4. SPECIAL TOPICS

“Estimated proportions of fatal crashes attributable to driver fatigue vary substantially.”

Driver fatigue has been identified as a leading contributor to roadway crashes among workers as well as the general population. Fatigue affects driving performance by impairing information processing, attention, and reaction times; it may also cause a driver to fall asleep. Time of day, duration of wakefulness, inadequate sleep, sleep disorders, and prolonged work hours have all been identified as major causes of fatigue [Akerstedt 2000].

Estimated proportions of fatal crashes attributable to driver fatigue vary substantially and may be higher among CMVs than in the general population. Driver drowsiness or fatigue was implicated in 1,773 fatal crashes (3.1%) in the United States in 2000 [NHTSA 2001a]. However, fatigue was noted for 7.4% of drivers of large trucks involved in fatal, single-vehicle crashes [FMCSA 2002a]. Fatigue was implicated in only 1.0% of large-truck drivers involved in fatal, multiple-vehicle crashes.

The Federal Highway Administration reported that driver fatigue contributes to an estimated 15% to 33% of crashes that were fatal only to occupants of large trucks. They note in contrast that only 1% to 2% of large-truck crashes fatal to pedestrians or occupants of other vehicles are judged to be fatigue-related [GAO 1999]. More recently, the FMCSA estimated that fatigue is involved in 15% of all fatal large-truck-related crashes. The agency estimated that fatigue is directly involved in 4.5% of these crashes, and that the mental lapses and inattention associated with fatigue contribute an additional 10.5%. Fatigue-related fatal crashes involving CMVs are most common in long-haul trucking, which accounts for 480 (63.6%) of the estimated 755 such events that occur annually [65 Fed. Reg. 25540 (2000)].

4.1 DRIVER FATIGUE

4.1.2 Factors Affecting the Estimated Role of Driver Fatigue

Several factors contribute to the wide range of estimates on the proportion of crashes attributable to fatigue:

• A lack of agreement about the definition of fatigue
  
  
• The subjectivity involved in making a determination about fatigue at the scene of a crash
  
  
• A frequent lack of witnesses or physical evidence at the crash scene (these fatigue-related crashes are disproportionately single-vehicle incidents)

Investigators may presumptively attribute a crash to fatigue based on the time of the incident, the driver’s work and sleep schedules over the previous few days, and evidence at the crash scene that the driver did not attempt to avoid the crash.

4.1.3 Research on Driver Fatigue

“In general, crash risk has been shown to increase with hours of driving.”

Most of the research on driver fatigue among those who drive on the job has focused on the motor carrier industry [Braver et al. 1992; Feyer et al. 1997; Hertz 1988; Jones and Stein 1987; Kaneko and Jovanis 1992; Lin et al. 1993, 1994; McCartt et al. 2000; NTSB 1995, 1999; Williamson et al. 1996; Wylie et al. 1996]. Numerous studies have addressed the associations of time of day and hours of driving with crash risk for large trucks, drawing varying conclusions about the relative contributions of each. In general, crash risk has been shown to increase with hours of driving [Jones and Stein 1987; Kaneko and Jovanis 1992; Lin et al. 1993, 1994]. Other research has demonstrated that truck drivers as a group do not necessarily obtain adequate rest under the current hours-of-service regulations [Brown 1994; McCartt et al. 2000; Wylie et al. 1996]. A number of studies have concluded that night driving is associated with increased crash risk—particularly for single-vehicle crashes, which are likely to be fatigue-related [Blower and Campbell 1998; Campbell 2002; Hamelin 1987; Hertz 1988; Kaneko and Jovanis 1992; Lin et al. 1993]. In addition, factors such as regularity of schedule, night rest, and taking rest breaks have all been cited as being associated with reduced crash risk [Hamelin 1987; Kaneko and Jovanis 1992; Lin et al. 1993, 1994].

Two related studies of the trucking industry concluded that number of hours driven had the strongest direct effect on crash risk. Crash likelihood increased steadily after 4 hours of driving, with risks in the 9th and 10th hours that were 80% and 130% higher, respectively, than those for the first 4 hours [Lin et al. 1993, 1994]. The finding that crash risk decreased in the 6th and 7th hours among drivers who had taken a rest break between hours 2 and 6 confirms the importance of breaks in reducing crash risk [Lin et al. 1994]. A case-control study of large trucks matched by roadway, time of day, and day of week found that drivers who were on the road for more than 8 hours had 1.8 times the crash risk of those who had driven 2 hours or fewer [Jones and Stein 1987]. Another study that compared drivers by weekly driving pattern and crash involvement reported significant increases in crash risk after 4 hours of driving, with the sharpest increase after 9 hours [Kaneko and Jovanis 1992]. An analysis of TIFA data for 1981–1996 reported nearly identical results [Campbell 2002]. Work practices such as driving for more than 10 consecutive hours, taking fewer than 8 hours off duty, and driving greater numbers of hours over a 7-day period were highly predictive of falling asleep at the wheel in a group of 593 long-distance truck drivers [McCartt et al. 2000].

“Night driving coupled with fatigue has also been associated with increased crash risk”

Night driving coupled with fatigue has also been associated with increased crash risk, higher proportions of injury-producing crashes, and greater likelihood of fatality in a crash. Between 1992 and 2000, work-related fatalities among semi-truck occupants were lowest at 9 p.m. (n=57), increasing steadily and peaking at 5 a.m. (n=172) (see Figure 4). In an analysis of TIFA data for 1981–1996 [Campbell 2002], the distribution of the risk of fatigue involvement, given a fatal large-truck crash, mirrors the pattern of early-morning fatalities in the CFOI data displayed in Figure 4. Another study reported that driving between 8 p.m. and 7 a.m. was associated with 1.9 times the crash risk during daytime hours. Driving 11 hours or more increased the relative risk associated with night driving to 2.4 [Hamelin 1987]. In two other studies, most of the driving patterns identified as posing the highest crash risk involved substantial periods of driving between midnight and 10 a.m. [Kaneko and Jovanis 1992; Lin et al. 1993]. In other research, all but one of the four 2-hour periods associated with the highest crash risk (independent of other variables) were during evening or early morning hours [Lin et al. 1994]. Lowest risk was associated with frequent driving between 6 a.m. and 2 p.m. and following a regular driving schedule. This group of low-risk drivers also tended to have off-duty time between 6 p.m. and 4 a.m., which increased the opportunity to obtain nighttime rest.

