This article was downloaded by: [Linköping University Library]On: 12 August 2014, At: 04:55Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Applied NeuropsychologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/hapn20

Neuropsychological Aspects of Driving After BrainLesion: Simulator Study and On-Road DrivingAnna Lundqvist , Johan Alinder , Hakan Alm , Bjorn Gerdle , Sten Levander & JerkerRonnbergPublished online: 07 Jun 2010.

To cite this article: Anna Lundqvist , Johan Alinder , Hakan Alm , Bjorn Gerdle , Sten Levander & Jerker Ronnberg (1997)Neuropsychological Aspects of Driving After Brain Lesion: Simulator Study and On-Road Driving, Applied Neuropsychology, 4:4,220-230

To link to this article: http://dx.doi.org/10.1207/s15324826an0404_3

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Applied Neuropsychology 1997, Vol. 4, No. 4,220-230

Copyright 1997 by Lawrence Erlbaum Associates. Inc

Neuropsychological Aspects of Driving After Brain Lesion: Simulator Study and On-Road Driving

Anna Lundqvist and Johrrn Alinder Department of Rehabilitation Medicine, Linkoping University, Linkoping, Sweden

Hiikan Alrn Department of Mechanical Engineering, Linkiiping Institute qf Technology, Linkoping, Swea'en

Bjorn Gerdle Department of Rehabilitation Medicine and Pain Clinic, University Hospital,

Linkoping University, Lin,&ping, Sweden

Sten Levander Department of Psychiatry, Lund University, Malmo, Sweden

Jerker Ronnberg Department of Education and Psychology, Linkzpiizg University, LinkZiping, Sweden

Twenty-nine patients with brain lesion and 29 matched controls participated in the studv. The patients were socially well recovered with a high rate of employment. Compared with the controls, they pe$ormed significantly worse on a neuropsychological test battery, especially on executive and cognitivefunctions. Patients drove as well as controls in predictable situalions in the advanced simulator used. In unpredictable situations, they demonstraied longer reaction times and safety margins, as well as dificulties in allocating processing resources to a secondary tusk. The patients showed significantly less attention, worse tr&c behavior, and less risk awareness when driving in real trafic. Forty-one percent of the patients did not pass the driving i'est. The neurops~~chological test battery was factor analyzed into four factors: executive capacity, cognitive capacity, automatic attentional capacity, and simple perceptual- motor capacity. The second factor was the mast significant with a simultaneous capaciql test predicting driving pe$ormance with 78% confidence.

Key words: brain lesion, cognitive impairment, driving, simulator driving

Driving is an activity of daily living that means auton- omy and independence. Too little empirical research has been done on neuropsychological aspects of driving (Christie, 1996; Van Zomeren, Brouwer, & Minderhoud, 1987). To decide whether a patient after brain lesion has regained sufficient cognitive ability to drive, one needs a multidimensional approach and systematic methods.

Residual signs of brain damage are, for instance, de- creased speed of cognitive processing, attentional prob-

lems, and decreased efficiency of motax performances (Stuss, Shallice, Alexander, & Picton, 1995; Van Zomeren & Brouwer, 1992). Patients' subjective, re- ported symptoms are distractibility and fatigue when concentrating, as well as difficulties in doing two tasks simultaneously.

Subtle neuropsychological deficits, due to cerebral lesion or disease, might remain for a long time or become permanent (Stuss et al., 1985). The patient may not have visible dysfunctions-that is, he or she can walk and talk-but a phjrsical impact to the head might

We thank The Swedish Nahonal Road Administration and The brain dysfunc.ons, even if the brain is not injured National Board of Health and Welfare for their financial support, and Ostergotland County Eoard Administration for practical help. directly. Moreover, impact damage to white matter due

Requests for reprints should be sent to Anna Lundavist. De~art- shear strain Or damage may cause a diffuse . . . ment i f ~ e h a b i l i t k o n Medicine, Link6ping University, Faculty of lesion (Filley, 1995; L,evin, Grafman, & Eisenberg Health Sciences, 58185 Sweden. 1982). Cortical frontotempord and subcortical struc-

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

NEUROPSYCHOLOGICAL ASPECTS OF DRIVING

tures are especially vulnerable. Mental slowness of peo- ple with cerebral lesion has been reported quite fre- quently (Stuss et al., 19135; Van Zomeren, Brouwer, & Deelman, 1984). The slowing down is found in executive functions and informati1on processing (Brooks, 1984), symptoms that are connected with different aspects of attention and processing. Speed of cognitive processing, allocation of attention, and control are suggested as relevant parameters for safe driving (Ranney, 1994).

Attention is a multipl~e system with different compo- nents. One anatomically posterior sensory attentional system is responsible for spatial attention, and an ana- tomically anterior system is concerned with control of attention (Posner, 1995). Optimal performance requires frequent shifts between Fast, less demanding automatic processing and attentional control, which is slow and requires cognitive effort and concentration (Posner, 1990; Stuss et al., 1995).

Ranney (1994) described different models of driving behavior that emphasize: cognitive demands. Informa- tion processing models suggest perception, decision making, and response selection (Wickens, 1992), whereas motivational models emphasize situation-spe- cific conditions and risk awareness (Fuller, 1984; NEitiinen & Summala, 1976; Wilde, 1982). Hierarchical control models include activities at strategic, tactical, and operational levels of control (Michon, 1985). Plan- ning the driving according to weather, traffic, and one's own condition is done at the strategic level. Making decisions in real traffic and adjusting driving to current demands (e.g., overtaking) are done at the tactical level. Instant reactions (e.g., braking suddenly) are done at the operational 1e:vel. The tirne available to make decisions is crucially different at thethree levels ofcontrol. Shiffrin and Schneider f 1977) introduced the concept of automt- icit), into models of driving behavior. With practice, behavior becomes automatized; for exampIe, steering, shifting, and braking car be performed simultaneously with more demanding ac:tivities, like decision making and listening to the radio (Wickens, 1984). Automatic processing is basically perceptual, whereas controlled processing is cognitive arid executive.

