Click for next page ( 11


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 10
10 medications found to increase risk in adults younger than age 65--namely, benzodiazepines (especially long-acting), cyclic antidepressants, and opioid analgesics. An excellent summary of the effects of specific drugs on driving plus guidelines for dealing with drug-impaired driv- ing problems can be found in the Drugs and Human Perfor- mance Fact Sheets developed by Couper and Logan (2004). This information reflects the conclusions of an international panel of experts in behavioral psychology, drug chemistry, forensic toxicology, medicine, and psychopharmacology as FIGURE 2 Proportion of crash-involved drivers in each age well as law enforcement personnel trained to recognize drug cohort taking multiple (two or more) PDI medications at time of effects on drivers in the field. crash. It is important to recognize the limitations of case-control Age-Related Functional Deficits that Predict studies. Although such studies can determine an association Crash Risk between a factor and an outcome, they cannot determine that the factor caused the outcome. It is difficult and often As emphasized earlier, a person's functional status--not impossible to separate the study factor of interest (e.g., tak- his or her age--determines fitness to drive. Medical condi- ing a medication) from other confounding factors that also tions and medication use deserve attention because of how could have been responsible for the outcome. Confound- they can compromise functional abilities, regardless of age. ing factors include a medical condition or multiple medical But even in the absence of these factors, the process of nor- conditions, severity of medical condition(s), drug interac- mal aging is accompanied by predictable functional deficits tion conflicts, driving exposure (number of miles driven), for the vast majority of the population--though individual use of other medications or alcohol, and length of exposure differences will grow ever more pronounced among older to the medication of interest. Furthermore, that a person cohorts of drivers. The following pages focus on age-related had a current prescription for a given medication does not changes in the abilities that research has demonstrated are guarantee that he or she was actually taking it on the day a most critical for safe driving. crash occurred. Also, there may be differences between the cases and the controls chosen, even though care is taken to Wherever possible, evidence tying functional loss to match them on specific characteristics. Nonetheless, these crash involvement--in particular, to at-fault crash involve- data are a "warning flag" about the increasing prevalence of ment--will be cited. This is most difficult when individuals single- and multiple-medication usage among older drivers. with significant deficits have already been screened out of For reference, Leroy and Morse (2008) found the most fre- the driving population. For example, because the visual acu- quently appearing drug combinations (in descending order ity test is already universally applied for private vehicle as of frequency) in the group of crash-involved drivers age 50 well as commercial vehicle operators [Ref. 49 CFR 391.42 and older to be as follows: (b)(10) Has distant visual acuity of at least 20/40 (Snellen) in each eye without corrective lenses or visual acuity sepa- Narcotics + NSAIDs rately corrected to 20/40 (Snellen) or better with corrective Skeletal muscle relaxants + NSAIDs lenses, distant binocular acuity of at least 20/40 (Snellen) in Narcotics + skeletal muscle relaxants both eyes with or without corrective lenses, field of vision of Narcotics + skeletal muscle relaxants + NSAIDs at least 70 in the horizontal meridian in each eye, and the Narcotics + antibiotics ability to recognize the colors of traffic signals and devices Gastric acid secretion reducers + narcotics showing standard red, green, and amber], very few driv- Anti-anxiety drugs + narcotics ers on the road have a significant deficit in this ability and Selective serotonin reuptake inhibitor antidepressants therefore crash analyses that compare them with drivers who + narcotics have "good" vision are statistically untenable. With such a Narcotics + NSAIDs + antibiotics restriction in the range of acuity levels in the current fleet, only very weak relationships, if any, are found when ana- In comparison, Wilkinson and Moskowitz (2001) lyzing differences in this functional ability to predict crash reviewed 11 epidemiological studies of medication use and outcomes. In this case, differences in driver performance on traffic safety risk (primarily in older drivers) in the United a reliable road test offer the best evidence available. States and Canada between 1991 and 2000, and concluded that the prescription drugs most likely to be associated with Performance decrements are often presumed to mediate motor vehicle crashes by older drivers include the same CNS crash risk, and crash studies typically involve large samples,

OCR for page 10