Other research provides further evidence of a link between night driving and the risk of large-truck crashes, reporting that long-haul trucks had more than twice the risk of crashes for the hours between 9 p.m. and 6 a.m. compared with daytime hours [Blower and Campbell 1998]. The hours between midnight and 6 a.m. were associated with the highest rate of injuries per 1,000 crashes (435 versus 320 for other hours of the day). Injuries during these hours were also more severe, with nearly twice as many fatalities per 1,000 injuries. Citing FARS data from 1993–1995, the authors noted an excess of single-vehicle crashes among long-haul truck crashes between midnight and 6 a.m. (21% at night versus 15% at other times of the day). For single-vehicle crashes involving long-haul trucks, fatigue was cited as a FARS driver-related factor for the truck driver in 19% to 21% of the crashes occurring between midnight and 6 a.m.—compared with fewer than 10% of the crashes that occurred throughout the day. This study estimated that 531 excess deaths occur per year among truck drivers and other road users as a result of night trucking operations.

Campbell [2002] used TIFA data to examine fatal fatigue-related crashes by (1) type of motor carrier employer (for-hire carriers whose primary business is transporting goods belonging to someone else versus private carriers who operate trucks to move their own goods), (2) type of truck (single-unit “straight” trucks versus semi-trucks), and (3) the length of trip one way (local versus 50 to 200 miles versus more than 200 miles). The highest relative risks of a fatal fatigue-related crash per vehicle mile traveled were found among drivers of straight trucks employed by for-hire carriers (3.41) and by private carriers (2.47). However, the number of fatal fatigue-related crashes in this population was quite small. More than half of fatal large-truck crashes in which fatigue was implicated involved for-hire semi-trucks on one-way trips of more than 200 miles. Compared with all other large-truck crashes, this group of semi-trucks had 1.9 times the risk of fatigue involvement per vehicle mile traveled. For all lengths of trips, drivers employed by for-hire carriers had a substantially higher risk of a fatigue-related fatal crash than did drivers employed by private carriers.

Insufficient sleep during off-duty time before and during a trip as well as chronic, accumulated fatigue are recognized crash risk factors that fall outside the scope of hours-of-service regulations in the United States and other nations [Brown 1994; Feyer et al. 1997; Hartley 1997; NTSB 1995]. A field study that evaluated four different driving schedules found that drivers got inadequate rest regardless of schedule, averaging 3.8 to 5.4 hours of sleep daily. Cumulative fatigue (as indicated by decreased scores on a test of vigilance) was observed during the last days of a trip, and driver self-assessment suggested higher fatigue levels after multiple trips [Wylie et al. 1996]. In another field study, fatigue increased markedly for all three driving regimens tested over the course of a 900-kilometer trip. Pre-trip sleep for the average driver in this Australian study was only slightly more than 6 hours [Williamson et al. 1996].

4.1.4 Regulations Addressing Driver Fatigue

“The U.S. regulations addressing driver fatigue and driving time have changed twice since their establishment in 1937.”

The U.S. regulations addressing driver fatigue and driving time have changed twice since their establishment in 1937 (see Section 3.1.5 for a more detailed discussion of existing and revised regulations). A rule change in 1962 reduced the mandatory off-duty period from 9 to 8 hours and required on-duty periods to be separated by 8 hours off duty. In addition, the 1962 rule change retained a limit of 10 hours of driving during a duty period, but it dropped a provision that the driving occur over a 24-hour period. Revisions to the hours-of-service regulations in 2003 increased the permissible hours of driving for property-carrying CMV drivers from 10 hours to 11 hours, but they decreased the maximum on-duty hours from 15 to 14 [68 Fed. Reg. 22456 (2003)]. These revisions take effect on January 4, 2004.

Under U.S. regulations applicable to property-carrying CMV drivers through 2003, two extreme work patterns are possible, each with characteristics that may contribute to driver fatigue. One pattern compresses the workday into an 18-hour drive-rest cycle of 10 hours of driving and 8 hours off duty. This pattern disrupts the normal 24-hour sleep-wake cycle. Alternatively, a driver may legally drive for 10 hours and spend an additional 5 hours on duty within a single shift, a practice that exceeds the usual maximum work shift in most other industries. In fact, only 10% of all U.S. workers worked more than a 12-hour shift during 1997 [NIOSH 2000].

The revised regulations make it possible for property-carrying CMV drivers to approximate a 24-hour work-rest cycle if they work the maximum number of duty hours (14 hours on duty and 10 hours off duty). However, the new regulations still allow a compressed drive-rest cycle of 21 hours (11 hours driving and 10 hours off duty) that disrupts the normal 24-hour sleep-wake cycle. The safety consequences of increasing maximum driving time to 11 hours remain to be seen, as research generally shows that crash risk increases with hours of driving.

Unlike U.S. regulations, European Union regulations require breaks within a driving trip. They allow a maximum of 9 hours of driving daily (up to 10 hours twice during a week), with a mandatory 45-minute break after 41/2 hours of driving. A minimum of 11 consecutive hours off duty is required daily; this number may be reduced to 9 hours three times per week [European Economic Community 1985]. In addition, European Union regulations require that drivers be compensated for compulsory rest time, and drivers must rest 45 consecutive hours for every 6 days worked [FHWA 2000]. Although European regulations permit fewer hours of driving per day and require mandatory breaks, shortcomings have been noted: European regulations do not address the issue of adequate sleep before a trip, and they do not consider the time of day that driving may begin and end [Brown 1994].

A recent European Union directive addresses the ancillary duties that may not be accounted for when calculating hours of service for truck drivers. This directive offers a broad definition of “working time” that encompasses driving, loading and unloading, vehicle cleaning, maintenance, and waiting time. Under this directive, average weekly working time for drivers may not exceed 48 hours per week over a 4-month period, and drivers may not work more than 6 consecutive hours without a break (using the above definition of “working time”) [European Union 2002].

In Australia, regulations that apply to the more populous eastern States generally allow 11 or 12 hours of driving per day. In contrast, there are no hours-of-service regulations in remote, sparsely populated Western Australia. Instead, truckers and motor carriers in Western Australia are expected to operate under a Code of Practice for fatigue management. The Code of Practice allows for long work hours (an average of 14 hours per day over 12 days), but it permits drivers to decide when they need to rest [Transport Western Australia 1998]. Although the Code of Practice addresses issues relevant to the U.S. transport industry, its effects on trucker safety have not yet been evaluated.

The FMCSA and Transport Canada are working cooperatively to develop and test another nonregulatory approach, the Fatigue Management Program. Educational components of the Fatigue Management Program that are designed to enhance drivers’ trip planning will address wellness, lifestyle, sleep hygiene, and alertness. Other interventions include screening for sleep disorders and development of guidelines for dispatchers to improve scheduling practices. Two pilot tests have been conducted in Canada, with a similar test to be conducted in the United States in the near future. If the Fatigue Management Program is shown to be effective, it will be offered for implementation in both nations [FMCSA 2002b].

