and is also a protection against theft. So far, the
only safety perspective is that of speed, but several other aspects of driver behaviour
are possible to measure.
Given a surveillance system, it would be possible to monitor the behaviour of the
drivers and see to that they adhere to company policy, for example regarding speeding
and breaks. However, these data can also be used for feedback to drivers, and for
spotting those who are in need of it concerning safety. Today, most managers will
probably not know anything about their drivers' behaviour on the road until crash
reports start to pile up. However, there are warning signs, which can be used to
identify these drivers. Speed is one such variable, but it is not so useful in urban areas,
where speed is usually rather limited anyway, by congestion. Under such
circumstances, acceleration (af Wåhlberg, 2006), and probably fuel consumption, are
stronger predictors of accidents (af Wåhlberg, 2002).
But how and why should variables like these be used? As pointed out, most managers
probably have no idea at all about the drivers' behaviours before accidents start
happening. However, it is also the case that the drivers themselves have little
understanding of their own behaviour in terms of relating it to how others drive. This
is apparent from many studies, where most people report that they are better drivers
than the average, and also overestimate the commonness of their own dangerous
behaviours, like speeding. The rationale seem to be that 'everybody's doing it, so why
shouldn't I?' Providing feedback on variables such as speed and acceleration to the
worst drivers will show them that they are actually different from the other drivers,
and that the management has noted this. Such a method was used by Larson et al
(1980), using the technically very primitive information from tachographs. When
feedback concerning speed, speed rate and other tachograph information was
provided for police officers from their sergeant, accidents were reduced by tens of
percent. Today, when such a method could be largely automated and instantaneous,
the effects would probably be much larger.
Sleep and fatigue
As noted, people with sleep disorders have an elevated accident risk, due to their
chronical fatigue. It can therefore be useful for the manager to take some interest in
the drivers' sleep quality. This can be done in an informal way (ask about snoring,
trouble falling asleep, waking up at night, getting enough sleep etc), where those who
respond positively can take a self-report test, like the Stanford sleepiness scale 1
(Connor et al, 2002), or be sent to a doctor for further evaluation. It is important to
stress that there are several possible treatments for different sleep problems, including
cognitive behaviour therapy for the milder cases, and continuous positive airways
pressure for sleep apnoea. However, whether such treatments actually lessen the
treated drivers' accident risk has still not been decisively proven (Connor, Whitlock,
The Epworth Sleepiness Scale has yielded uneven results in relation to accidents (Philip & Åkerstedt,
Norton & Jackson, 2001), although a few researchers have reported positive results
(e.g. Cassel et al, 1996; Findley, Smith, Hooper, Dineen & Suratt, 2000).
Fatigue is a normal effect of work, even for healthy subjects without sleep problems.
This is the reason for work breaks. However, the problem is that there are large
individual differences in when a rest period is actually needed, given the same work.
Milosevic (1997) reported from four to six hours as the time when most of his sample
of dump-truck drivers reported being fatigued, while the range for all drivers was
apparently broader. The conclusion from such findings is that it is very hard to
recommend an exact cut-off for when driving becomes dangerous. It is possible that
instead of using number of hours as decision rule, emphasis should be put upon
teaching drivers to recognize their own signs of fatigue and react upon them.
Milosevic found, for example, that back pains and other physical discomfort was
commonly reported by drivers as signs of fatigue. Such tangible feelings may be more
easily recognized by the stressed driver as signals that it is time to take a break, and
should therefore be emphasised by the company.
Were possible, company guidelines could also be issued, where it is recommended
that drivers refrain from driving if they have had less sleep than usual. In general, the
company should stress the importance of good sleeping and health when it comes to
One of the conclusions of the sleep/fatigue review is that it is important for companies
where the drivers can influence the timing of their rest breaks to try to teach the
drivers to plan their schedules so that breaks are inserted at times appropriate for each
driver, with enough flexibility to be able to counter the odd night of bad sleep and
sudden dips of unknown origin. For long-haul truck drivers (who are known to sleep
too little during their drives (Milosevic, 1997), this is an appropriate way, while for
taxi drivers it would seem more useful to teach them the importance of simply
refraining from taking another customer when they feel tired.