11 lengthy intervals, and considerable expense. This helps According to this literature review, prior studies comparing explain why the bulk of research in this area is cross-sec- the safety records of hearing-impaired with those of non- tional and part-task, whether it is carried out in a laboratory, impaired drivers of private (noncommercial) vehicles have a simulator, or on a closed course. Fortunately, the growing found that hearing-impaired drivers had roughly 1.8 times concern with medical and functional fitness to drive in an the number of accidents as nonhearing-impaired drivers. aging population has focused increasing resources on this This finding was fairly consistent across the studies reviewed problem in recent years, making it possible to point with in this area. greater confidence to a relatively small set of functional abilities that appear to predict crashes most strongly. Another study, Hearing Disorders and Commercial Motor Vehicle Drivers, has suggested that when visual infor- Deficits in Sensory Ability mation is diminished, an auditory warning becomes more important to prevent crashes (Songer et al. 1993). In addi- Sensing (detecting) hazards is critical to safe driving. If a tion, sounds sometimes give drivers the first awareness that conflict with a pedestrian, another vehicle, or an obstruc- a problem or condition demanding a timely driver response tion in the roadway is not sensed or detected, the driver will exists, including both emergency sounds such as sirens and not recognize the hazard and may fail to execute the control warnings of equipment failure. movements necessary to avoid it. Alternately, if a hazard is sensed late, the subsequent perceptual/cognitive and psycho- Attempts to quantify the safety impact of hearing loss motor stages of information processing will also be delayed, for commercial vehicle operators remain inconclusive, as the slowing the driver's response. FHWA hearing standard ensures that the same "restriction of range" for this functional ability applies as in the case The technical literature commonly describes driving as of visual acuity. But in the private vehicle fleet, where this primarily a visual task in the sense that an alert individual driver ability is not regulated, a relative risk associated with can obtain all the information he or she needs to safely oper- hearing loss has been demonstrated. This elevated crash ate a motor vehicle through visual inputs. But the role of involvement for hearing-impaired drivers of private vehicles hearing cannot be discounted for commercial drivers, par- raises concern about the consequences of a sensory deficit ticularly in local and short-haul operations that take place in that can delay hazard perception for operators of commercial congested urban settings and include frequent backing vehicles, with their larger sizes, heavier weights, and longer maneuvers using vehicles from which visibility is severely braking distances. limited. Robinson et al. (1997) investigated the functional requirements for safe commercial vehicle operation and, In consideration of visual deficits, the existing standard after gaining a detailed understanding of what commercial and associated "restriction-of-range" problem in studies drivers must do to safely perform their jobs, identified a examining the relationship between (static) visual acuity number of specific tasks for which hearing is critical. These and motor vehicle crashes and convictions has already been researchers concluded that an FHWA hearing requirement is noted. One study on this subject carried out by Rogers and necessary, and provided recommendations not only for hear- Janke (1992) compared the driving records of heavy-vehicle ing standards [See 49 CFR Part 391.41 (b) (11) First perceives operators whose vision was unimpaired with the records of a forced whispered voice in the better ear at not less than 5 heavy-vehicle operators who were found to have a deficit feet with or without the use of a hearing aid or, if tested by in static acuity. Overall, the visually impaired drivers were use of an audiometric device, does not have an average hear- found to have a higher incidence of crashes and convictions ing loss in the better ear greater than 40 decibels at 500 Hz, than the unimpaired drivers. However, when the visually 1,000 Hz, and 2,000 Hz with or without a hearing aid when impaired group was further subdivided into those who were the audiometric device is properly calibrated [American moderately impaired and those who were severely impaired, National Standard (formerly ASA Standard) Z24.51951] the incidence of crashes for moderately impaired and unim- but also for improved testing methods to address the ques- paired drivers did not differ significantly regardless of age. tionable reliability of the "whisper test." Pure tone audiomet- Indeed, the older drivers in the broader visually impaired ric tests were recommended for assessing hearing level in sample had lower conviction and crash rates, despite their the future, with different sensitivities required at different overall poorer visual acuity as a group. frequencies as opposed to an average hearing level. Given the desire to predict future crash risk based on func- The Robinson et al. (1997) research was performed in tional status measurement, it is a stronger research design four phases, including a literature review; the most exten- to use prospective rather than retrospective crash data. Sev- sive commercial driving task analysis ever undertaken; field eral studies involving older drivers of private vehicles have measurements of driver hearing, truck-cab noise, and noise used this approach. Rubin et al. (2007) found no association exposure; and evaluations of the audibility of speech, warn- between static acuity and future crashes in a population- ing signals, and other sounds during commercial operations. based, prospective study, using 1,801 current drivers ages 65