4.1.5 Noncompliance with Hours-of-Service Regulations

Research in the United States and Australia has reported rates of noncompliance with hours-of-service regulations ranging from 38% to 73% [Braver et al. 1992; Hartley 1997; Hertz 1991]. Economic pressures may encourage violations. In the United States, truckers’ pay may be based on miles driven or value of cargo, which may increase incentive to violate the regulations. This method of compensation is prohibited in some European countries, including Sweden and France [FHWA 2000]. CMV drivers covered by Federal Motor Carrier Safety Regulations are exempt from the overtime provisions of the FLSA, and they are therefore not guaranteed pay for hours worked beyond 40 hours per week. In the absence of opportunity for overtime earnings, paying truck drivers by the mile can increase their incentive to drive additional hours, thereby increasing the likelihood of fatigue.

Compensation and the potential for fatigue are also linked with respect to nondriving tasks. During on-duty, nondriving time, truck drivers perform other tasks such as cargo loading and unloading, maintenance, and recordkeeping. These nondriving tasks, particularly those involving physical labor, also contribute to fatigue. Although truck drivers may be paid an hourly or mileage rate for time spent driving, they may not be paid for the time they spend performing nondriving tasks. A recent survey of truck drivers found that 45% were paid for time spent loading and unloading, and 21% reported being paid for maintenance tasks [Belman et al. 1999]. Both these statistics have implications for safety. Lack of compensation may reduce the incentive for drivers to perform thorough safety inspections. In addition, it may increase the incentive to work too quickly during loading and unloading, thereby increasing the risk of injury from overexertion and fatigue.

An infrastructure issue that has been cited as a contributor to driver fatigue in both the United States and Australia is the shortage of suitable parking areas for drivers of large trucks to stop and rest [Braver et al. 1992; 64 Fed. Reg. 28237 (1999); FHWA 1999; Hartley 1997]. Demand for parking may be so great that truck stop owners may set time limits that make it difficult for drivers to get adequate rest. Participants at a Federal Highway Administration forum on the topic suggested that the use of alternative parking areas such as weigh stations and freight terminals be considered as a short-term solution [FHWA 1999].

Individual driver characteristics and work habits may contribute to driver fatigue and risk of a fatigue-related crash. These individual factors include age, individual differences in ability to adjust to irregular or extended work hours or night work, physical health and fitness, driving experience, undiagnosed sleep disorders, drug or alcohol consumption, and electing to drive for longer periods without taking a break [Brown 1995; NHTSA 1998a]. The importance of individual differences is underscored by research showing that small numbers of drivers contributed disproportionately to the total number of drowsiness episodes noted [Hartley 1997; Wylie et al. 1996]. However, these findings do not mean that the experience of fatigue is related only to individual driver characteristics. The unique working conditions of the trucking industry call for creative efforts to modify work schedules and ancillary on-duty task assignments so that drivers’ opportunities for rest are more similar to those of workers in other professions. Such an approach may help motor carriers retain valued employees, ensure employment for qualified drivers, and enhance roadway safety.

4.1.6 Federal Research to Address Driver Fatigue

In response to these concerns, Federal agencies have sponsored or conducted the following research to provide a scientific basis for revising hours-of-service regulations in the motor carrier industry:

• A 1995 National Transportation Safety Board study reported that the three most important predictors of fatigue involvement in large truck crashes were the duration of sleep in the last sleep period before the crash, the total hours of sleep obtained in the 24 hours before the crash, and the presence of split sleep periods. Major recommendations from this study were to complete rulemaking within 2 years to (1) revise 49 CFR 395.1 to require sufficient rest provisions for drivers to obtain at least 8 hours of continuous sleep after driving for 10 hours or being on duty for 15 hours and (2) eliminate the portion of 49 CFR 395.1 that allows drivers with sleeper berths to split the 8 hours of required off-duty time into 2 separate periods [NTSB 1995].
  
  
• A 1996 Federal Highway Administration field study of commercial driver fatigue assessed 80 drivers who followed driving regimens that differed by length of time on duty (10 versus 13 hours), type of work shift (regular hours versus rotating hours), and start time (day versus night). Drowsiness was assessed through performance tests, physiologic tests such as electroencephalographs, and video recordings. The authors concluded that time of day (i.e., driving at night) was the most important predictor of fatigue—not number of hours driven or cumulative number of trips made. Highest levels of drowsiness were noted in the 8-hour period between late evening and dawn. However, it should be emphasized that drivers in this study got very little sleep on average: 3.8 to 5.4 hours per day, depending on their driving schedule. In addition, performance tests and self-assessments done as part of the study showed some evidence of cumulative fatigue. The authors concluded that insufficient opportunity for sleep and failure of drivers to place high priority on sleep were the key contributors to lack of sleep [Wylie et al. 1996].
  
  
• In 1999, the National Transportation Safety Board published a review of progress by DOT in addressing the Board’s 1989 recommendations regarding operator fatigue. The Board reported that the DOT had made progress in implementing a coordinated research program on fatigue and had developed and disseminated educational materials about shift work to the transportation industry. However, the Board noted that little progress had been made in revising hours-of-service regulations to reflect the most current research findings on sleep and fatigue. This report restated the two 1995 recommendations for changes to 49 CFR 395.1 described above. It also recommended that the revised hours-of-service regulations should limit hours of service, provide for predictable work and rest schedules, and consider circadian rhythms and normal requirements for sleep and rest [NTSB 1999].
  
  
• In 2000, the FMCSA published results of a study of the effects of sleep on CMV driver performance. The first portion collected data on sleep patterns of long- and short-haul drivers over 20 days under normal work conditions. Both groups of drivers averaged about 7.5 hours of sleep per night. Short-haul drivers were more likely to have a single sleep period, while long-haul drivers obtained almost half their daily sleep during work hours, mostly in sleeper berths. The authors noted that drivers in both groups frequently had inadequate sleep during off-duty time [Balkin et al. 2000].

The laboratory portion of the study assessed performance on physiological, psychomotor, and driving simulation tests for 66 drivers assigned to 3, 5, 7, or 9 hours in bed each night over a 14-day period. For all but the group assigned to spend 9 hours in bed, test performance declined further with each successive day, even among the group that was assigned to spend 7 hours in bed. For the group assigned to only 3 hours in bed each night, performance did not return to baseline levels even after 3 consecutive nights with 8 hours in bed [Balkin et al. 2000].

4.1.7 Preventing Fatigue-Related Crashes

Federal regulations address fatigue in the motor carrier industry by specifying maximum hours of driving time and duty time and minimum hours of off-duty time. Yet, drivers in other industries have limited protections from work schedules that can lead to fatigue. Few studies have addressed driver fatigue among workers who operate company-owned or personal motor vehicles other than large trucks when performing their jobs. These driving environments are largely unregulated, especially those in which personal vehicles are used for work purposes. Thus, for employers not covered by the motor carrier regulations whose workers are expected to drive on the job, a driver fatigue management program is a critical element of the overall safety program.