For most drivers, it would probably be a good thing if they had the possibility of
stopping and having a nap in the afternoon or whenever their individual alertness dip
happens. Except for long-distance truck drivers, who often have a sleeper berth, this
would seem to necessitate going home for most drivers. However, there is also the
possibility of the company arranging for a rest room. See further the suggestions
about alleviating stress.
Alcohol and other drugs
Although the possibilities of controlling driving under the influence, and general
alcohol intake, is only under partial control by the manager, the emerging technology
of alco-locks make this an interesting future area of safety work. However, today it
would seem to be more fruitful to stress a generally healthy and safety-conscious
living for employees, by encouraging exercise and giving out information about the
dangers of driving under the influence, including hangover. It should also be
explicitly stated by the company that being hung over is not acceptable for drivers.
The first thing to do is to review the actual working situation of the drivers. For how
long do they actually drive without taking a break? Do they experience time pressure
and/or stress in their daily driving, trying to make it on time?
Are the drivers on commission? In that case, the company policy should be reviewed,
because this will entice drivers to shorten their time of travel. This effect is due to
their being a reward for them (extra pay) for making many contacts/sales, while the
possible punishment (an accident) will mainly be a cost to the company.
Are cars equipped so that the drivers can use them as offices? If so, the company can
issue a policy for using certain hours (say rush traffic and the afternoon attention dip)
for office work in the car, a cafeteria, or why not a small local office set up by the
company for drivers passing through?
How is the day's driving planned? Do drivers actually try to drive the shortest way
around, and avoid congested areas, or is the planning done solely from when
appointments can be made?
In this review, the main orientation has been variables that have been shown to
correlate with accident record, and are also possible for the manager to influence.
Furthermore, a number of specific and general methods for increasing driving safety
in companies have been discussed. The emphasis has also been on features that have
been well researched, with reliable methods.
In general, it can be said that most methods advocated here do not demand any special
expertise, at least in their initiation. Most of them are also fairly cheap.
It should be noted that the possible savings are surprisingly large, as indicated by
some successful safety intervention programs (Larson et al, 1980; Saari & Näsänen,
1989). We are talking about tens of percent of reduction, and often a stable low level
after the program has been withdrawn. Accidents are not unavoidable, although a zero
level may be very hard to attain. Furthermore, the knowledge is in existence. It just
need to be dug out and implemented.
Aarts, L, & van Schagen, I. (2006). Driving speed and the risk of road crashes: A
review. Accident Analysis and Prevention, 38, 215-224.
Arthur, W., Jr., Barrett, G. V., & Alexander, R. A. (1991). Prediction of vehicular
accident involvement: A meta-analysis. Human Performance, 4, 89-105.
Barbé, F., Pericas, J., Munoz, A., Findley, L., Anto, J. M., Agusti, A. G., de Lluc
Joan, M. (1998). Automobile accidents in patients with sleep apnoea syndrome.
American Journal of Respiratory Critical Care Medicine, 158, 18-22.
Blom, D. H., Pokorny, M. L., & van Leeuwen, P. (1987). The role of age and
experience in bus drivers' accidents. International Journal of Epidemiology, 16, 35-
Broughton, J., Baughan, C. J., Pearce, L., Smith, L., & Buckle, G. (2003). Work-
related road accidents. TRL Report 582. Crowthorne: Transport Research
Boufous, S., & Williamson, A. (2006). Work-related traffic crashes: A record linkage
study. Accident Analysis and Prevention, 38, 14-21.
Bylund, P-. O., Björnstig, U., & Larsson, T. J. (1997). Occupational road trauma and
permanent medical impairment. Safety Science, 26, 187-200.
Carstensen, G. (2002). The effect on accident risk of a change in driver education
Denmark. Accident Analysis and Prevention, 34, 111-121.
Cassel, W., Ploch, T., Becker, C., Dugnus, D., Peter, J. H. & von Wichert, P. (1996).
Risk of traffic accidents in patients with sleep-disordered breathing: reduction with
nasal CPAP. European Respiratory Journal, 9, 2606-2611.
Chapman, P., Roberts, K., & Underwood, G. (2000). A study of the accidents and
behaviours of company car drivers. In G. B. Grayson (Ed.), Behavioural Research in
Road Safety X. Crowthorne: Transport Research Laboratory.
Clarke, S., & Robertson, I. T. (2005). A meta-analytic review of the Big Five
personality factors and accident involvement in occupational and non-occupational
settings. Journal of Occupational and Organizational Psychology, 78, 355–376.