As earlier research has proved that cognitive proc- essing is often affected by brain lesion, one important question is whether brain damage also influences driv- ing performance. Therefare, we propose that decreased performance on neuropsychological tests measuring speed and acc:uracy in cognirive processing, especially required at the tactical level of control, predicts lower driving perfarmance.

A variety of approaches for assessing driving skill have been assessed. One way to study driving perform-

ance is in real driving. Artother way is In an advanced driving simulator. On-road evaluations have been re- garded as a direct measure of driving performance, how- ever, often with poor reliability (Hopewell &Price, 1985). Simulator driving can be standardized but it is not quite reaJistic. Simulators have often been too simple (Hopewell & Price, 1985; Quigley & I)e Lisa, 1983). The predictive value of cognitive capacities, measured with neuro- psychological tests, for actual driving performance has given mixed results @rcoke, Questad, Patterson, & Valois. 1992; Galski, Bruno, & Ehle, 1993; Galski, Ehle, & Bruno, 1990; Sutndet, Cioffeng, & EIc~fft, 1995; Van Zon~eren, Brouwer, Rothengatter, & Snoek, 1988). 'Brain-damaged drivers could not, in general, be seen as risky drive~x, although some individuals show decreased driving skill and risky behavior in traffic? (Van Zomeren et al., 1987, p. 697). Much earlier research is criticized because of lacking control in terms of agc, sex, education, and driving experience (Christie, 1996).

This study focused on ,n group of patients suffering from brain lesion, but with "good recovery," and a con- trol group. The participants were assessed in the follow- ing ways. First, a neuropsychological test battery was assumed to measure perceptual, cognitive, and executive functions. Perceptual functions were measured with per- ceptual and visuomotor tests. Cognitive Functions were measured with tests requiring informabon processing and divided attention. Executive functions were meas- ured with a test requiring planning, shifting of set, and sustained control. Next, sirnulator driving, having built- in varying predictable driving situations, was used to assess the participants. A separate information process- ing test, requiring allocation of cognitive resources, was also presented during driving. Finally, driving in real traffic was used to validate and evaluai-e the predictive value of the neuropsyeholcrgical test battery for driving pe~formance. Ratings of risk awareness were supposed to measure an overall driving-skill performance.

U7e hypothesized that people with brain lesion have decreased efficiency of cognitive prclccssing, which influences driving performance. We al:;ol assumed that neuropsychological tests measuring speed and accuracy of cognitive processing ccruld be used ELS a screening instrument for driving performance.

Patient Group

A sample was selected from patients who had been treated at the neurosurgical department of the Univer- sity Hospital of Linkoping between 1990 and 1993 and

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

LUNDQVIST ET AL.

who were diagnosed with traumatic head injury or subarachnoidal hemorrhage. Criteria for medically veri- fied brain lesion were neurological symptoms of more than 24 hr, visible damage or cerebral edema observed during surgery, or brain lesion verified by CT or MRI. To make sure that the investigated cognitive processes were not affected by other deficits, the following criteria for exclusion were used: alcohol abuse, mental retarda- tion, nonfluency in Swedish, hemianopsia, herniparesis, serious aphasia, and serious intellectual disturbances due to injury that made participation impossible. Twelve persans who did not have a driving license or were not driving were excluded. Six persons did not want to participate. Twenty-nine patients participated in the study, 16 suffering from traumatic brain injury and 13 suffering from subarachnoidal hemorrhage. Accord- ing to the CT or MEU, 9 patients had right hemisphere brain damage, 5 had left hemisphere damage, and 12 had bilateral brain damage. In 3 patients, the brain damage could not be localized. In 23 cases, the lesion was frontal or multiple. Mean age of participants was 44.1 years (SD = 11.7), The group was .made up of 17 (59%) men and 12 (41 %) women. Mean education level was 1 1.2 years (SD = 3.0). Eighty-three percent of the participants were employed or were students. The mean interval between injury or illness and this study was 3.7 years (range = 1.8-$3 years).

Control Group

Twenty-nine participants without documented neu- rological disorders or cerebral lesion served as controls. They were matched by age, gender, education, and driving experience (mileage per year). Mean age was 44.8 years (SD = 11.8). Mean education level was 12.2 years (SD = 2.8). Ninety-six percent of these partici- pants were employed or students. With respect to edu- cation and professic+n, the two groups were intellectually comparable. In both groups, 82% had their driving license for more than 10 years.

lyeurologicd screening. The neurological screen- ing was carried out by a specialist in rehabilitation medi- cine according to a protocol involving examination of distant vision, visual fields, ocular movement, muscle power, tendon reflexes, coordination, equilibrium, gait, sensation of pain, touch, and vibration.

Interview. The structured interview explored con- cerns about driving experience, changed driving behav- ior after injury or illness, cognitive disability affecting

driving, awareness about cognitive deficits not related to driving, and subjective opinion about driving ability.

Neuropsychological Examination

From a neuropsycholo~cal test battery used for clinical purposes, a reduced battery was selected, con- sisting of tests assessing pe:rceptual, cognitive, and ex- ecutive functions.

