Factors associated with falling asleep at the wheel among long-distance truck drivers

https://doi.org/10.1016/S0001-4575(99)00067-6Get rights and content

Abstract

Data on the prevalence and hypothesized predictors of falling asleep while driving were gathered through face-to-face interviews with 593 long-distance truck drivers randomly selected at public and private rest areas and routine roadside truck safety inspections. Hypothesized predictor variables related to drivers’ typical work and rest patterns, extent of daytime and night-time drowsiness, symptoms of sleep disorder, measures of driving exposure, and demographic characteristics. A sizeable proportion of long-distance truck drivers reported falling asleep at the wheel of the truck: 47.1% of the survey respondents had ever fallen asleep at the wheel of a truck, and 25.4% had fallen asleep at the wheel in the past year. Factor analysis reduced the large set of predictors to six underlying, independent factors: greater daytime sleepiness; more arduous schedules, with more hours of work and fewer hours off-duty; older, more experienced drivers; shorter, poorer sleep on road; symptoms of sleep disorder; and greater tendency to night-time drowsy driving. Based on multivariate logistic regression, all six factors were predictive of self-reported falling asleep at the wheel. Falling asleep was also associated with not having been alerted by driving over shoulder rumble strips. The results suggest that countermeasures that limit drivers’ work hours and enable drivers to get adequate rest and that identify drivers with sleep disorders are appropriate methods to reduce sleepiness-related driving by truck drivers.

Introduction

Until recently in the US, driving while drowsy has not been a major focus of highway safety initiatives aimed at the general public1. This may be attributable to studies of crash data that suggest that sleepiness-related driving is a relatively low crash risk, when compared to such other factors as alcohol-impaired driving or excessive vehicle speed. According to research conducted by the US National Highway Traffic Safety Administration, drowsiness or fatigue has been identified as a causal factor in 1.2–1.6% of all police-reported crashes and 3.6% of fatal crashes (Knipling and Wang, 1994, Knipling and Wang, 1995). It is believed, however, that drowsiness is under-reported as a contributory crash factor, because there generally is little physical or other evidence that a crash-involved driver was drowsy or fell asleep, and a lack of standardization among the states in the definition and reporting of sleepiness-related crashes.

An increased interest in the role played by driver drowsiness in highway crashes is evidenced by the recent publication of reports on sleepiness and highway crashes by a national Expert Panel on Driver Fatigue and Sleepiness (1997) and by the Council on Scientific Affairs, American Medical Association, 1998. Both reports emphasize that certain groups of drivers, including commercial drivers, are at increased risk for sleepiness-related crashes. In this regard, government agencies, transportation professionals, researchers, the motor carrier industry, safety advocates, and others have long been concerned with the extent and causes of sleepiness-related driving among long-distance truck drivers. Research on the causes and effects of sleepiness suggests that the schedules of long-distance truck drivers may place them at higher risk for driving while drowsy and for sleepiness-related crashes. Driving is the type of task known to be fatiguing; it is both a monotonous, repetitive task and one that requires sustained attention (Monk and Folkard, 1979, Williamson et al., 1996). Not only do most truck drivers have high levels of driving exposure, many also work long, irregular hours, at times that conflict with natural circadian rhythms; they also frequently drive alone, on long, monotonous, high-speed roadways. Many truck drivers may have limited opportunities to obtain sufficient sustained, restorative sleep and thus may accumulate a sizeable sleep debt. Many may also have difficulty maintaining an otherwise healthy lifestyle, including a nutritious diet and regular exercise.

The federal hours-of-service (HOS) regulations are a primary tool used by the US Federal Highway Administration (FHWA) to promote safe travel by interstate motor carriers. Title 49, Part 392.3, of the Code of Federal Regulations stipulates that no truck driver may operate a motor vehicle while the driver’s ability or alertness is impaired, or likely to become impaired, through fatigue. Title 49, Part 395 stipulates that except under limited circumstances, interstate truck drivers may not drive more than 10 consecutive h or work more than 15 h before having 8 h off-duty; a driver may split the 8 h off-duty into two rest periods if the truck has a sleeper berth. In addition, total weekly work time is limited to 60 h in 7 consecutive days or to 70 h in 8 consecutive days, depending on whether the carrier operates on a 6-day or 7-day weekly schedule. Drivers may continue to perform non-driving work-related duties past the 60/70 h limit.

In initiating rule-making in 1996 to revise the HOS regulations, the Federal Highway Administration intends to replace the current regulatory approach with one that couples regulations with a performance-based system that assesses driver alertness and fitness for duty. As outlined in the Federal Register, November 5, 1996, a primary consideration in the regulatory process is the extant research on the scope, causes, and effects of drowsiness among commercial vehicle drivers; the crash risks associated with different levels of drowsiness and with various scheduling, work, and compensation practices; the effectiveness of current regulations in reducing drowsiness and mitigating its effects; and the estimated effects on highway safety of proposed alternative approaches.