Connor, J., Norton, R., Ameratunga, S., Robinson, E., Civil, I., Dunn, R., Bailey, J., &
Jackson, R. (2002). Driver sleepiness and risk of serious injury to car occupants: a
population based case control study. BMJ, 324, 1125.
Connor, J., Whitlock, G., Norton, R., & Jackson, R. (2001). The role of sleepiness in
car crashes: a systematic review of epidemiological studies. Accident Analysis and
Prevention, 33, 31-41.
Cornwall, C. J. (1962). The accident experience of London bus drivers. Annuals of
Occupational Hygiene, 5, 69-82.
Donovan, D. M., Marlatt, G. A., & Salzberg, P. M. (1983). Drinking behaviour,
personality factors and high-risk driving. Journal of Studies on Alcohol, 44, 395-428.
Downing, A. (1988). The effectiveness of a retraining programme for bus drivers in
Pakistan. In J. A. Rothengatter & R. A. de Bruin (Eds.) Road User Behaviour: Theory
and Research (pp. 682-689). Maastricht: van Gorcum.
Evans, L., & Gerrish, P. H. (1996). Antilock brakes and risk of front and rear impact
in two-vehicle crashes. Accident Analysis and Prevention, 28, 315-323.
Evans, W. A., & Courtney, A. J. (1985). An analysis of accident data for franchised
public buses in Hong Kong. Accident Analysis and Prevention, 17, 355-366.
Farmer, C. M. (2001). New evidence concerning fatal crashes of passenger vehicles
before and after adding antilock braking systems. Accident Analysis and Prevention,
Farmer, C. (2004). Effect of electronic stability control on automobile crash risk.
Traffic Injury Prevention, 5, 317–325.
Farmer, C. M., Lund, A. K., Trempel, E. R., & Braver, A. R. (1997). Fatal crashes of
passenger vehicles before and after adding antilock braking systems. Accident
Analysis and Prevention, 29, 745-757.
Findley, L., Smith, C., Hooper, J., Dineen, M., & Suratt, P. M. (2000). Treatment with
nasal CPAP decreases accidents in patients with sleep apnoea. American Journal of
Respiratory and Critical Care Medicine, 161, 857-859.
Findley, L., Unverzagt, M. E., & Suratt, P. M. (1988). Automobile accidents
involving patients with obstructive sleep apnoea. American Review of Respiration
Disorder, 138, 337-340.
Gold, D. R., Rogacz, S., Bock, N., Tosteson, T. D., Baum, T. M., Speizer, F. E., &
Czeisler, C. A. (1992). Rotating shift work, sleep, and accidents related to sleepiness
in hospital nurses. American Journal of Public Health, 82, 1011–1014.
Gonzales-Rothi, R., Foresman, G., & Block, A. (1988). Do patients with sleep apnoea
die in their sleep? Chest, 94, 531-538.
Goodenough, D. R. (1976). A review of individual differences in field dependence as
a factor in auto safety. Human Factors, 8, 53-62.
Grayson, G. B. (1996). Behavioural adaptation: a review of the literature. TRL
Report 254. Crowthorne: Transport Research Laboratory.
Gregersen, N-. P. (1994). Systematic cooperation between driving schools and parents
in driver education, an experiment. Accident Analysis and Prevention, 26, 453-461.
Gregersen, N-. P. (1996). Young drivers’ overestimation of their own skill - An
experiment on the relation between training strategy and skill. Accident Analysis and
Prevention, 28, 243-250.
Hedlund, J. H. (1994). "If they didn't drink, would they crash anyway?": The role of
alcohol in traffic crashes. Alcohol, Drugs and Driving, 10, 115-125.
Herms, B. F. (1972). Pedestrian crosswalk study: accidents in painted and unpainted
crosswalks. Highway Research Record, 406, 1-13.
Horne, J. A., & Reyner, L. A. (1995). Sleep related vehicle accidents. British Medical
Journal, 310, 565-567.
Johnson, H. M. (1946). The detection and treatment of accident-prone drivers.
Psychological Bulletin, 43, 489-532.
Jonah, B. A. (1997). Sensation seeking and risky driving: A review and synthesis of
the literature. Accident Analysis and Prevention, 29, 651-665.