Paper and pen tests. Paper and pen tests were Trail Making Test, Part 3 (TMTB), Digit Symbol, Raven Progressive Matrices set I, and Rey Complex Figure Test (CFT) recall after 3 min. Most of these tests are well-known neuropsychological tests used in clini- cal practice with patients with brain injuries. Standard procedures and scoring were followed (Lezak, 1995).

Two tests presented on tape. Two tests were presented on tape. First was the Paced Auditory Serial Addition Test (PASAT), which was applied and scored in a conventional way (Lezak, 1995). PASAT requires the person to add 60 randomized digits so that each is added to the digit immediately preceding it (Gronwall & Wrightsoin, 1974). The digits in the current version were presented at the rate of one digit e.~:ry 1.6 sec. The other tape-recorded test was a modified ]Listening Span test (Baddeley, 1990; Daneman & Carpenter, 1980). The test consists of simple three-word sentences pre- sented in groups of three to six sentences. The task is to answer "yes" or "no" whether the statement is sensible or nonsense. Then the individual is requi:red to recall the first or the last word, without knowing which in ad- vance, in the set of sentences. The score was the total number of words correctl!~ recalled.

Computerized tests. The compilterized tests were Finger Tapping, K test, Reaction Time test, and Simultaneous Capacity test-all included in Automated Psychological Tests (Levander, 1988)-and the Wis- consin Card Sorting Test (WCST; Loong, 1990). The computerized Simultaneous Capacity test consisted of one background task running continuously, which re- quired the participant to identify three consecutive odd digits in a continuous stream of digits pi-esented one by one in the center of the screen, and one foreground task, which required the participant to respond to messages in the upper part of the screen by pressing 1 of the 10 numeric keys. An overall performance measure, the T score, was calculated from the mean of the background and the foreground task. In the K test, the task of the

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

NEUROPSYCHOLOGICAL ASPECTS OF DRIVING

participant was to decide as quickly as possible if the letter k was present in a set of 10 distractor characters (squares or letters) displa~yed at random positions on the screen. An overall performance measure (speed and accuracy) was calculatecl and expressed in T scores.

Simulator Driving

The Swedish Road and Transport Research Institute (VTI) driving simulator is an advanced simulator that consists of a moving-base system, a wide-angle (120") visual screen, a vibration-generating system, a sound- regulating system, and a temperature-regulating system (Nilsson, 1989; Nordn~ark, Jansson, Lidstrom, & Palmqvist, 1986, 1988). These five subsystems can be controlled to operate in a way that gives the driver a realistic dynamic impression of traffic scenes. The sub- systems developed by WTI are by far the fastest systems used in any simulator at present. The road type presented to the participants in the simulator was a two-lane asphalt road. It contained curves, oncoming traffic, a leading vehicle, speed-limit signs, and constant friction. The conditions corresponded to a dry road in summer and cloudy weather. The car body was an ordinary Saab 9000. Each participant was given a 20-km training session before the actual recording took place. The test route was 80 km. The pa.rticipants were asked to drive normally. The simulator had already been used in dif- ferent studies concerning driving behavior and cogni- tive demands (Alm & Nilsson, 1995).

The predictability of the driving situation was varied during driving. High predictability indicated good sight, straight road, and catching up to the leading vehicle, which drove at slightly varying speed. Low predict- ability indicated that the leading vehicle suddenly braked, a parked car suddenly appeared behind a curve, or a visual stimulus suddenly appeared on the screen. Here, the participant was expected to brake as quickly as possible.

A distracting task was presented three times during drivin~. When a mobile telephone was ringing, the participants answered by pushing a button on the panel. Then they listened to a modified Listening Span test (Baddeley, 1990; Baddeley, Logie, Nimmo-Smith, & Brereton, 1985; Daneman & Carpenter, 1980) similar to the one used in the neuropsychological examination. Dependent measures were speed, lateral position, and complex reaction time. Time to collision (TTC min) and distance to collision (DTC) reflect safety margins in time and distance. TTC min was the least time to colli- sion when the relative speed between the vehicles was

not changed. DTC was the distance to the leading vehi- cle at ITC min. Also, amount of words recalled at the Listening Span test was me.asured. After completing the test route, the participants rated the ~vorkload during driving (N.4SA-RTLX; Byers, Bittnel-, & Hill, 1989; Hart & Staveland, 1988).

On-Road Driving

On-road driving of 25 km was done iin actual traffic in the participant's own car. The evaluation was con- ducted by one of two certified driving inspectors (inter- rater reliability, r = .96,p < .01), who made ablind rating about to which group the participant belonged on the basis of five specific criteria: speed, maneuvering, lat- eral position, traffic behavior, and attention. Speed was evaluated in relation to current traffic demands. Maneu- vering was the handling anti control over the vehicle, for example, steering and braking. Lateral position was placing of the vehicle in the road lane, ro~ndabouts, and intersections. Attention concerned traffic signs, direc- tions, and other road user!;. Traffic behavior included planning the driving and adjusting to traffic rules and other road users. The scorirg was made on a rating scale ranging from 1 to 5, with a rating of 1 or 2 failing. Based on these criteria, an overall evaluation of risk awareness was carried out. The same inspector rated both the patient and the corresponding control person. The in- spe.ctors did not know abcwt the results of the neuro- psychological testing or about the simulator driving results. The route was selected to involve a variety of driving situations regarding action and locality (Figure 1). All individuals were driving under similar circum- stances with regard to traffic density, .weather, road conditions, and driving route.