OCR for page 10
12 to 85. Another investigation found no significant relationship the ability to recognize and avoid road hazards. Specifically, between impaired acuity (defined as resolution worse than a correlation of r = 0.57 described the association between 20/40) and injurious crash risk in a prospective cohort study contrast sensitivity level and detection of 1 2.2-m sheets of with 3 years of follow-up crash data, using 294 drivers ages 80-cm-thick gray foam rubber placed on the roadway. These 55 to 87 (Owsley et al. 1998). targets represented objects that were large relative to the resolution limits of the eye--so performance was not related Dynamic visual acuity--the ability to resolve the details to acuity--but were of low contrast. Wood (2002) concludes of a moving (instead of a stationary) target--is arguably that older drivers, even in the early stages of an ocular dis- more relevant to the actual demands of driving. Some activi- ease that produces contrast sensitivity deficits, are likely to ties that appear to rely on dynamic acuity are reading street have significant problems with seeing and avoiding potholes, signs while in motion, locating road boundaries when nego- highway debris, speed bumps, pedestrians, and other vehi- tiating a turn, and making lateral lane changes. In these situ- cles, especially when driving in poor visibility conditions ations, greater speeds are associated with greater demands such as rain or fog. for dynamic visual acuity. For many years, a slight advan- tage for dynamic visual acuity versus static visual acuity has Uc et al. (2005) found that contrast sensitivity was sig- been observed in crash studies; but the correlations between nificantly correlated with detection of road signs and com- dynamic visual acuity and crash rates have also been con- mercial landmark signs in an on-road study using 170 older sistently weak. As reported by Hills and Burg (1977), this drivers (rs = 0.44, p < 0.0001). Subjects with better contrast correlation was too low (r = 0.054) to be of any practical sensitivity were able to identify a higher percentage of the significance for identifying at-risk drivers. signs than subjects with poorer contrast sensitivity. Although this deficit could affect all drivers, its impact on commercial Another visual ability that declines with age, and vehicle operators is arguably more important. appears to be more strongly related to motor vehicle crash risk, is contrast sensitivity. Tests of this ability measure the Finally, in a study of 12,400 Pennsylvania drivers ages 16 response to the full range of spatial frequencies, including to 75+, Decina and Staplin (1993) did not find a significant not only sharply defined black-on-white targets as in an acu- relationship between contrast sensitivity and crash rates in ity test, but also those that are grayer and have less distinct a 3.67-year retrospective analysis. Neither was visual acu- edges. Contrast sensitivity makes it possible to drive in rain ity nor horizontal field measures, in isolation, significantly or fog, at twilight, or in other poor visibility conditions. It related to crash involvement. However, they did find that a is also especially important for detecting curbs, barriers, or composite measure including contrast sensitivity, binocular objects that do not stand out sharply from their background, visual acuity, and horizontal field measurement was signifi- including pedestrians at night, lane and road edge boundar- cantly related to crash involvement for drivers age 66 and ies that are not marked conspicuously with fresh paint, and older. Failure on a combined criterion that incorporates a reflectors. Under the same lighting conditions, the median binocular acuity standard of 20/40, a horizontal visual field 75-year-old requires at least twice the contrast in a scene to standard of 140 degrees, and below-normal scores for one or see as well as a 55-year-old (Brabyn et al. 2000). more of the three spatial frequencies of sine wave gratings in a contrast sensitivity test using a countertop vision tester In a population-based study with older drivers, Rubin et produced the strongest relationship linking poor vision and al. (2007) found that reduced contrast sensitivity was associ- high crash involvement, especially for the ages 6675 and ated with poor driving performance, self-reported driving age 76+ driver groups (Decina