While research on truck driver fatigue or sleepiness extends back to the 1930s, much remains unknown about the extent, causes, and effects of sleepiness-related driving among commercial drivers. Case studies of crashes conducted by the National Transportation Safety Board (NTSB, 1990, NTSB, 1995) are among the most frequently cited studies of the extent of sleepiness-related crashes among truck drivers. A 1995 NTSB study of single-vehicle large-truck crashes, in which the drivers survived, found that 58% of the crashes were fatigue-related and that 19 of the 107 drivers (17%) interviewed stated that they had fallen asleep while driving (NTSB, 1995). In a self-report study of truck drivers in an Australian state without hours-of-service regulations, 5% of drivers reported having a hazardous fatigue-related event, such as nodding off, on their current trip; 14% said they had nodded off at least occasionally while at the wheel over the previous 9 months (Arnold et al., 1997). Of those drivers who had a crash in the previous 9 months, 12% cited fatigue as a contributory factor.

Most studies have examined the extent and causes of sleepiness in drivers, rather than the link between sleepiness and crash risk or between causes of sleepiness and crash risk. Synthesizing these studies is difficult, given the variety of research designs, differing outcome and predictor measures, differing sampling frames and methods, and sometimes inconsistent findings. These studies suggest, however, that sleepiness-related driving among long-distance truck drivers is associated with variables that are related to drivers’ schedules, particularly their patterns of working and resting, and drivers’ individual sleep needs and patterns.

First, the number and pattern of hours worked and hours off-duty have been linked to sleepiness-related driving. Studies of the general driving population and shift workers have linked sleepiness-related driving with rotating shifts and night shifts (Mitler et al., 1988, Gold et al., 1992, Marcus and Loughlin, 1996, McCartt et al., 1996, Lauber and Kayten, 1998). Researchers, using operational data from a national motor carrier, reported that total driving time had a greater effect on crash risk among truck drivers than either time of day or driving experience (Lin et al., 1994). An experimental study by Mackie and Miller (1978) found that driving performance among truck drivers declined with an irregular schedule, more than 8 h of driving for regular schedules, and more than 5 h of driving for irregular schedules. In another experimental study, Harris et al. (1972) found that driving performance, as measured by number of crashes, declined after 4 h of driving. Williamson et al. (1996) found that fatigue increased over the duration of trips, regardless of the driving regime, although the pre-trip level of fatigue was a primary factor in fatigue experienced while on the road. In a 1992 survey of over-the-road tractor-trailer drivers, violators of the hours-of-service regulation were more likely to report that they had fallen asleep at the wheel (Braver et al., 1992). Hertz (1988) identified an elevated fatal crash risk when drivers split the required 8 h off-duty into two sessions in a sleeper berth. Truck drivers’ schedules may necessitate sleeping during the daytime, and daytime sleep may not achieve the restorative quality of night-time sleep (Lavie, 1986).

A second variable, the time of day, has also been identified as predictive of sleepiness-related driving among truck drivers. Based on physiologic and performance data for 80 US and Canadian long-haul truck drivers on revenue-generating trips, the US/Canadian Driver Fatigue and Alertness Study found that the strongest and most consistent factor influencing driver fatigue and alertness was time of day, rather than time on task or cumulative number of trips (Wylie et al., 1996). It might be noted that drivers participating in this study followed driving schedules that, although demanding, were permissible within current US and Canadian HOS regulations; the study did not examine the effects of schedules that violate regulations. Mackie and Miller (1978) and Harris et al. (1972) also found an association between time of day and level of fatigue, while Hertz (1988) and Jovanis et al. (1991) found night-time driving resulted in a higher crash risk for truck drivers.

Substantial research has also demonstrated a link between sleepiness-related driving and the quantity and quality of sleep. In general, a person’s tendency to fall asleep during normal waking hours is increased and psychomotor performance declines with fewer hours of sleep and successive days of restricted sleep (Wilkinson et al., 1966, Carskadon and Dement, 1981, Mitler et al., 1997). The 1995 NTSB study found that predictors of sleepiness-related single-vehicle large-truck crashes were the duration of a driver’s last sleep period, the total sleep obtained during the 24 h preceding the crash, and fragmented sleep patterns. The study also suggested that night driving after relatively little sleep was a better predictor of fatigue-related crashes than night driving alone. In the US/Canadian Driver Fatigue and Alertness Study (Mitler et al., 1997), researchers concluded that drivers obtained less sleep on the road than is required for alertness on the job. Although drivers vary with regard to the portion of off-duty time devoted to sleep, longer off-duty periods have been associated with longer periods of sleep (Mitler et al., 1997).

A fourth variable linked to sleepiness-related driving is obstructive sleep apnea, characterized by severely disturbed breathing during sleep. Symptoms of this disorder include obesity, snoring, sleep interrupted by intermittent gasping for breath, and excessive daytime sleepiness. Research suggests that drivers with untreated sleep apnea, snoring, or sleep-disordered breathing are at increased risk for motor vehicle crashes (Findley et al., 1988, Aldrich, 1989, Findley et al., 1989, Stoohs et al., 1993, Young et al., 1997) and that truck drivers may be at increased risk for obstructive sleep apnea (Stoohs et al., 1993, Stoohs et al., 1994).

