Skip to main content
European Commission logo
Mobility & Transport - Road Safety

Fatigue and crash risk



Fatigue and crash risk

The finding that fatigue is involved in 10-25% crashes does not in itself prove that fatigue increases crash risk. For example, it could be that drivers who are more fatigued also drive more kilometres than other drivers so that the risk per kilometre is the same for fatigued and non-fatigued drivers. Several studies have investigated the relationship between driver fatigue and crash risk and have attempted to quantify the increased risk. Often increased risk of particular groups such as young drivers or professional drivers derives from a combination of factors.


How dangerous is fatigued driving?

Several studies have investigated the relationship between driver fatigue and crash risk and have attempted to quantify the risk increase. Reviewing these studies, Connor et al [21] concluded that nearly all studies were limited in their ability to establish a causal relationship. Study limitations concerned design, biases, and in many cases, small sample sizes. Despite these limitations the better quality cross-sectional studies do suggest a positive relationship between fatigue and crash risk. A reliable estimate of the strength of the association cannot yet be given.

In a case control study of New Zealand drivers, Connor et al [20] compared 571 crash-involved drivers with 588 non-crash involved drivers driving in the same area and at the same times. Driver variables were taken from accident registration and additional interviews. Taking into account possible confounding variables (gender, age, socio-economic status, annual kilometres, speed, road type), they found a strong relationship between acute fatigue (based on loss of sleep the night before) and crash involvement. Crash risk was eight times higher for drivers with a score ≥ 4 on the Stanford Sleepiness Scale (95% confidence interval 3.4-19.7); 5,5 times higher for driving between 2 and 5 am (95% interval 1.4-22.7); and almost 3 times higher when drivers had slept for less than 5 hours in the past 24-hour period (95% confidence interval 1.4-5.4).

In a case-control study, Cummings et al [23] compared crash-involved drivers with a similar group of non-crash involved drivers at the same location, direction, time and day. They found the crash risk was fourteen times higher for drivers who had reported to have almost fallen asleep behind the wheel (95% confidence interval 1.4-147).

The data collected in The 100 Car Naturalistic Driving Study shows that driving while fatigued increases a driver's risk of involvement in a crash or near-crash by nearly four times [59].

Studies of professional drivers (bus, lorry, truck) show that it takes around 9 or 10 hours of driving, or 11 hours of work, before crash risk starts to rise [68]. Hamelin found that after 11 hours of work span the crash risk doubles. The effect of task duration is practically always entangled with the effects of the time of day and sometimes also with the length of time awake and previous lack of sleep. The duration of a trip may be of lesser importance compared to these other factors - many fatigue-related accidents occur after driving for only a few hours [98][102]. Short trips can also end up in fatigue-related crashes because time of day and long and irregular working hours are stronger predictors of fatigue than time spent driving [15][114].

Connor et al. [21]also note blind spots in the research on driver fatigue. The association of non-medical (lifestyle) determinants of fatigue with crash has not been the subject of thorough research. There is still a lack of knowledge concerning the contribution of increasing total hours of work, and shift schedules to driver fatigue. Whereas research into fatigue and sleep apnoea in truck drivers has led to awareness of these problems and some modification of work conditions [34][73][77], occupationally induced fatigue in potentially much larger numbers of commuters has received little attention.


Combination of factors

Frequently a combination of situational and individual factors contributes to increased risk of being involved in a fatigue-related crash. The increased risk may results from a mix of biological, lifestyle, and work-related factors. For teenage drivers, the strong biological need for sleep and going out in weekend-nights may combine to increase fatigue and risk [38]. For professional drivers and long distance drivers, both reduced sleep and long working hours combine to increase fatigue and risk [53][76][91].

Stutts et al [102] investigated both situational factors and individual differences in fatigue-related traffic risk. The database consisted of police accident reports and surveys from 312 drivers who fell asleep at the wheel, and surveys from 155 drivers who had caused an accident as a result of fatigue. The study used as a control group 529 drivers, who were responsible for an accident, which was not caused by fatigue and 407 accident-free drivers.

The researchers found that drivers responsible for a fatigue related accident had more often several jobs, were shift workers or had unusual working hours. In addition, these drivers reported to sleep less hours at the average, to feel more tired during the day, to drive more often at night-time and to have experienced drowsiness at the wheel more frequently. In comparison to drivers with accidents without fatigue origin, these drivers drove on average longer, were awake more time, slept less at night and used barbiturates more often. 23 % of the drivers with fatigue related accidents reported to have driven in the past year 10 or more times in a fatigued state. 19% of these drivers reported to have been awake more then 20 hours before the crash. The authors conclude, that the crash risk due to fatigue is significantly increasing, if the driver sleeps less than 7 hours. Compared to driver averaging 8 hours of sleep or more, drivers who sleep less than 5 hours per night on average are 6 times as likely to be involved in a fatigue related crash (versus not being in a crash at all).