and Staplin 1993). difficulty, self restriction, and prior crash involvement; how- ever, the association with future crashes was not significant These findings raise the issue of visual field loss among after adjusting for miles driven. Yet in Brown and others' older drivers. It has long been established that the isopters, or 1993 study of 1,475 ITT Hartford Insurance Company poli- borders, of the visual field are constricted as a consequence cyholders (ages 5080+), who were divided into two groups of aging, with older adults exhibiting a generalized loss in based on the presence or absence of recent at-fault crashes, sensitivity throughout the central 30 degrees of vision and the PelliRobson contrast sensitivity test consistently yielded a slightly greater sensitivity reduction in more peripheral the highest correlation to crashes (r = -0.11, p < 0.05). These areas (Jaffe et al. 1986; Johnson et al. 1989). As with visual researchers noted that because contrast sensitivity was nega- acuity, a minimum visual field size is required under 49 CFR tively correlated with age itself (r = -0.40), the relationship Part 391.41, as cited earlier, with the same consequences for between performance on the PelliRobson chart and crash analyses relating visual field loss to safety outcomes for involvement was probably understated. (older) commercial drivers. In a study by Wood (2002) of visual impairment, age, and Historically, studies among drivers of all ages examining driving performance, contrast sensitivity as measured by the visual field sensitivity and crashes have typically failed to PelliRobson test chart emerged as a significant predictor of find a relationship between them (Council and Allen 1974;

OCR for page 10
13 Shinar et al. 1977; Waller et al. 1980); this includes an early sory inputs a driver receives from his/her eyes and ears. Key study involving bus and truck drivers (Henderson and Burg perceptual-cognitive abilities include (1) working memory, 1974). One exception is a large sample study (n = 10,000) by including speed of processing; (2) attentional processes, Johnson and Keltner (1983) who found that the small sub- including selective and divided attention; and (3) visuospa- set of drivers with severe binocular visual field loss (mostly tial abilities, such as pattern perception and visualization older drivers) had crash and conviction rates twice as high of missing information. These abilities have been empiri- as those with normal visual fields. More recent research cally validated as significant predictors of at-fault crash risk suggests that the effect of a visual field loss on driving is among older (private vehicle) drivers, but they have strong strongly related to its location, with the greatest safety risk construct validity as crash predictors for commercial vehicle resulting from losses in the central field and in the horizontal operators as well. meridian (Lovsund et al. 1991), including the lower periph- ery (Rubin et al. 2007). One other cognitive ability with strong construct validity as a crash predictor is executive function. In technical litera- Even as the relationship between visual field loss and ture, this term is variously used to encompass the evaluative crash risk remains difficult to quantify, given existing con- aspects of cognition, such as planning and judgment, as well trols to remove those with severe deficits from the fleet, it is as the process through which the memory, attentional, and crucial to distinguish reduced visual field size or sensitivity visuospatial abilities noted earlier are integrated to support as a sensory function from the related component of visual driver decision making. This review will not focus explicitly attention commonly termed "useful field of view." This is on executive function, however, for several reasons. First, actually a perceptual-cognitive ability, for which reliable this construct remains only vaguely defined, and accord- age differences and a significant relationship with at-fault ingly there are no methodologies for testing it that are widely crashes have been demonstrated, as discussed in the follow- accepted and which produce reliable results with a clearly ing section. established relationship to safety measures. Perhaps more important, this review targets potential safety concerns with To conclude the review of functional changes in vision the workforce of normally aging commercial operators, not with advancing age that have a potential impact on traffic those older individuals with symptoms of dementia; that is, safety, the regulatory requirement to "recognize the colors the group where evidence of deterioration in executive func- of traffic signals" must be acknowledged. A differential tion is most prominent. increase in blueyellow errors as a function of age is most prominent (e.g., Verriest et al. 1982); the mean error for naive Working memory allows the integration of continuous sen- subjects older than age 70 is greater than 100, compared with sory information over time, the manipulation of information a mean error score of 37 for the ages 20 to 30 group on the in memory for problem solving