Statistical Methods

All statistical analyses were made using SPSS for Windows (version 6.1; Norusis, 1993). liesults in the text and tables are given as mean values with one standard deviation (+.SD). For statistical tests, p I .O1 was considered significant where multiple t tests were used. Where single tests were carried out or where the tests were planned in advance, p I .05 was considered significant. To evaluate differences between groups, the t test (for independent samplles) was used for continuous variables, and Pearson x2 was used for categorized variables. A one-tailed t test was chosen when group differences were hypothesized. A stepwise logistic re-

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

LUNDQVIST ET AL.

Number Condition Evaluation

Traffic light Attention Roundabout or junction Positioning, attention Turning nght crossing Attention

cycle path Rural public road 50 Speed, positioning

km/hr, 70 km/hr Prohibition to turn left Attention, planning,

positioning Highway 90 kmthr Speed Residential area, Maneuvering,

turning in junction attention, speed Distraction, talking Attenuon

with instructor Participant given Planning

instruction, "Follow signs to. ." (e.g., "city." "Stockholm"

Figure 1. On-road driving route: Conditions and examples of tasks evaluated.

gression analysis was carried out to find out the tests that had the best predictive power with regard to differ- entiation of the patient and control groups, as well as the high and low risk awareness groups, respectively. In addition, separate stepwise multiple regression analyses were carried out for each of the five rating scales per- taining to the on-road performances, with the variables in the neuropsychological test battery as independent variables.

Factor analysis was used to find out whether the test variables reflect a smaller number of underlying factors. Factors with eigenvalues 2 1.00 (Kaiser's criterion) were considered as nontrivial factors. A varimax rota- tion of the factor solution was made, and the rotated loadings together with communalities are presented in the tables. Factor scores were saved according to the

regression method, which is based on standardized vari- ables. The factor scores were used to compare the patient and control groups and the high and low risk awareness. The factor scores were also used in logistic regression analyses to predict the most significant factor,

Results

Neurological Screening

Twenty-eight individuals went through the physical examination. Four bad definite signs of lesion in the central nervous system, such as raised muscle tone, hemianesthesia, or abnormalities in coordination or gait. Eight patients had minor abnormalities such as moderate side difference in the rapidity of complex, asymmetrical movements. Thus, 72% had no or minor neurological symptoms at the time of the study. No one had physical. disabilities considered to be of any practi- cal consequence for the ability to take part in the neuro- psychological tests or to drive a car.

Interview

Thirty-eight percent of the patients reported that driving was discussed by the doctor at discharge. On average, they had avoided driving for 4 months after brain damage. Generally, patients had reduced their driving. Still, however, 7695 were driving more than 5,000 krn per year at the time of the study. In both groups, 93% stated that they drive "as weH as, or better than people in common." Both patients and controls seldom or never experienced problems related to driv- ing. Among patients, 3 1 % avoided long-distance driv- ing compared with about 7% in the control group, X2(1) = 5.50, p < .05. Patients reported significantly more memory problems than the controls did, X2(l) = 8.34, p < .01.

Newropsycholsgical Examination

A factor analysis on the neuropsychological test bat- tery, including 14 measures, produced four factors, ac- counting for 74% of the total variance (Table 1). The WCST test variables loaded on Factor 1. PASAT, Si- multaneous Capacity test, Listening Span test, and CFT recall loaded on Factor 2. The K test, Complex Reaction Time, TMTB, and Digit Symbol loaded on Factor 3, and Simple Reaction Time and Finger Tapping loaded on Factor 4. Listening Span test and T m loaded greater than .SO on more than one factor. Both Factor 1 and

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

NEUROPSYCHOLOGICAL ASPECTS OF DRIVlNG

Factor 2 differed significantly between the patient and patient group, except for the estimate of simple and control groups, t(56) = 3.07, p < .01, and, t(56) = 3.94, complex reaction time. There was also a significant p < .01, respectively. difference between the grctups in tests where demands

Table 2 presents the results of patients and controls on speed are not so marked. on the neuropsychological test battery. In all tests, the In the subjective evaluation of performance on the control group performed significantly better than the co~nputerized tests, the patients rated h hems elves as

Table 1. Factor Analysis of fhe Neuropgchological Test Battery With Variimw Rotation

Factor - -

Test/Factor 1 2 3 4 Communality -

WCST Percentage Correct 0.93 0.19 0 18 0 11 0.95 WCST Total Categories Completed 0.88 0.09 0.18 0.13 0 83 WCST Percentage Perseverative Errors -0.84 -0 18 -0.41 -0.04 0 92 PASAT 1.6 0.06 0.81 0 16 0.12 0.70 Simultaneous Capacity, T Score 0.18 0.70 0.39 0.15 0 71 Listening Span Test 0.51 0.69 -0.08 -0.07 0.75 Rey CFT Recall 0.13 0.58 0.35 0.32 0.57 K Test, T Score 0.31 0.09 0.82 -0.06 0.78 Reaction Time, Complex, T Score 0.19 0.19 0.75 0.27 0 70 Trail Making Test B -0.14 -0.57 -0.63 -0.12 0.76 Digit Symbol 0.22 0.46 0.58 0.30 0.69 Reaction Time, Simple, T Sco~e -0.01 -0.Cb5 0 16 0.86 0.77 Finger Tapping, T Score 0.16 0.36 0.01 0.75 0.72

Eigenvalue 6.34 1.70 1.25 1.11 Percentage of Variance 45.3 12.2 9.0 8.0

Note: aumbers in bold represent factor loadings > 50. WCST = Wisconsin Card Sorting Test; PASAT = Paced Auditory Serial Addition Test; C f l = Complex Figure Test

Table 2. Neuropsychologzcal Test Data Means and Standard Deviations t Test for Independent Samples