4.2 SPECIAL ISSUES IN MOTOR CARRIER SAFETY

4.2.1 Other Safety Concerns

The highest priority of the FMCSA is to reduce the number of fatalities from crashes of large trucks by at least 50% from the 1998 baseline (5,374 deaths among truck drivers and other road users) by the end of 2009. Although the issue of driver fatigue is central to safety concerns about the motor carrier industry, numerous other factors related to the driver, the vehicle, the road, and the environment influence injury and fatality risk for truckers and other motorists. In its Safety Action Plan for the years 2000 through 2003, the FMCSA cited challenges to progress in improving safety in the motor carrier industry. These included the rapid growth in large-truck mileage and in the number of motor carriers, the need for resources to improve compliance reviews and the Commercial Driver’s License Program, the need for additional research on causal factors in large-truck crashes, slow progress in rulemaking, and the need for further evaluation of collision avoidance technologies [FMCSA 2000]. In addition, characteristics of the broader roadway work environment (such as those discussed in the following subsections) affect the safety of long-haul truck drivers and others with whom they share the road.

4.2.2 Vehicle Safety and Design Issues

Vehicle safety standards for large trucks appear, in some instances, to be less protective than similar standards for passenger vehicles. For example, trucks with a GVWR greater than 10,000 lb are excluded from standards addressing protection of the head from interior impact, protection against impact from the steering control system in a crash, requirements for roof crush resistance over the passenger compartment, and requirements for head restraints to reduce frequency and severity of injuries that may occur in rear-end and other collisions [49 CFR 571, Standard Nos. 201, 202, 203, and 216].

“The adoption of safety standards for trucks has lagged behind the adoption of similar standards for passenger vehicles”

In addition, the adoption of safety standards for large trucks has lagged behind the adoption of similar standards for passenger vehicles. For example, specifications to minimize the likelihood of inadvertent operation of powered windows, roof panels, and partitions went into effect in 1971 for passenger vehicles, but not until 1988 for trucks [49 CFR 571, Standard No. 118]. Another example is Standard No. 121, which applies to air brake systems. Originally implemented in 1975, this standard was struck down by a court decision in 1978. The standard’s key performance requirements were not reinstated until 1997 [Krall 2002].

In some instances, safety protections for truck drivers have improved in the absence of regulatory change. The provisions of Standard No. 208 (Occupant Protection) that require installation of lap belts in heavy trucks went into effect in 1972. In response to research findings that three-point shoulder-belt restraint systems reduce the incidence of head and upper body injuries, truck manufacturers have voluntarily equipped trucks with these restraint systems since the late 1980s [Krall 2002].

Requirements for application of retroreflective sheeting or reflectors on large trailers and semi-trailers (incorporated into 49 CFR 393 in recent years) have yielded substantial safety benefits. Overall, the use of these materials on trailers reduced by 29% the incidence of other motorists’ vehicles striking the rear or sides of semi-trailers under dark conditions. Retroflective sheeting was most effective in reducing injury-producing crashes (1) under dark conditions in which no additional lighting was present, (2) when the driver of the impacting vehicle was under age 50, and (3) when the material was used on a flatbed trailer [NHTSA 2001b].

4.2.3 Vehicle Maintenance

“Research has linked inadequate maintenance to increased crash risk for large trucks.”

Research has linked inadequate maintenance to increased crash risk for large trucks. An analysis of data from FARS and GES estimated that 4.5% to 5.0% of all CMV crashes have a mechanical component [Randhawa et al. 1998]. This study identified brakes, securing of loads, tires, and wheels or rims as the most common mechanical contributors. However, these results may underestimate the actual proportion of crash-involved trucks with mechanical defects. FARS and GES data are based on reports from law enforcement officers, who are generally not trained to identify vehicle defects. Furthermore, the amount of information about mechanical defects collected on police crash reports varies from State to State [Blower 2002; Randhawa et al. 1998].

Other studies using case-control design or in-depth investigative methods have reported considerably higher proportions of mechanical defects among crash-involved large trucks. A case-control study reported that large trucks with equipment defects were 1.7 times more likely to be involved in a crash than trucks with no defects [Jones and Stein 1989]. However, this study of crash-involved tractor-trailers matched with noncrash-involved controls from the same traffic stream found high proportions of equipment defects in both groups: 76% for the crash-involved trucks and 66% for the controls. The defects were serious enough in 41% of the crash-involved trucks and in 31% of the controls that these trucks should have been removed from service. A more recent study of crashes in Michigan found that nearly 55% of trucks involved in fatal crashes had at least one defect related to the truck’s mechanical condition [Blower 2002]. This study reported that 28.5% of trucks had at least one out-of-service condition—that is, a defect serious enough to require that the truck be parked until the defect was corrected.

Brake defects were found in 56% of the crash-involved trucks in the study by Jones and Stein [1989], and steering equipment defects were found in 21%. Furthermore, the authors were able to demonstrate a relationship between the type of defect and the crash configuration: 50% of trucks that rear-ended other vehicles had out-of-service brake defects, and 10% of trucks that sideswiped other vehicles had steering defects. A related publication noted that 23% of the controls with equipment defects had been on the road for less than 2 hours, suggesting that these defects were likely to have been present at the beginning of the trip and should have been identified in a pre-trip inspection [Jones and Stein 1987].

Compared with Jones and Stein [1989], Blower [2002] found lower proportions of brake violations (34.5%) and steering defects (5.6%). Like Jones and Stein, Blower found that the crash configuration and type of defect were related. Brake defects were associated with incidents in which the truck was the striking vehicle in a rear-end collision: 27.3% of trucks that were struck from the rear had a brake violation, compared with 50.0% of trucks that were the striking vehicle [Blower 2002]. This study also examined fatal crashes in which a vehicle crossed the center line and struck another vehicle. The trucks that crossed the center line were more likely to have brake defects (46.7%) than trucks that were struck by another vehicle that crossed the center line (19.7%). Similarly, trucks that crossed the center line were much more likely to have steering defects (26.7%) than trucks that were struck by another vehicle that crossed the center line (2.8%).

Defective lights and signals (found in 23.7% of large trucks involved in fatal crashes) were the second most common type of vehicle defect reported by Blower [2002]. These violations were more common in crashes in which trucks were struck from the rear (40.0%) than in crashes in which the truck was the striking vehicle (15.4%), suggesting that approaching vehicles were unable to see these trucks [Blower 2002].