and decision making, and Farnsworth Munsell 100 hue test. the division of attention between multiple relevant sources of information such as an intersection control display and Traffic safety researchers including Temple (1989), Brown oncoming traffic. As a practical example, this ability comes et al. (1993), and Tarawneh et al. (1993) have reported that into play when a driver is fully engaged in vehicle guidance correlations of deficiencies in color vision with on-road driv- and maintaining a safe headway in relation to a lead vehicle, ing performance were not significant. Where significant cor- while at the same time recalling information about the route relations with simulator performance could be demonstrated, he or she wishes to follow and the next navigational land- such findings have not suggested any practical consequence mark for which to watch. for performance of critical driving tasks. Also, color vision was not significantly correlated with older drivers' prior at- One way that researchers have operationalized working fault crashes in studies by Owsley et al. (1991) or by Ball et memory is to provide drivers with a to-be-remembered set al. (1994). In other words, available data do not indicate that of words, number, or ideas; confirm that they have learned an increase in crash risk specifically as the result of deficits this material; then require performance of another task that in color vision will be a major concern, with the aging of the interferes with rehearsal of the new information; and finally commercial driver population. ask that they recall what they memorized earlier. Using this type of "delayed recall" measure with a sample of 1,876 Deficits in Perceptual-Cognitive Ability Maryland drivers age 55 and older, Staplin et al. [2003 (b)] found that those who scored poorly (2 out of 3 verbal stimuli Vision, and to a lesser extent hearing, are essential to provide recalled incorrectly after a 5-minute intervening task) were drivers with the information they need to make decisions and 2.92 times more likely to be in an at-fault crash and 1.72 execute timely vehicle control movements, yet from moment times more likely to be cited for a moving violation than to moment these decisions and actions critically depend on older drivers who performed well on the test. Crash and a core set of perceptual-cognitive abilities. These abilities violation history included 1 year of retrospective data and are needed to filter, attend to, and make sense of the sen- 20 months of prospective data. In an updated analysis from

OCR for page 10
14 this sample that added 12 more months of prospective driv- Next, two complementary functions that are essential to ing history, the odds ratio for crashing increased to 3.34 for safe driving performance and have been associated with sig- drivers who performed poorly on the delayed recall measure nificant age differences are selective attention and divided [Staplin et al. 2003(a)]. attention. Selective attention involves the earliest stage of (visual) attention used to quickly capture and direct attention Similarly, Johansson et al. (1996) used a five-item delayed to the most salient events in a driving scene. Because of the recall test, where the subject was required to name and recall vast quantity of information that is continuously available five objects viewed on a desk after a 10-minute period, in a in the driving environment, the ability to selectively attend case-control study to compare 23 cases with crashes and 29 to information that is of primary relevance for maintaining control subjects with no crashes in the past 5 years. Results driving function is key. Divided attention pertains to the showed that the crashed drivers had poorer five-item recall ability to monitor and respond effectively to multiple sources ( p < 0.003). And, in McKnight and McKnight's (1999) study of information at the same time; for example, a driver enter- of 407 drivers age 62 and older, split into incident-involved ing a freeway must track the curvature of the ramp and steer and nonincident-involved groups, measures of delayed short- appropriately, keep a safe distance behind the car ahead, and term memory showed a significant and fairly strong correla- check for gaps in traffic on the highway, while at the same tion with unsafe driving (r2 = 0.32). time accelerating just enough to permit a smooth entry into the traffic stream. Salthouse and Babcock (1991) reported that age-related variance in working memory was largely accounted for by Pietras et al. (2006) found that older drivers with selective measures of processing speed. In other words, age-related attention deficits had shorter time-to-collision values (5.60 s decline in working memory is not the result of differences vs. 6.86 s), took longer to cross the roadway (5.42 s vs. 4.84 in the capacity of working memory, but rather in that older s), and had shorter safety cushions (the difference between adults require more time to retrieve information from mem- time-to-collision and time to cross the roadway) than older ory (Waugh et al. 1978). Given this limitation, and that infor- drivers with