Patientsa Controlsa t Test

Tests M SD M SD t - - --

Raven PM Set I (0-1 2) 8.7 2.1 10.2 2 0 2 83** Trail Making Test B (sec) 76 5 27.3 55.9 16.6 3.48** Digit Symbol (0-93) 51.7 12.4 60.1 11.0 2.73** Listening Span Test (0-54) 23 9 4.9 31.6 5 2 5.78** PASAT 1.6 (0-60) 36.3 10 6 45.4 8.8 3 58** Rey CFT Recall (0-36) 20.2 6.0 24.7 5.4 3.05** WCST (Total Categories Completed; 0-1 I) 3.4 2.1 5.0 1.6 3 42** WCST (Percentage Correct) 60 0 14.3 73.5 12.7 3.79** WCST (Percent Perseverative E m ) 20.9 9.9 13.6 7.1 3 79** Finger Tapping, T Score 51.6 8.4 57.0 6.7 2.70** Reaction Time, Simple, T Score 52.7 5.3 54.7 5.9 1.33 Reaction Time, Complex, T Score 52.2 10 8 55.8 4.6 1.67' K Test, T Score 52.2 9.4 57.2 4.0 2 62** Simultaneous Capacity Test, T Score 54.1 10 3 60.5 5.0 3.03**

Note. Raven PM = Raven Progressive Matrices; PASAT = Paced Au&tory Serial Addition Test; CFT = Complex Figu:e Test; WCST = Wisconsin Card Sorting Test "N = 29. *p 5.05. **p < .01, df= 56

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

LUNDQVIST ET AL.

slow compared with the controls, t(56) = 3.06, p < -01. The controls also rated their overall performance higher than the patients did, t(56) = 3.38, p < .01.

Simulator Driving

A factor analysis of all the simulator driving variables produced three factors, accounting for 76% of the total variance (Table 3). Factclr 1 accounted for 'ITC min and DTC in unpredictable situations, whereas Factor 2 ac- counted for ?TC min and DTC in predictable situations. Factor 3 accounted for Listening Span test and Complex Reaction Time. DTC in a predictable situation loaded greater than .50 on more than one factor.

Two patients and three controls had to stop the simu- lator driving because they did not feel well in the simu- lator. The patients and the controls did not differ with regard to speed and lateral position. In unpredictable situations, the patients had significantly longer TTC min, t(51) = 1.72, p = .C!5; also complex reaction time differed significantly between the groups, t(51) = 1.81, p = .05 (Table 4). The amount of produced words at the

Listening Span test during driving was significantly better for the control group, t(51) = 3 . 2 7 , ~ c .Ol. As for the subjective evaluation of driving in the simulator, the controls estimated the phq sical demands and effort, as well as time pressure as harder than ihe patients did: physical demands, t(51) = 2.10, p < .05; effort, t(51) = 2.07, p < .05; time pressure, t(51) = 2.71, p < .01.

On-Road Driving

Out of the five variables evaluated iin the on-road driving, attention, t(56) =-- 2.75, p < .01, and traffic behavior, t(56) = 2.56, p < .01, differed significantly between the two groups (Table 5). The scores in the control group were generally higher, but not signifi- cantly so for speed, maneuvering, and lateral position. The five scores together constituted the overall risk- awareness score. The traffic inspectors classified 12 patients (41%) as having lc~w risk awareness, whereas 3 controls received the same classification, X"l) = 13.52, p < .01. The risk-awareness score was nut dependent on driving experience or frequency.

Table 3. Factor Analysis ofSinzulator Driving Measures With Vaninax Rotatzon -

Factor

'ITC min. Unpredictable Situation Distance to Collision, at TTC min, Unprehctable Situation 'ITC min, Predictable Situation Distance to Collision, at TTC mln, Predictable Situation Listening Span, Words Complex Reaction Time for Visual Stimulus

Eigenvalue Percentage of Variance

Note: Numbers in bold represent factor loadings > 50. TTC min = time to collision.

Table 4. S~rn~~Iator Drtvtng Data Means and Standard DevtaDons, t Test for Independent Samples

- Patientsa controlsb t Test

-- Task M SD M SD t

Complex Reachon Time for Unpredictable Visual Stimuli (sec) 15 0.7 1.2 0 5 1.81* 'ITC min, Predictable Situation (sec) 3.1 1 . 1 2.7 0 9 1.30 Dlstance to Collision, at TTC min, Predictable Situation (m) 11.5 2.0 11.3 2.3 0 31 'ITC min, Unpredictable Situation (sec) 2.5 1 0 2.1 0.7 171* D~stance to Collision, at TTC min, Unpredictable Situation (m) 16.5 3.5 18.0 3.6 1.48 Listening Span Task, Words (0-15) 8.3 3.4 11.1 2.7 3.27**

Note Two-tailed t test was used for testing differences in predictable situations. 'ITC min =time to collision a ~ = 2 7 . b ~ = 2 6 * p 1 . 0 5 **p<.Ol ,d f=51 .

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

NEUROPSYCHOLOGICAL ASPECTS OF DRIVING

Table 5. On-Road Dnvzng Means and Standard Deviations, t Test for Independent Samples

- Patientss Control8 t Test --

Variable h4 SD M SD t - Speed 3.4 1.0 3.8 1.0 1.47 Maneuvering 3.6 0.7 3.9 0.9 1.46 Lateral Position 3.1 1.1 3 5 0.9 1.47 Traffic Behavior 2.9 0.9 3.5 1.0 2.56*

Attention 2.8 0.8 3.5 0.9 2.75*

"N = 29. * p < .01, df = 56.