4.2.4 Crash Location, Highway Design, and Emergency Response

For large trucks, crash risk on interstate highways is relatively low compared with U.S. and State highways. One study of FARS data estimated that interstate highways accounted for 40% of truck miles driven in 1996 but only 23% of truck-related fatal crashes [Stuster 1999]. This difference may be due partly to the more consistent speeds on interstate highways and to the relatively small number of access points that provide advance warning of entry and exit options. In contrast, U.S. and State highways may have numerous access points to allow local traffic to enter and exit, and these may not be as well marked as they are on interstate highways. Where there is a high volume of CMVs such as large trucks using the same road, the risk of CMV crashes with passenger vehicles increases. This traffic situation on U.S. and State highways, with their greater speed differentials, more frequent braking and accelerating, and greater number of access points, requires greater vigilance on the part of CMV operators and greater appreciation of CMV operating capabilities by passenger vehicle drivers. Limiting the number of highway entrance and exit ramps is a component of Dutch transportation policy. In some areas of the Netherlands, “truck-only” lanes are being designed and tested to reduce conflicts between trucks and passenger vehicles [FHWA 2000].

Large-truck crashes occur disproportionately in rural areas, a fact that is consistent with the high proportion of these crashes that occur on interstate, U.S., and State highways (see Table 4). The rural location of many large truck crashes has implications for response by emergency medical services as well. In 2000, 16% of large-truck crashes were single-vehicle events [NHTSA 2001a]. Emergency medical services personnel may have to travel a greater distance to reach the scene of a rural crash: they were able to respond within 10 minutes of notification for 56% of fatal rural crashes in 2000, compared with 89% of urban crashes [NHTSA 2001a]. Extrication of injured truck occupants may be complex and prolonged in rollover and jackknife events in which loads may have shifted, doors may be jammed, and the cab area may be deformed [Baker et al. 1976]. When a passenger vehicle strikes the rear of a large truck, extrication of the passenger vehicle occupant may be complicated if the vehicle has underridden the truck.

4.2.5 Research Addressing Large-Truck Safety Issues

The FMCSA is conducting or sponsoring a variety of research projects that may ultimately make large trucks inherently safer and may help motorists and truckers drive near one another more safely. The Intelligent Vehicle Initiative features collaboration with vehicle manufacturers to develop electronic brakes, on-board sensing of safety-critical systems, devices to improve truck stability, collision warning devices, and hazard location technologies [FMCSA 2000]. Other research under the Intelligent Vehicle Initiative includes a study to identify human factors and possible countermeasures for CMV rear-end collision avoidance and lane changing, and an instrumented vehicle study of car-truck interaction to describe the actions of other vehicles around trucks. These two studies could contribute information essential for developing training programs for truckers and public education programs directed toward motorists. In addition, the information collected about driving behaviors of other motorists around large trucks may help refine collision warning systems.

The FMCSA, in cooperation with NHTSA, initiated the multiyear Large Truck Crash Causation Study in 2002. This study uses sites already used for collection of Crashworthiness Data System and GES data to investigate at least 1,000 large-truck crashes that result in fatality or serious injury. Teams made up of Crashworthiness Data System researchers and State truck inspectors will collect detailed data on the crash, vehicles, and occupants. Assessment of these data will help determine the critical event that made a collision unavoidable and the reasons for the critical event [Craft and Blower 2002].

4.2.6 Future plans

Future rulemaking by the FMCSA is slated to address training requirements for entry-level drivers and unique training needs of multiple-trailer combination vehicle drivers. The agency is also planning to study the relationship between driver payment methods and safety [FMCSA 2000].

4.3 DRIVER DISTRACTION AND CELL PHONE USE

Driver distraction has been defined as “capture of the driver’s attention by information that is irrelevant to the driving situation to a degree where insufficient information is left for the primary task” [Janssen 2000]. Some distractions affect the driver by requiring physical maneuvers that may threaten vehicle control, whereas others are mental distractions from sources inside or outside of the vehicle. Among all the elements of driver distraction, cell phone use has perhaps received the most attention. In recent years, cell phone ownership has increased rapidly in the United States, with more than 137 million cell phone subscriptions as of August 2002 [Cellular Telecommunications & Internet Association 2002]. Cell phone use while driving has been questioned because it may contribute to increased risk of motor vehicle crashes. Despite this safety concern, the availability of a cell phone in a vehicle offers a number of benefits, including prevention of unnecessary trips, peace of mind through improved access to family and friends, the ability to report emergencies, improved response time to accidents, and the ability to handle household errands during commuting time [Brookhuis et al. 1991; Lissy et al. 2000]. For workers, the availability of a cell phone may offer increased productivity, efficiency, and access to clients and coworkers [Lissy et al. 2000].

4.3.1 Data on Crashes Involving Cell Phone Use While Driving

“Cell phone use while driving has been questioned because it may contribute to increased risk of motor vehicle crashes.”

National estimates from NHTSA observational studies conducted during 2000 indicate that at any given time during daylight hours, 3% of passenger vehicle drivers in the United States were actively using a hand-held cell phone. An additional 0.9% of drivers were estimated to be using a hands-free phone [Utter 2001]. However, no estimates exist for the number of persons who use cell phones while driving for work. Although improper use of cell phones and other devices has been documented as contributing to roadway crashes in the general population, determining the role of cell phones in work-related crashes is difficult. CFOI, the primary source of data on occupational fatalities, does not collect this information. FARS collects information about the presence of driver-related factors such as inattention, drowsiness, and cell phone use, but it is less comprehensive than CFOI in its coverage of occupational roadway crashes. Death certificates are the only means FARS uses to ascertain work relationship; and although death certificates are the single source shown to identify the greatest number of work-related deaths, at least 20% may not be captured [Stout and Bell 1991].

The fundamental problem is that currently only 15 States are required by law to collect information about cell phone involvement on police crash reports [Rushing 2002]. Therefore, although FARS and the National Automotive Sampling System added cell phone use as a driver-related factor in 1995, the police crash reports that are a primary source of data for these systems do not necessarily collect this information [NHTSA 1997]. Even for States that do collect it, the true extent of the involvement of driver distraction and cell phones may be underestimated: as with fatigue-related crashes, assessment of these elements is largely subjective. And police crash reports are not designed to provide a scientific assessment of crash causation.

Although DOT data systems provide a mechanism to collect national data on roadway crashes associated with cell phone use, published data indicate that very small numbers are actually reported through FARS—a total of 76 cell-phone-related fatal crashes in 1994 and 1995 [Lissy et al. 2000]. Examination of more recent on-line FARS data revealed that 81 fatal crashes in 1999 and 101 fatal crashes in 2000 were reported to be related to cell phone use—0.2% of the total for those years. One of the crashes in 1999 and two of the crashes in 2000 were identified as being a fatal injury at work [NHTSA 2002b]. Other electronic devices (e.g., computers, fax machines, and on-board navigation systems) were identified as contributing to a total of five fatal crashes in 1999 and 2000, one of which was a fatal injury at work. It is not yet known whether the small number of fatal crashes related to use of cell phones or other electronic devices reflects under-reporting or whether these devices indeed pose little risk [NHTSA 1997].