Katila, A., Keskinen, E., Hatakka, M., & Laapotti, S. (2004). Does increased
confidence among novice drivers imply a decrease in safety? The effects of skid
training on slippery road accidents. Accident Analysis and Prevention, 36, 543-550.
Laberge-Nadeau, C., Maag, U., Bellavance, F., Lapierre, S. D., Desjardins, D.,
Messier, S., & Saidi, A. (2003). Wireless telephones and the risk of road crashes.
Accident Analysis and Prevention, 35, 649-660.
Larson, L. D., Schnelle, J. F., Kirchner, Jr., R., Carr, A. F., Domash, M., & Risley, T.
R. (1980). Reduction of police vehicles accidents through mechanically aided
supervision. Journal of Applied Behaviour Analysis, 13, 571-581.
Lester, J. (1991). Individual differences in accident liability: review of the literature.
TRRL Research Report 306. Crowthorne: Transport and Road Research Laboratory.
Lie, A., Tingvall, C., Krafft, M., & Kullgren, A. (2006). The effectiveness of
electronic stability control (ESC) in reducing real life crashes and injuries. Traffic
Injury Prevention, 7, 38-43.
Lund, A. K., & Williams, A. F. (1985). A review of the literature evaluating the
defensive driving course. Accident Analysis and Prevention, 17, 449-460.
Masa, J. F., Rubio, M., & Findley, L. (2000). Habitually sleepy drivers have a high
frequency of automobile crashes associated with respiratory disorders during sleep.
American Journal of Respiratory and Critical Care Medicine, 162, 1407-1412.
Maycock, G. (1997). Sleepiness and driving: the experience of heavy goods vehicle
drivers in the UK. Journal of Sleep Research, 6, 238-244.
Maycock, G., Lester, J., & Lockwood, C. R. (1996). The accident liability of car
drivers: the reliability of self report data. TRL Report 219. Crowthorne: Transport
Maycock, J., Lockwood, C., & Lester, J. F. (1991). The accident liability of car
drivers. TRRL Research Report No. 315. Crowthorne: Transport and Road Research
Mayhew, D. R., Donelson, A. C., Beirness, D. J., & Simpson, H. M. (1986). Youth,
alcohol and relative risk of accident involvement. Accident Analysis and Prevention,
McCartt, A., Ribner, S., Pack, A., & Hammer, M. (1996). The scope and nature of the
drowsy driving problem in New York State. Accident Analysis and Prevention, 28,
Meuleners, L. B., Lee, A. H., Legge, M., & Cercarelli, L. R. (2005). Health conditions
of heavy vehicle drivers involved in a crash in Western Australia: a retrospective
study using linked data. Health Promotion Journal of Australia, 16, 37-40.
Miller, T. R., & Galbraith, M. (1995). Estimating the costs of occupational injury in
the United States. Accident Analysis and Prevention, 27, 741-747.
Milosevic, S. (1997). Drivers' fatigue studies. Ergonomics, 40, 381-389.
Mitchell, R., Driscoll, T., & Healey, S. (2004). Work-related road fatalities in
Australia. Accident Analysis and Prevention, 36, 851-860.
Morrow, P. C., & Crum, M. R. (2004). Antecedents of fatigue, close calls, and crashes
among commercial motor-vehicle drivers. Journal of Safety Research, 35, 59-69.
Peck, R. C. (1993). The identification of multiple accident correlates in high risk
drivers with specific emphasis on the role of age, experience and prior traffic violation
frequency. Alcohol, Drugs and Driving, 9, 145-166.
Petersen, A., Barrett, R., & Morrison, S. (in press). Driver-training and emergency
brake performance in cars with antilock baking systems. Safety Science,
Philip, P., & Åkerstedt, T. (2006). Transport and industrial safety, how are they
affected by sleepiness and sleep restriction? Sleep Medicine Reviews, 10, 347-356.
Redelmeier, D. A. & Tibshirani, R. J. (1997). Association between cellular-telephone
calls and motor vehicle collisions. New England Journal of Medicine, 336, 453-458.
Saari, J., & Näsänen, M. (1989). The effect of positive feedback on industrial
housekeeping and accidents: a long-term study at a shipyard. International Journal of
Industrial Ergonomics, 4, 201–211.
Sagberg, F. (2006). Driver health and crash involvement: A case-control study.