no impairment in their selective attention capa- mation in immediate memory has a limited "life span," older bilities. Their traffic entry judgments were made from an adults will be penalized on tasks that require substantial instrumented vehicle parked in an entry driveway perpen- attentional resources, or on tasks that require the reorganiza- dicular to a busy two-way, four-lane highway. Oncoming tion of "to-be-remembered" information. As a consequence, traffic entered the view of the participant approximately older drivers will be at greater risk in situations such as inter- 1,000 ft away, and drivers were asked to press a button to sections that require rapid mental operations for appropriate mark the last possible moment that he or she would cross the vehicle control, especially when simultaneously required to road in front of a specific oncoming vehicle. perform such operations and retain other (e.g., navigational) information for future use. The most promising work addressing issues of selective attention and traffic safety arose, interestingly, from the gen- Using the Visual Attention Analyzer, Useful Field of eral failure of earlier studies to find a reliable relationship View subtest 1, Hennessy (1995) found that speed of pro- between visual field sensitivity and motor vehicle crash expe- cessing deficits in a sample of 1,235 California drivers accel- rience. Driving involves complex scenes with moving and/or erated after age 70. Comparing performance on this test to distracting stimuli, plus the necessity of constantly dividing California Department of Motor Vehicle crash records, poor one's attention between central and peripheral vision. These performance on the speed of processing subtest accounted attributes are not present in conventional visual field sensi- for 2.8% of the variance in crash involvement ( p < 0.05) for tivity tests. Another approach examining the "functional" drivers age 70 and older, without adjusting for demograph- or "useful" field of view involves the detection, localization, ics or driving exposure. After adjusting for age, gender, and and identification of targets against complex visual back- driving exposure in the group of drivers age 70 and older, grounds, often with distractors in the scene, and a limited performance on the speed of processing subtest accounted time to process the display. Tests assessing the useful field for 4.1% of the variance in crash involvement. of view are better predictors of driving problems than are standard visual field tests. The impact of the decline in speed of processing with advancing age derives from its cross-cutting nature--it Ball et al. (1994) found that drivers with restrictions in appears to mediate a number of other cognitive functions-- their useful field of view had 15 times more intersection and from its prevalence among older drivers. While keep- crashes than those with normal visual attention, and the cor- ing in mind that individual differences in (cognitive) ability relation between crash frequency and useful field of view increase with age, an overall slowing of mental processes has exceeded r = 0.55. In an analysis of 278 drivers from this been postulated beginning as early as the fifth decade and sample, McGwin et al. (1998) found that the odds ratio for accelerating for most individuals as they continue to age into crashes for drivers with a 40% or more reduction in use- their 70s and beyond (Cerella 1985). ful field of view was 13.7 compared with drivers with less

OCR for page 10
15 than 40% reduction when the dependent measure was state- ful field of view protocol separately, the strongest association recorded crashes. Goode et al. (1998) found that the useful was with divided attention (with a hazard ratio of 1.47, p < field of view measure reliably distinguished between older 0.0001). The hazard ratio for processing speed was 1.27 ( p < drivers who had experienced one or more state-recorded 0.04), and for selective attention was 1.45 ( p < 0.22). at-fault crashes in the prior 5-year period from older driv- ers who were crash free. Their sample was 239 drivers age An additional aspect of perceptual-cognitive function 60 and older (mean 70.4), drawn from the Ball et al. (1994) that has been identified as a significant predictor of crash study, 115 of who were crash free and 124 who were crash risk is visuospatial ability. The ability to visualize missing involved. The overall classification rate was 85.4%, with a information is valuable in many situations: when a stop sign sensitivity of 86.3% and a specificity of 84.3%. is partially obscured by foliage; when only an arm or leg of a pedestrian stepping out from behind a parked car is visible Age-related deficits in selective attention may also explain to an approaching motorist; or when the leading edge of a older driver failure to detect pedestrians and other vehicles potential conflict vehicle first appears at an intersection with at intersections. Caird et al. (2005) used a change blindness a road or driveway where the sight triangle is restricted. method in their laboratory to assess turn decision accuracy. Change blindness--the