Finally, based on studies of the general driving population, the extent of daytime sleepiness has been related to sleepiness-related driving. Based on drivers’ self-report, McCartt et al. (1996) found that the frequency of drowsy driving was related to the frequency of trouble staying awake during the day and the number of hours that could be driven before the onset of drowsiness. Also using self-report data, Maycock (1997) found an association between the probability of almost falling asleep at the wheel and the Epworth Sleepiness Scale score, a measure of general daytime sleepiness (Johns, 1991).

In sum, previous studies, using a variety of methods and measures, have identified a number of variables associated with fatigue-related driving and decrements in driving performance. However, few of these studies have used multivariate analyses to establish the relative contribution of various predictors of sleepiness-related driving. Especially given the high costs and practical difficulties of conducting in situ over-the-road studies, and the considerable limitations of crash data, research methods relying on drivers’ self-reports are particularly useful to gather in-depth data on drivers’ experiences with sleepiness-related driving under typical circumstances and on risk factors related to drivers’ typical work and rest patterns. In the study reported here, data were collected through a survey of long-distance truck drivers to determine the prevalence of sleepiness-related driving among these drivers and to identify associated risk factors and their relative predictive importance. These data represent one of the most comprehensive data sets related to drivers’ typical work and rest patterns and their symptoms of sleep disorders, patterns of sleepiness, and experiences with sleepiness-related driving.

Preliminary bivariate analyses of the survey data indicated that falling asleep at the wheel in the past year was associated with a number of variables related to job characteristics and drivers’ work/rest schedules (McCartt et al., 1997, McCartt et al., 1998). However, the large set of hypothesized predictors and the significant bivariate associations between many predictors suggested that a multivariate analytical approach would provide a deeper understanding of the key driver-related and work-related constructs predictive of sleepiness-related driving. Thus, this paper reports the results of a two-stage multivariate analysis to accomplish the following: (1) condense the large set of hypothesized driver-related and work-related predictors of falling asleep at the wheel into a smaller, more useful set of constructs; and (2) measure the relative importance of these general constructs in predicting sleepiness-related driving.

Section snippets

Survey method

In-depth interviews were conducted in Spring 1997 with a sample of long-distance truck drivers selected to be representative of long-distance truck drivers traveling New York’s interstate roadways. A total of 593 drivers were interviewed. Interviews were conducted with 192 drivers (32.4%) at public full-service and limited-service rest areas, with 233 drivers (39.3%) at private full-service truck stops, and with 168 drivers (28.3%) at routine truck safety inspections conducted by state and

Respondent characteristics

Almost all drivers interviewed were men (98.8%). Less than one-quarter of drivers (22.3%) were licensed by New York; 61.2% were licensed by another US state, and 16.5% were licensed by a Canadian province. Drivers most often worked for a private fleet (37.7%), a company that owns trucks and employs drivers, or for a for-hire fleet (35.3%), a company that leases trucks with drivers to other companies. Of the remaining drivers who described themselves as owner-operators, 22.3% worked under a

Discussion

While prior studies of fatigue and sleepiness-related driving among truck drivers have largely focused on specific sleepiness-related driving episodes and the variables linked with those episodes, this study probed drivers with respect to self-reported prior incidents of falling asleep at the wheel and their usual ‘real-world’ work, sleep, and rest patterns. The study departs from other studies of sleepiness-related driving in terms of the wide range and large number of hypothesized predictors

Acknowledgements

This study was funded, in part, by the New York State Department of Transportation with research funds from the Motor Carrier Safety Assistance Program, Office of Motor Carriers, Federal Highway Administration.

References (40)

  • L.J Findley et al.

    Severity of sleep apnea and automobile crashes

    New England Journal Medicine

    (1989)
  • L.J Findley et al.

    Automobile accidents involving patients with obstructive sleep apnea

    American Review of Respiratory Disease

    (1988)
  • D.R Gold

    Rotating shift work, sleep, and accidents related to sleepiness in hospital nurses

    American Journal of Public Health

    (1992)
  • R.L Gorsuch

    Factor Analysis

    (1974)
  • Harris, W., et al., 1972. A Study of the Relationships among Fatigue, Hours of Service, and Safety of Operations of...
  • L Hartley

    Recommendations from the Second International Conference on Fatigue in Transportation, Fremantle, Western Australia

    (1996)
  • R.P Hertz

    Tractor-trailer driver fatality: the role of nonconnective rest in a sleeper berth

    Accident Analysis and Prevention

    (1988)
  • M.W Johns

    A new method for measuring daytime sleepiness: the Epworth sleepiness scale

    Sleep

    (1991)
  • Jovanis, P.P., Kaneko, T., Lin, T.D., 1991. Exploratory analysis of motor carrier accident risk and daily driving...
  • Knipling, R.R., Wang, J.-S., 1994. Crashes and Fatalities Related to Driver Drowsiness/Fatigue, Research Note, Office...
  • Cited by (0)

    View full text