Accident Analysis and Prevention, 38, 28-34.
Sagberg, F., Fosser, S., & Sätermo, I.-A. (1997). An investigation of behavioural
adaptation to airbags and antilock brakes among taxi drivers. Accident Analysis and
Prevention, 29, 293-302.
Salminen, S., Perttula, P., & Merjama, J. (2005). Use of rest breaks and accidents by
professional drivers. Perceptual and Motor Skills, 101, 665-668.
Shinar, D. (1993). Traffic safety and individual differences in drivers' attention and
information processing capacity. Alcohol, Drugs and Driving, 9, 219-237.
Struckman-Johnson, D. L., Lund, A. K., Williams, A. F., & Osborne, D. W. (1989).
Comparative effects of driver improvement programmes on crashes and violations.
Accident Analysis and Prevention, 21, 203-215.
Stutts, J. C., Wilkins, J. W., Scott Osberg, J, & Vaughn, B. V. (2003). Driver risk
factors for sleep-related crashes. Analysis and Prevention, 35, 321-331.
Tse, J. L., Flin, R., & Mearns, K. (2006). Bus driver well-being review: 50 years of
research. Transportation Research Part F, 9, 89-114.
Tzamalouka, G., Papadakaki, M., & Chliaoutakis, J. El. (2005). Freight transport and
non-driving work duties as predictors of falling asleep at the wheel in urban areas of
Crete. Journal of Safety Research, 36, 75-84.
Ulfberg, J., Carter, N., & Edling, C. (2000). Sleep-disordered breathing and
occupational accidents. Scandinavian Journal of Work Environment Health, 26, 237–
Violanti, J. M. (1997). Cellular phones and traffic accidents. Public Health, 111, 423-
Violanti, J. M., & Marshall, J. R. (1996). Cellular phones and traffic accidents: An
epidemiological approach. Accident Analysis and Prevention, 28, 265-270.
Williams, A. F., & Wells, J.-A. (1994). Driver experience with antilock brake
systems. Accident Analysis and Prevention, 26, 807-811.
Wu, H., & Yan-Go, F. (1996). Self-reported automobile accidents involving patients
with obstructive sleep apnoea. Neurology, 46, 1254-1257.
af Wåhlberg, A. E. (2002). Fuel efficient driving training - state of the art and
quantification of effects. Manama: E141 Proceedings of Soric'02. Available at
af Wåhlberg, A. E. (2003). Some methodological deficiencies in studies on traffic
accident predictors. Accident Analysis and Prevention, 35, 473-486.
af Wåhlberg, A. E. (2005). Differential accident involvement of bus drivers. In L.
Dorn (Ed.) Driver Behaviour and Training, Volume II, pp. 383-391. Aldershot:
Ashgate. Second International Conference on Driver Behaviour and Training.
Edinburgh 15-17 November, 2005.
af Wåhlberg, A. E. (2006). Speed choice versus celeration behaviour as traffic
accident predictor. Journal of Safety Research, 37, 43-51.
af Wåhlberg, A. E. (in press). Long term effects of training in economical driving;
fuel consumption, accidents, driver acceleration behavior and technical feedback.
International Journal of Industrial Ergonomics,
Zylman, R. (1974). A critical evaluation of the literature on "alcohol involvement" in
highway deaths. Accident Analysis and Prevention, 6, 163-204.
Information processing: A cognitive psychological research area, which studies how
humans handle information from their senses. For example, can people
simultaneously process information from hearing and vision?
Meta analysis: A quantitative review of several research results, and also a set of
statistical techniques for this. For example, two dozen studies may have found varying
effects of sleep problems on accident liability. A meta-analyses would take all these
results and analyse them, sometimes taking into consideration samples size, sampling
method, general methodology etc, and arriving at a figure which contain less of the
error which every single study inevitably suffers from.
(Obstructive) Sleep Apnoea Syndrome (SAS): Sleep-disordered breathing, which
leads to not enough oxygen being provided for the body. This semi-suffocation makes
the subject's sleeping very uneven, as he (this afflicts (middle-aged, obese) men more
often than women) partially wakes up many times during sleep. However, this partial
awakening does not necessarily register in memory, and sleep apnoeas may be quite
unaware of their problem, apart from general tiredness etc. A spouse may be much
more knowledgeable, because snoring and stopped breathing are common symptoms.