Prediction Analyses

A logistic regression analysis using 14 neuro- psychological predictor variables was made. Hence, 86% of the cases were correctly classified with respect to group belongingness, with Listening Span and TMTB as the most potent tests (Model x2 = 33.1, p < .01), and the neuropsychological test battery accounted for 78% of the dichotomous outcome variable, risk awareness, with Si- multaneous Capacity test as the most potent test (Model X2 = 9.3, p < .01). Both analyses have in common that Factor 2 is the most sigmficant factor, needed for fast and efficient cognitive proct:ssing. Factor 2 also differed significantly between high and low risk awareness, t(56) = 2.61, p < .01. In the stepwise multiple regression analyses of each of the five rating scales pertaining to the on-road performances, Rs varied between .65 and .34, predicting lateral positiorc attention, and traffic behavior best. None of the tests in the neuropsychological test battery predicted any of the outcome variables of the simulator driving. Simulator driving outcome was not able to predict on-road driving outcomes.

Overall, the results suggest the following: the pa- tients performed signific;mtly worse on tests measuring perceptual, cognitive, and executive functions. They were performing at the same level as the controls in predictable simulator diving situations. In unpre- dictable situations, they had significantly longer reac- tion times to visual stimuli, longer TTC, and lower Listening Span performance during driving. The patient group showed significan1:ly less attention, worse traffic behavior, and less risk awareness in real traffic driving. The neuropsychological test battery could predict risk awareness with 78% conFidence.

Additional variables were studied in order to find out whether they were correlated with risk awareness. Neither gender, age, nor diagnosis or residual neurological symp- toms were significantly cctrrelated to risk awareness.

Discussion

There is a growing need for an easi~ly administered screening assessment of cognitive functions for driving after brain damage. The hypothesis addressed here was that people with brain lesion have decreased speed and accuracy of cognitive processing, which influences driving performance, especially at the tactical level of control. We also assumed that neurops~~chological tests measuring cognitive processing speed and accuracy are useful in screening driving performance. Twenty-nine patients with brain damage and 29 contrclls participated in rhe study. Eighty-two percent had driving experience of more than 10 years. Thl: patients were socially well recovered after cerebral lesions, with ia high rate of qmployment and a mean time of 3.5 pears after lesion. Colnsequently , the neuropsychological condition is sup- posed to be stable.

Models of driving emphasize the strategic level, where executive functions are salient. Speed of cogni- tive processing is necessary at the tactical level, and pe~ceptual and motor speed are needed at the opera- tional level.

Earlier research has focused mainly on the opera- tional level of control. When cholosing neuro- psychological tests for this study, the jntention was to include tests assumed to load on cogniti.ve processing speed and corresponding 110 the tacticid control level. Other tests were also included to provide a comparison with functions correspondling to strategic and opera- tional levels of control.

Neuropsychological Examination

Analysis of the test battcry showed that it contained four factors, with item loadings of .93 to .58, measuring executive capacity (Factor 1) corresponding to the stra- tegic level of control, co,gnitive capa.city (Factor 2) co~responding to the tactical level of cont.ro1, automatic attentional capacity (Factor 3), and simple perceptual- motor capacity (Factor 4), both probably corresponding to the operational level of control. The: results show a significant difference on most of the neuro- psychological test perforn~ances between the patient and control group. Accordi:tig to the resulits, the patients have deficits in attentional, cognitive, and executive functions, despite the long interval sinc:e their cerebral lesion.

.4t the level of factor analysis, however, only the executive and cognitive factors proved to be signifi- cantly different between the patient and control groups. Thus, the data suggest that controlled prclcessing is more

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

LUNDQVIST ET AL.

affected by brain lesion than automatic processing. These results are in accordance with earlier studies showing residual executive and information-processing impairment after brain damage (Hugenholtz, Stuss, Stethem, & Richard, 1988; Lezak, 1995; Stuss et al., 1985).

Sirnulatar Driving

The results showed significant differences between the groups concerning complex tasks; for example, complex reaction time for unpredictable visual stimuli. Also, in the Listening Span test, the difference already found in the neuropsychological testing was confirmed. No difference between the groups concerning safety margins was expected in the predictable situations, but contrary to our expectations, the patients had signifi- cantly longer TI% in unpredictable situations. It might imply that the patients sloweddown when getting closer to the vehicle in front of them. They presumably antici- pated the increased cognitive demands in getting closer to another car. Therefore, they compensated with an increased safety margin. Because there was no signifi- cant difference in DTC, the results could also imply that the patients were slower to brake, which is in line with their longer reaction time in unpredictable situations.

Because the simulator driving situation probably rep- resents a new situation, all participants compensated for insecurit~ by driving very carefully (Ranney, 1994). The design was comparable to a country road with low-density traffic. It might explain the generally low correlation between simulator driving measures and the on-road driving task, which contained bath city and rural areas. If the simulator design had been more com- plex, however, there would have been a risk of failing, which would have increased the number of dropouts from the study.

On-Road Driving

The on-road driving was the external criterion of driving performance. It involved dynamic demands and varying complexity. l l e patients performed as well as their controls in tasks where they could rely on automat- ized functions, measured by speed, lateral position, and maneuverinlg. This was also confirmed by speed and lateral position in the simulator driving and in simple and complex reaction time, measured with the neuro- psychological tests. Again, automatic processing is not necessarily affected by brain damage. Driving experi- ence probably leads to automatization of skills required at the operational level and should, therefore, be taken

into account. Difference in driving experience did not influence the driving performance in this study with participants having extensive driving experience.