Data from the Crashworthiness Data System indicate that in 1995, small proportions of nonfatal towaway crashes in the United States were related to cell phone use. Of nearly 2.4 million crashes in 1995, talking on a cell phone was associated with 0.1%, and dialing a cell phone was associated with an additional 0.1% [Wang et al. 1996]. Unlike FARS, the Crashworthiness Data System does not collect information about work relationship. However, this system does record a detailed description of the vehicle and information about gross vehicle weight, both of which may be of some value in assessing whether a crash is occupational in nature.

Lissy et al. [2000] suggest that fatal cell phone-related crashes may appear infrequently in data systems that focus on fatal crashes and towaway crashes because they are likely to occur during rush hours, when congestion may lead to lower-speed collisions and hence fewer crashes resulting in fatalities and injuries. In addition, cell phone use may be most common during daytime hours, when overall crash risk is lower.

4.3.2 Research on the Risks of Cell Phone Use While Driving

Two case-control studies found a statistical but not necessarily causal relationship between cell phone use and increased risk of motor vehicle crashes [Violanti and Marshall 1996; Violanti 1998]. The methods used in the first of these studies did not allow researchers to establish whether a cell phone was in use at the time of the collision. A study by Redelmeier and Tibshirani [1997] is generally considered to be the most scientifically sound study of the relationship between cell phone use and crash risk. They examined drivers involved in property-damage-only crashes, comparing their cell phone use on the day of a crash with their cell phone use on the day before the crash during the same time period. This research found that drivers who had used a cell phone in the 10 minutes before a crash had 4.3 times the crash risk of those who had not. This study, like several others, was limited by its inability to establish that cell phone use caused the crash.

“The use of cell phones while driving has a negative effect on driving performance.”

In general, studies conducted in driving simulators and in instrumented vehicles have found that the use of cell phones while driving has a negative effect on driving performance [Brookhuis et al. 1991; McKnight and McKnight 1993; Nilsson and Alm 1991; Tijerina et al. 1995]. In these studies, the decreased performance associated with cell phone use was measured by decreased response time for manual or cognitive tasks, decreased time looking at the road, increased workload (as indicated by increased heart rate and steering wheel movements), less mirror checking, longer braking times, failure to maintain safe following distances or consistent speed, and poor lane-keeping.

Results are mixed from research addressing whether hands-free phones with voice-activated dialing pose less risk than hand-held phones that require manual dialing. In general, research suggests that hands-free phones are safer than hand-held models, although they still pose distractions for drivers [Stevens and Paulo 1997]. One study conducted with instrumented vehicles found that drivers using hands-free cell phones performed better than those using hand-held phones but less well than those who did not use cell phones at all [Brookhuis et al. 1991]. A study conducted in a driving simulator reported better driving performance for voice-activated dialing versus manual dialing [Serafin et al. 1993]. In contrast, another simulator study found equivalent declines in driving performance among users of hand-held and hands-free cell phones [Strayer and Johnston 2001].

“Studies outside the laboratory, like simulator-based studies, have failed to establish that hands-free phones offer clear safety advantages.”

Studies outside the laboratory, like simulator-based studies, have failed to establish that hands-free phones offer clear safety advantages. A study of professional semi-truck drivers found that lane-keeping was improved and distraction from visual driving tasks was reduced during voice-activated versus manual dialing [Tijerina et al. 1995]. In contrast, two other studies did not report lowered crash risk for users of hands-free phones versus hand-held phones [Dreyer et al. 1999; Redelmeier and Tibshirani 1997]. However, it is important to note that the two types of studies are not directly comparable. The first study [Tijerina et al. 1995] measured differences in performance that may be associated with crash risk but was not designed to follow drivers’ crash experience. The other two studies reported differences in crash risks among large groups of cell phone users [Dreyer et al. 1999; Redelmeier and Tibshirani 1997].

Also unclear is whether crash risk is greater while dialing, conducting a conversation, or reaching to answer a phone or retrieve a dropped phone. Studies done in the United States have generally found that conversation, not dialing, is involved in greater numbers of cell-phone-related crashes. A multiyear review of North Carolina police crash reports found that reaching for a dropped cell phone and talking on a cell phone were the primary circumstances associated with crashes involving cell phones [NHTSA 1997]. A review of FARS and GES cases found that talking on a cell phone was implicated in 17 of 28 crashes involving cell phones (61%). Reaching for a phone was identified in two cases, and dialing was identified in only one case [NHTSA 1997]. Another study, conducted in a simulator, assessed differences in driving performance among users of hand-held cell phones given simple versus more complex conversational tasks. The researchers reported that more driving errors were committed while performing a more complex word-generation task than during a simple word repetition task [Strayer and Johnston 2001]. These findings have implications for the safe use of cell phones in the workplace. Although little is known about the content of work-related cell phone calls made from vehicles, it is reasonable to assume that some of these calls place substantial demands on driver attention and may increase crash risk.

4.3.3 In-Vehicle Internet and Other Information Systems

New technologies have the potential for further eroding driver attention, especially when they are coupled with cell phone use while driving. Although in-vehicle Internet technology would ideally provide information that has positive effects on traffic and fleet management [Burns and Lansdown 2000], others have cautioned that little is known about the effects of multiple information systems on driver attentiveness [NHTSA 1997]. Trends toward miniaturization of electronic devices might also compromise safety by placing greater visual demands on the driver [NHTSA 1997].

“Little is known about the effects of multiple information systems on driver attentiveness.”

A recent study conducted on a test track assessed the effects of multiple devices on driver performance [NHTSA 2000a]. Drivers were asked to manually tune a car radio, manually dial an unfamiliar number on a cell phone, and enter information into a route guidance system. The route guidance systems tested had various types of user interfaces: manual keypad entry, joystick, and voice activation. Overall, drivers needed much more time to interact with the route guidance system than to dial the cell phone or tune the car radio.

Compared with drivers aged 35 or younger, drivers aged 55 or older took more than twice as long to enter information into the route guidance system, took their eyes off the road about twice as often, and had much more difficulty staying in the travel lane while entering information. For all drivers, using the voice-activated route guidance system was associated with far less time with eyes off the road and no declines in lane-keeping. Among older drivers using the voice-activated system, times with eyes off the road decreased almost to that observed among younger drivers. The voice-activated system did not necessarily require less time than interacting with the manual entry systems. However, the eyes-off-the-road time associated with its use was much lower—nearly as low as that observed for dialing a cell phone or tuning a radio. Thus voice-activated route guidance systems appear to be preferable to manual entry systems, although research is needed to further assess the effects of voice interaction on driver attentiveness [NHTSA 2000a].

The potential for task and sensory overload as a result of more and smaller devices is of particular concern in the workplace, where installation of Internet-based information systems may offer increased productivity through streamlined customer contact and scheduling. Given the potential economic benefits, it is possible that introduction of this technology into fleet vehicles may precede its widespread use in personal vehicles. Thus the first evidence of any safety effects of in-vehicle Internet combined with other technologies may be seen among workers.