inability to detect changes made to The Goode et al. (1998) study included two measures an object or scene during a flicker, or blink--was used in this of visuospatial ability, the ReyOsterrieth Complex Fig- study to assess visual attention. The study sample included ures test, which involves first copying a complex figure and 62 drivers distributed across four age groups: 18 to 25, 26 to then attempting to draw the figure from memory, and the 64, 65 to 73, and 75+. For intersections where pedestrians Visual Reproduction subtest of the Wechsler Memory Scale appeared, drivers age 65 and older had especially low accu- (WMS-VR), which requires memorization of a visual stimu- racy scores compared with younger drivers, even when the lus and construction of the stimulus from memory. Perfor- pedestrians were in a crosswalk. Failure to detect the pedes- mance on these measures distinguished between 124 older trians may have led older drivers to decide that the intersec- drivers who had experienced one or more state-recorded at- tion was clear and a turn maneuver was safe to complete. fault crashes in the prior 5-year period from 115 crash-free Older drivers also missed detecting relevant vehicles that older drivers. The WMS-VR had a sensitivity of 66.1% and a were relatively large and conspicuous. specificity of 52.2%. The ReyOsterrieth test had a sensitiv- ity of 50% and a specificity of 61.7%. Drivers' difficulties in negotiating intersections also may reflect the divided attention demands they face in these situ- Lundberg et al. (1998) also employed the ReyOsterri- ations. In their study of 1,876 Maryland drivers age 55 and eth test to compare older, crash-involved drivers who had older, referenced earlier, Staplin et al. [2003(b)] found that suspended licenses because of serious moving violations older subjects who scored poorly (took 300 msec or longer) on (running red lights, not yielding right-of-way rules, and not the useful-field-of-view subtest that measures divided atten- heeding stop signs) with older, noncrashinvolved drivers tion were 3.11 times more likely to be in an at-fault crash in whose licenses had been suspended for violations, and with the 20 months following assessment than older drivers who older control drivers with clean driving records. The former performed well on the test. Looking at 20 months of prospec- group performed significantly worse on the visuospatial tive driving history data plus 1 year of retrospective driving memory component of this test. history data, drivers performing poorly on the useful-field- of-view subtest were 2.48 times more likely to be involved in In the research with older Maryland drivers reported a crash and 1.67 times more likely to be cited for a moving by Staplin et al. [2003(b)], those who scored poorly on the violation. In the updated crash analysis conducted by Staplin Motor Free Visual Perception TestVisual Closure subscore et al. [2003(a)] that added 12 more months of prospective (5 or more incorrect out of 11) were 6.22 times more likely crash data, the odds ratio for at-fault crashes was 2.23. to be in a crash in the 20 months following assessment than older drivers who performed well on the test. This measure Rubin et al. (2007) examined the relation of selective atten- depends on the visualization of missing information, by tion, divided attention, and processing speed to prospective requiring the identification of which of four incomplete fig- crash involvement in a sample of 857 older drivers. Overall, ures (line drawings) could be completed to match an exam- they found that a useful field of view reduction of 40% or ple figure. Crash and violation history were obtained from greater was associated with a 2.12 increase in crash risk in the state motor vehicle administration. Adding 1 year of the following 2-year period (using state-recorded crashes, retrospective crash history to the 20 months of prospective but not determining fault) compared with drivers with no crash history resulted in an odds ratio for at-fault crashes of loss, after adjustment for demographic and health-status 4.96 and an odds ratio for moving violations of 4.53. In a sub- variables (95% CI = 1.323.39, p < 0.01). After adjustment sequent analysis by Staplin et al. [2003(a)] that looked at 32 for miles driven, the hazard ratio increased to 2.21 (95% CI = months of prospective, plus 1 year of retrospective d riving 1.323.39, p < 0.01). Analyzing the three subtests of the use- history data, drivers age 55 and older performing below the

OCR for page 10