Conclusions

1. Speed and accuracy of cognitive processing are often impaired after brain damage (Stuss et al., 1995). When it comes to complex tasks, our suggestions re- garding the patients' lower capacity of cognitive proc- essing were confirmed by the neuropsychological tests, Also, in the practical driving task, the patients per- formed significantly worse than the controls, with 41 % failing in attention and 38% failing in traffic behavior. Adjusting to the continuously changing traffic situation, corresponding to the tactical level of control, presum- ably requires high cognitive processing speed. In the simulator driving, the patients were also slower in com- plex tasks, and when the task load was increased by the Listening Span test, the patients did not manage to allocate processing resources to both tasks. The patients showed, however, compensatory behavior while driv- ing in the simulator. When there was a cognitive over- load by the Listening Span test, they gave priority to the driving task and in the unpredictable situation, they increased safety margins. Both compensation by giving priority to the driving task and increased caution are decisions made during driving at the tactical level.

2, Forty-one percent of the patients had a low risk awareness. Awareness of oneself and evaluating one's own driving eapacity can be seen as a kind of metacog- nition that is required at the strategic level. One can notice that the patients estimated their driving skill to be as good as the controls' was. Still, 41% of the patients failed the on-road driving test. The patients' own judg- ment of their driving ability seems to be unreliable. Most people, however, tend to overestimate their own driving skill and safety (Svenssan, 1481). In daily driving, the patients might compensate better when they decide themselves where to drive instead of getting directions from another person. People often compensate for irn- paired driving performance by making the right judg- ments and decisions at the! strategic and tactical levels. A significant number of the patients avoided long-dis- tance driving, which can also be seen as a decision made a1 the strategic level.

3. The neuropsycholo~cd test battery classified about 80% of the individuals correctly into patienz and control group andinto low or high risk awaieness gr~up, with Factor 2 as the most significant factor. The Simul- taneous Capacity test was the test with the highest predictive power, classifying low risk awareness with a

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

NEUROPSYCHOLOGICAL ASPECTS OF DRIVING

cutoff T score of less than 40 (referred to the norm data of the Automated Psychological Test battery, current manual). The test required speed of cognitive process- ing and divided attention, which supports our hypothe- sis that neuropsychological tests measuring speed and accuracy of cognitive prc~cessing are useful in screening driving performance. Awaiting further studies, how- ever, the results should be interpreted with caution given the small sample size.

4. Automatic processing is used to a great extent durinp driving. In perceiving a threat, there is a demand on allocation of attention to controlled processing, which is dependent on the anterior attentional system (Stuss et al., 1995). 'Re controls experienced more effort and time pressure in simulator driving, which suggests that they mobilized executive capacity to a larger extent than the patients did. The patients' lower executive function was also confirmed in the neuro- psychological testing.

References

Alm, H., & Nilsson, L. (1995) The effects of mobile telephone task on driver behavior in a car following situation. AccidentAnaly- sis and Prevention, 27, 797-71 5.

Baddeley, A D. (1990). Hurnan memory: Theoq and practzce. London Lawrence Erlbaum Associates, Inc.

Baddeley, A. D.,Logie, R., Nimmo-Smith, I.. & Brereton. N. (19853. Components of fluent reading. Journal of Memory and h- guuge, 24, 119-131.

Brooke, M. M., Questad, K. A., Patterson, D R.. & Valois. T A (1992). Driving evaluation after traumatic brain injury. Amen- can Journal of Physical Medzcine and Rehabllitatron, 71, 177-182.

Brooks, N. (1984). Cognitive Geficits after head injury. In N Brooks (Ed.), Closed head injury. Psychologrcal, soclal, and family consequences (pp. 44-73) New York: Oxford University Press.

Byers, J. C., Bittner, A. C., Jr., &Hill, S G. (1989) Traditional and raw task load index (TLX) correlations. Are paired compari- sons necessary? In A. Miyal (Ed.), Advances in zndustrial ergonornzcs and safety (Vol 1, pp. 481-485). London: Taylor & Francis.

Christie, N. (1996). Assessing drzvingfirness following bmuz znjuiy or zllness: A research review (TRL Rep. No 208). Crowthome, England: Transport and Road Research Laboratory.

Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19,450-466.

Filley, C . M (1995). Neurobehavioral aspects of cerebral white matter disorders In B. S Fogel, R. S. Schiffe, & S. M. Rao (Eds ), Neuropsychratiy A comprehensive textbook (pp 913-933). Baltimore Williams & Wilkins.

Fuller. R. A. (1984). Conceptualisations of driver behavlor as threat avoidance. Ergonomzcs, 27, 1139-1 155

Galski, T., Bruno, R. L., & Ehle, H T. (1993). Prediction of behind the wheel driving performance in patients with cerebral brain damage: A discriminant function analysis American Journal ofOccupatzona1 Therapy, 47, 391-396

Galski, T., Ehle, H. T., & Bruno, R. L. (1990) An assessment of measures to predict the outcome of driving evaluations in patients with cerebral damage. Amerrcan Journal of Occupa- rional Therapy, 44,709-7 1 3.

Gronwall, D.. QE Wrightson, P. r 1974) Delayed recovery of intel- lectual function after minor head injury. Luncet, 11, 605-609.

Hart, S. G., & Staveland, L. E (1988). Developn~ent of NASA-TLX (Task Load Index): Results of empirical and theoretical re- search. In P. A H a n m k & N. Meshkati (Eds.), Human mental workload (pp. 139-177). Amsterdam: Elsevier.

Hopewell, C. A., & Price, R. J. (1985). Driving after head injury Journal qf Clinical Experimental Neurops-vchology, 7, 148

Hugenholtz, PI.. Stuss, D. T., Stethem, L. L., 6r Richard, M. T. (1988). How long does it take to recover from a mild concus- sion? Neurosurgeq, 22,853-858

Levander, S. (1988). An automatl?dp~~chological test battery: IBM- PC verszon (APT-PC) (Research Reports froin the Department of Psychiatry & Behavioral Medicine, Uqiversity of Trond- heim, 11, No. 65). Trondheim, Norway: University of Trond- heim.