The European Union has developed principles for in-vehicle Internet systems that address design, installation, presentation of information, and drivers’ interactions with displays and controls [Commission of the European Communities 2000]. The European Union principles stress the importance of maintaining the driver’s attention to the driving task, view of the road, and view of critical vehicle displays and controls. In addition, they emphasize that in-vehicle Internet systems should not (1) place the driver under pressure to respond in a certain time frame, (2) require long, uninterruptible sequences of interactions, or (3) visually entertain the driver. Others have recommended that in-vehicle Internet systems be capable of automatically restricting information when traffic conditions demand it [Burns and Lansdown 2000].

4.3.4 Policy and Legislative Issues Related to Cell Phone Use While Driving

Policy decisions to guide cell phone use while driving are unusually difficult given that cell phones offer clear safety benefits along with risks. In addition, the rapid growth in cell phone ownership has not been accompanied by regulations governing their use, either in the general population or in the workplace. Between 1995 and June 2002, at least 41 States considered legislation addressing the use of cell phones and other in-vehicle electronic devices in passenger vehicles [Rushing 2002]. To date, no State has instituted a complete ban on cell phone use while driving. However, the State of New York passed legislation (effective in December 2001) that prohibits motorists from using hand-held cell phones while driving on public roadways. A number of localities have passed similar ordinances prohibiting the use of hand-held devices [Rushing 2002]. An additional concern is that varying restrictions on cell phone use by locality has the potential to create confusion among residents and among those who are in an unfamiliar location on business travel and unaware of local laws. As of June 2002, five States had passed legislation that would make State law override any local ordinances related to cell phones. However, to date only New York has acted to place significant restrictions on the use of cell phones statewide [Rushing 2002].

Of concern to employers is the possibility of increased legal liability of workers (and perhaps employers) who are involved in cell-phone-related crashes. One source cited recommendations that employers consider placing limits on worker cell phone use and convey to workers that cell phone use is not a necessary condition of employment [Buschman 2000]. To date, no research studies have addressed cell phone use on the job, including whether pressure on workers to use cell phones for conducting business has a detrimental effect on vehicle safety.

4.4 AGE-RELATED FACTORS

4.4.1 Young Drivers

Workers aged 16 and 17 have lower fatality rates attributable to work-related roadway crashes than do workers aged 18 and 19 (which may reflect lower levels of exposure to driving because of FLSA prohibitions). However, these workers are still exposed to crash risks as vehicle passengers and as pedestrians. Fatality rates among workers aged 18 and 19 (those not restricted under FLSA) are comparable with those for adult workers aged 20 to 44, and rates for workers aged 18 are higher than those for workers aged 35 through 44 (see Tables 7 and 8).

“Novice drivers may acquire vehicle handling skills quickly, but they require much more time to develop skills needed to recognize hazards and respond appropriately.”

Several factors contribute to the increased crash susceptibility of young drivers. Numerous studies indicate that young novice drivers may acquire vehicle handling skills quickly, but they require much more time to develop higher-order perceptual and cognitive skills needed to recognize hazards and respond appropriately [Deery 1999; Pelz and Krupat 1974; Regan et al. 1998b]. Novice drivers may also lack skills for determining what factors in the driving environment require their attention at a given time, adjusting to differences in intensity of the driving workload and matching their performance to demands of the task [Deery 1999]. Immaturity is another important factor. Young drivers typically possess less well developed judgment, engaging in more risky driving behaviors than their more experienced adult counterparts [Deery 1999]. Also related to immaturity is the tendency for young drivers to overestimate their own driving skills [Gregersen 1996]. Another contributor to injury risk is low levels of safety belt use among adolescents and young adults compared with older persons [NHTSA 1998b]. Sixty-one percent of vehicle occupants aged 16 to 20 who died in automobile or truck crashes in 2000 were not wearing a safety belt [NHTSA 2001a].

Fatigue may also contribute to crash risk for young workers who drive on the job. Lifestyle factors can result in insufficient sleep for adolescents at a time when maturational changes can make them more susceptible to the effects of fatigue [NHTSA 1998a]. Employed youth, who must balance school, home, and social life with job responsibilities, typically get less sleep than their counterparts who are not employed, and they are more likely to report daytime sleepiness [Carskadon 1990].

Graduated driver licensing laws, now in place in many States, provide novice drivers the opportunity to gain additional driving experience before full driving privileges are extended to them. Features of graduated driver licensing laws vary from State to State. Most, however, provide for a gradual progression from the learner’s permit stage to full licensure, extending the length of time during which novice drivers can gain skills and driving experience. Some require that novice drivers log a minimum number of supervised driving hours. Most graduated driver licensing programs have zero tolerance for illegal drugs or alcohol. Many limit night driving as well as the number of teenage passengers. In New Jersey, the graduated driver licensing law prohibits permit holders who have not yet passed the road test required for initial licensure from using a cell phone or other wireless communication device while operating a vehicle [Assembly of the State of New Jersey 2001].

“Young drivers’ newness to the workplace compounds occupational safety concerns for a population that is already at high risk for vehicle crashes.”

Many of the risk factors that increase the likelihood that younger drivers in the general population will be involved in vehicle crashes are also present in the workplace. Young drivers are not only new behind the wheel, but their newness to the workplace compounds occupational safety concerns for a population that is already at high risk for vehicle crashes.

Federal regulations under the FLSA address vehicle safety concerns for young workers by prohibiting all on-the-job driving for 16-year-olds and placing limitations on the nature and amount of driving permitted for 17-year-olds (see Section 3.4.1). However, the FLSA does not cover young workers aged 18 and older, who are still in the process of developing driving skills and gaining experience. For this group of inexperienced young adult drivers, employers should consider postponing the assignment of intensive or time-sensitive driving tasks, thereby continuing to act in the spirit of both the FLSA and graduated driver licensing laws.

Research is now under way to develop driver training modules designed to improve skills that may be underdeveloped in young drivers. Many of these have been tested successfully in driving simulators but have yet to be evaluated outside the laboratory. One example is insight training, which is intended to make young drivers more aware of the unpredictability of typical driving situations and the limitations of their own skills [Gregersen 1996]. Another is variable priority training, which involves assignment of multiple tasks unrelated to driving, with the goal of improving novice drivers’ skills in dividing their attention between competing tasks and priorities [Regan et al. 1998a]. Other approaches seek to improve young drivers’ ability to (1) apply what they have learned to increasingly complex and dissimilar situations, (2) improve their ability to predict a sequence of events, and (3) evaluate their own performance and perceive hazards better by describing their thought processes to an instructor while actually driving [Deery 1999; Regan et al. 1998b]. If field testing of these approaches is successful, they may eventually be incorporated into driver training programs. Employers should consider implementing programs that use innovative teaching methods tailored to younger drivers, as they may be particularly useful in remedying potential deficiencies among this group.