16 cut point for this measure were at 3.6 times greater risk of Physical reaction time was the time it took for the foot to an at-fault crash. start moving from the accelerator until the brake pedal was fully depressed. The finding that physical condition was Deficits in Psychomotor Ability associated with initial reaction time was unexpected, but the authors suggested that participants with pain in the lower The physical capabilities (psychomotor functions) needed extremities may find it more difficult to initiate movement, for driving include strength, range of motion of extremi- and the delay may contribute to slower initial reaction time. ties, trunk and neck mobility, and proprioception. With the exception of proprioception--the ability of the kinesthetic A study of 283 community-dwelling individuals ages 72 receptors to determine where one's limbs are at any given to 92 found that a measure of leg strength and mobility, the moment, allowing for the coordination of movement--it rapid-pace walk, was a significant ( p < 0.05) predictor of has been well established that physical capabilities decline who reported an "adverse driving event" in the year follow- as a function of age. National estimates of the prevalence of ing testing (Marottoli et al. 1994). Adverse events included age-associated functional impairments for physical mobil- crashes, violations, and being stopped by police. ity are available from studies such as the National Health Interview Survey (Pleis and Lethbridge-ejku 2007). In Next, the extent and effect of restricted head movement the National Health Interview Survey, age was positively were investigated by Isler et al. (1997), for drivers in four age associated with an arthritis diagnosis and the presence of groups: under age 30, ages 40 to 59, ages 60 to 69, and age chronic joint symptoms. Forty-four percent of adults age 70 and older. Their methodology required drivers arriving 75 and older had chronic joint symptoms, and those ages 65 at the decision point of a T-intersection to judge the distance to 74 had an almost equal amount (43%); whereas percent- and speed of traffic approaching on the intersecting roadway, ages for those ages 45 to 64 dropped to 34%, and to 15% to find a safe gap in which to merge. In this situation, con- for those ages 18 to 44. siderable head movement was needed to bring intersecting traffic into central vision, to support these perceptual judg- The relevance of such physical limitations for safe driving ments; and the oldest subjects exhibited an average decre- is indicated in a general sense by the findings of Diller et al. ment of approximately one-third of head range of movement (1999). These researchers found that licensed drivers in their compared with the youngest group of subjects. As a conse- medical program in Utah, who had functional motor impair- quence, the oldest drivers could not bring oncoming traffic ment but no license restrictions, had a significantly higher into central vision at distances exceeding 65 ft (20 m) with- citation and at-fault crash rate than a comparison group of out additional eye movements. Unfortunately, in addition drivers who were not part of the medical program matched to their restricted head movements, the oldest drivers also on age, gender, and county of residence. The relative risks for had restricted horizontal peripheral vision. The authors state citations and at-fault crashes were 1.42 and 1.18, respectively. that the combination of restricted head movement, combined with deficits in peripheral vision, increases the difficulty of The specific age-related changes in psychomotor func- bringing an approaching vehicle into central vision, where tion of greatest concern for safe driving appear to be a loss acuity is the highest, and may help explain why older drivers of leg strength and range of motion, and reduced head/neck have higher rates of intersection crashes that result in injury flexibility. These physical deficits have been linked to dif- or death. ficulties with pedal (brake) control and with visual search to the sides and rear of the vehicle, respectively; and associated Finally, there is solid evidence about the safety impact increases in crash risk have been documented. Thus, despite of these age-related physical limitations from the Mary- the reduced physical demands for operating modern passen- land Pilot Older Driver Study [Staplin et al. 2003(b)]. This ger and commercial motor vehicles, concerns remain about research included the rapid-pace walk as a measure of leg the safety impact of changes in physical function. strength and mobility, plus a measure of head/neck rotation designed to determine whether an individual could look Zhang et al. (2007), in a study including studies of 1,039 directly over his or her shoulder as needed to safely change drivers age 67 and older, found that those who reported three lanes or merge; for the latter measure, the lower torso was or more complaints of pain in the feet, hips, legs, or cur- fixed in place with a seatbelt, as when driving. Both of these rent treatment for arthritis, had significantly slower brake measures were significant predictors of at-fault crashes in reaction speeds than drivers with no complaints of pain an observation period that included 1 year of retrospective in these areas, in terms of both initial reaction speed and crash data plus an average prospective observation interval physical response speed. Initial reaction time was measured of 20 months, for 1,876 drivers age 55 and older; their respec- from the time the signal turned red until the time there was tive odds ratios were 2.64 (leg strength/mobility) and 2.56 a motor response (foot moving off the accelerator pedal). (head/neck rotation).