Levin, H. S., Grafman, J , & Eisenberg, H. M. (1 982). Neurobehav- ioral recover),fiom head i n j u ~ . New York: Oxford University Press.

Lezak, M. D. (1995) Neurops)..chologzcal assessment (3rd 4 . ) . New York. Oxford University Press.

Loong, J. (1990). Wiscoizsin Card Sorting Test [Neuropsychological Assessment software and user's manual]. Lutz. FL: Psycho- loscal Assessment Resourc:es.

Michon, J A (1985). A critica view of dnver behavior models. What do we know, what should we do? In L Evans & R. Schwing (Eds ), Human 5ehavzor and trafic safe5 (pp. 485-520). New York: Plenum

Naatanen, R., & Summala, H. (1976). Road user behaviorand ti-afic accidents New York. North-Holland.

Nilsson, L. (1989). The VTI d,.ivmng szmulator (DRIVE Project V1017 [BERTIE], Rep. No. 24). Linkoping, Sweden Swedlsh Road and Traffic Research [nstitute.

Nordmark, S., Jansson, H., Lidst~om, M , & Pal~~qvist, G. (1986). A movmg base dr~vmg simu!ator wrth wrde angle vzsual system (VTI 106A). Linkping, Sweden: Swedish Road and Traffic Research Institute.

Nordmark, S.. Jansson, H . Lidstrom, M., & Palrnqvist, G. (1988). The VTI driving simulator: Recent developments. Linkoping, Sweden: Swedish Road and Traffic Research Institute.

Norusis, M J (1993). SPSS for WindowsTM [Computer software] Chicago. Advanced Statistics.

Posner, M. I. (1990). Hierarch~cal distributed networks in the neuropsychology of selective attention. In A. ~Zaramazza (Ed.), Cognitive neuropsychology and neurolzng~nstics. Advances m nzodels of cognrtrve functzorz and rmpairment (pp 1 87-2 10) Hillsdale, NJ: Lawrence ErlSaum Associates, Inc.

Posner, M. I. (1995). Attention in cognitive neuroscience: An over- view. In M. S Gazzaniga (Ed.), The cognitzve neuroscrences (pp. 615-624). Cambridge, 8MA: MIT press.

Quigley, F. L , & De Lisa, J. A. (1983). Assessir~g dnving potential of CVA patients. American Journal of Occupatiotuzl Tlterapy, 37,474-478

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4

LUNDQVIST ET AL.

Ranney, T. A. (1994). Models of driving behavior. A review of their evolution. Accident Analysis and Prevention, 26, 733-750.

Shiffrin, R. M., & Schneider, W. (1977). Controlled and automatic human information processing II. Perceptual learning, auto- matic attending and a general theory. Psychokogical Review, 84,90-127

Stuss, D. T., Ely, P., Hugenholtz, H., Richard, M. T., LaRochelle. S., Poirier, C. A., &Bell: I. (1985). Subtle neuropsychological Geficits on patients with good recovery after closed headinjury. Neurosurgery, 17,41-47.

Stuss, D. T., Shallice, T., Alexander, M. P.; & Picton. T. W. (1995) A multidisciplinary approach to anterior attentional functions. Annals of the Ne7-c York Academy ojsciences, 769, 191-21 1.

Sundet, K., Goffen~, G.. &Horn, E. (1995) To drive or not to drive: Neuropsycho;logical assessment for driver's license among ptroke patients. Scandinavian Joumal of Psych~logy~ 36, 47-5 8.

Svensson. 0. (1981). Are we all kss risky and more skillful than our fellow drivers? Acta Psychologia, 47, 143-1 48.

Van Zomeren, A. H., & Brouwer, W. H. (1992). Assessment of attention. In J. R. Crawford, D. M. Parker, & W. W. Mc Kinlay (~ds.). A handbook of neuropsychological assessment (pp. 241-266). Hove, England: Lawrence Erlbaum Associales, Inc.

Van Zomeren? A. H., Brouwer, W. H., & Deelman, B G. (1984). ettentional deficits: The riddles of selectivity, speed and alert-

ness In R;. Brooks (Ed.), Closed head injury: Pq~choIog~cal, soczal and f a t l y consequences (pp. 74-107). New York: Oxford University Press.

\'an Zomeren. A. H., Brouwer, W. H., & Minderhoud, J. M. (1987). Acquired brain damage and driving: A review. Archives of Physical Medicine and Rehabilitation, 68, 697-705.

Van Zomeren, A. H., Brouwer, W. H., Rothengatter, J. A., & Snoek, J. (1988). Fitness to drive a car after recovery from severe head injury. Archives of Physical Medicine and Rehabilitation, 69, 90-96.

Wickens, C D. (1984). Processing resources in attention. In R. Parasuraman-Br D. R. Davies (Eds.), Varieties of attention (pp. 63-102). San Diego, CA: Academic.

Wickens, C. D. (1992). Engineerin$ psychology and human per- formance (2nd ed.). New York: HqerCoIiins.

Wilde, G. J. S. (1982). The theory of risk homeostasis: Implications for safety, and health risk analysis. Risk Analysis, 2, 209-225.

Original submission April 18, 1997 Accepted July 21, 1997

Dow

nloa

ded

by [

Lin

köpi

ng U

nive

rsity

Lib

rary

] at

04:

55 1

2 A

ugus

t 201

4