4.4.2 Older Drivers

The need to accommodate older drivers is receiving increasing attention in the traffic safety community at large. As increasing numbers of Americans continue to work beyond the traditional retirement age of 65, the special needs of older drivers become a workplace safety issue as well. Older drivers have two traits in common with younger drivers: difficulty in responding to traffic hazards and a tendency to overestimate driving skills [Gregersen 1996; Holland 1993]. However, the reasons for these traits differ for the two age groups. Younger drivers may be at increased risk for crashes because they do not have enough experience to recognize, assess, and respond to hazards, and because they may be willing to accept higher levels of risk [Deery 1999]. With older drivers, the issue is not necessarily a lack of knowledge about what constitutes a hazard. Instead, the danger is that they may not anticipate and react to hazards quickly enough. Furthermore, they may not recognize their failure to deal with these situations as effectively as they did in the past [Holland and Rabbitt 1994]. Although their understanding of what is required may not decline, their ability to respond appropriately in a real-world driving situation may diminish.

Both younger and older drivers tend to overestimate their driving skills relative to other persons of their age [Gregersen 1996; Holland 1993; Pelz and Krupat 1974]. However, older drivers may overestimate their skills because of a decreased ability to react to high-risk situations—not from lack of knowledge about the risk (which is more likely to be the case among younger drivers). Older drivers report compensating for what they perceive to be high-risk situations by avoiding rush-hour driving, complex traffic situations, night driving, and long trips [Holland and Rabbitt 1994]. Avoiding such situations may not be an option for commercial drivers and others who drive for work.

Reduced reaction times (both physical and cognitive), reduced ability to divide attention between tasks, and increased difficulty in handling complex and unfamiliar situations are associated with the normal aging process and are widely recognized and well documented in the scientific literature [Brouwer et al. 1991; Holland and Rabbitt 1994; Maycock 1997; Stelmach and Nahom 1992]. In addition, normal aging results in declining visual acuity from reduced field of vision, less effective peripheral vision, and reduced ability to cope with glare from oncoming headlights and other sources [FHWA 2001; Maycock 1997]. The reduced range of head and neck motion associated with the normal aging process may also diminish the driver’s skill in scanning the driving environment [FHWA 2001]. Night driving poses particular risks for older drivers. Night vision depends on seeing contrasts between objects, not on visual acuity alone, and this sensitivity to contrasts decreases with age [Burnham and Abrams 1998].

Certain driving situations and maneuvers may increase the crash risk among older drivers. Intersections are problematic for older drivers, especially where they must make decisions about yielding the right of way [Maycock 1997]. Older drivers may have trouble interpreting pavement markings and reading street signs [FHWA 2001]. They also have difficulty negotiating interchanges. For example, those who are cited at freeway interchanges are most often cited for failure to yield and improper use of lanes [FHWA 2001]. Situations such as highway construction zones may be particularly hazardous, since older drivers may not react quickly enough to signs, traffic control devices, decreases in lane width, and lane closures and shifts. Construction zones may also be problematic for older drivers because they violate drivers’ expectations of how the roadway will be laid out.

“As the number of older workers increases, so will the number of older workers who drive on the job.”

Methods for assessing an older driver’s fitness to continue driving on or off the job should ideally draw on the expertise of various safety and health professionals. Some researchers have concluded that general medical screening alone is not sufficient to identify older persons who can no longer safely operate a vehicle; they believe that tests of cognitive functioning can more effectively identify older drivers who may be impaired [Johansson et al. 1996; Lundberg et al. 1998]. Other researchers have focused on developing new methods to evaluate older drivers’ visual perception skills, proposing that simple tests of visual acuity alone cannot reliably assess the ability to process the complex visual information needed for safe driving [Ball et al. 1988]. Related studies have shown that poor performance on more sophisticated tests of visual performance were highly predictive of crash involvement among drivers aged 55 to 90 [Ball et al. 1993; Ball and Owsley 1993; Owsley et al. 1998]. These tests assessed visual processing speed, ability to divide attention between centralized and peripheral objects, and ability to pay attention selectively in the presence of target objects and distractions.

As the number of older workers increases, so will the number of older workers who drive on the job. Employers will increasingly need to evaluate methods for giving older drivers continued opportunities for employment while ensuring that safety is not compromised. For many older persons, giving up driving is a life-changing event associated with a loss of independence and competence. However, employers will inevitably face the prospect of limiting or revoking driving duties of valued older workers. Such decisions should be made objectively after evaluating cognitive and visual ability and current levels of driving performance. Occupational medicine professionals, geriatric health professionals, and specialists in vision screening can help employers ensure that their medical screening programs effectively identify older drivers who may have trouble performing their duties safely. If it becomes necessary for an older worker to give up driving, employers should ideally make every effort to reassign the worker to nondriving duties that he or she can safely perform.

4.5 FLEET SAFETY ISSUES

Although workers employed in the transportation industry (which includes motor carriers) experience the greatest numbers of occupational fatalities because of vehicle-related roadway crashes, fully two-thirds of these occupational fatalities occur in industries other than transportation. In contrast to the unique regulatory climate in which the motor carrier industry operates, employers in other industries are governed by relatively few regulations specific to the operation of motor vehicles. The first of the prevention measures that are listed in Section 5.1 are intended to address fleet safety in these less regulated industries, but many are also relevant to the motor carrier industry. Conversely, numerous safety measures that are required for the motor carrier industry may be of value in other industries

Mandatory use of seat belts is the single most important driver safety policy that employers can implement and enforce. NHTSA estimated that in 2000, the use of seat belts prevented 11,889 fatalities in the United States and could have prevented 9,238 fatalities that did occur [NHTSA 2002a]. Seat belt use by front seat occupants reduces the risk of fatality by 45% for passenger car occupants and by 60% for light-truck occupants [NHTSA 2000b]. NHTSA also determined that nearly 143,000 moderate to severe injuries could have been prevented had all vehicle occupants worn seat belts [NHTSA 2002a]. As of February 2001, approximately 70% of vehicle occupants wore seat belts. Although no data are available to distinguish between belt use on or off the job, it is clear that nonuse of seat belts has substantial direct and indirect impact on employers.

As with any workplace safety policy, driver safety policies such as mandatory use of seat belts can be effective only if employers (1) tell workers how important the issue is to the company and (2) enforce safety policy with fairness and vigilance. High proportions of drivers in the general population report that they are more attentive to safe driving practices than the average driver, thus drivers may view safety messages as meant for someone else [Williams et al. 1993]. Delivering information about safe driving practices alone may not be sufficient to motivate workers to drive safely. Employers may need to provide additional motivation by (1) emphasizing the potential catastrophic consequences of a motor vehicle crash to both worker and family, and (2) clearly communicating that safety infractions will not be tolerated [Kedjidjian 1994; Lin and Cohen 1997; Williams et al. 1993].