Patient samples consist of individuals with brain injury.
1988a n = 92
1998b n = 121
1989 n = 175
1990 n = 215
C = 41 |
Cognitive Behavioral Driver’s Inventory
10 Neuropsychological tests:
-Attention.
-Concentration.
-Rapid decision making.
-Stimulus discrimination/ response differentiation.
-Visual scanning and acuity.
-Attention shifting.
Yields 27 response measures.
On-Road test
Criterion measure. |
1988a
Reliability
Good internal consistency (Cronbach’s alpha = 0.949).
Validity
(based on General Driver’s Index and on-road performance as criterion measure).
Of 44 passing the CBDI, 42 passed road test.
Of 48 failing the CBDI, only 6 were allowed to attempt the road test and all failed.
1988b
Normative tables derived from 121 brain-injured patients provided, complete with decision-making rules.
1989
Double blind validity study.
Validity
(CBDI and on-road measures).
Of 42 passing the CBDI, 40 passed road test.
However, of 39 patients failing the CBDI, 7 passed the road test.
1990
109 brain injured passed the road test and 54 failed.
118 brain injured passed the CBDI and 97 failed.
CBDI was sensitive to discriminating between controls, brain injured who passed, and those who failed road test. Importantly, age was a confounding factor. |
CVA = 23
THI =14 |
Physical and Neuropsychological Tests (21 tests)
-Attention.
-Concentration.
-Reaction time.
-Memory.
-Visual acuity.
-Visuospatial skills.
Scores converted to pass/fail by driver evaluator (criteria not defined).
On-Road test
-Consisted of 26 tasks believed to require an integration of basic driving skills.
-Pass/fail if sufficient skill demonstrated to driving evaluator (criteria not specified). |
Physical and Neuropsychological tests
Only 4 of the tests predicted pre-driver evaluation outcome (Benton Visual Retention Test, cancellation test, visual acuity measure, and observations of inattention).
Neither the pre-driver evaluation outcome nor any of the pre-driver evaluation tests predicted on-road evaluation outcome.
On-Road Test
Only six of the 26 on-road measures correlated with on-road evaluation outcome (caution, backing up in parking lot, on highway, parking on grade, lane use, and evaluating right of way).
Conclusion
Pre-driver evaluations must change from an attempt to measure abilities assumed to predict driving to an effort to screen out patients who are unsafe behind the wheel.
Items used in driving evaluation, although high in face validity, were low in predictive validity.
Test battery accounted for very small percentage of variance.
Authors call for research involving the empirical evaluation of tests used in pre-driver evaluations and on-road |
n = 13 (CVA)
n = 21 (THI)
Time since injury of stroke .08 to 17 years (mean = 1.8 + 3.6 years |
Neuropsychological Testing Battery
-sensory input.
-scanning and attention.
-calculation and construction.
-general and specific driving knowledge tests.
-resident diagnostic program(executive functions, awareness of deficits, etc.).
-integration (from simulator and on-road testing- seat and mirror adjustment, signaling, steering and tracking, etc.).
Doron Simulator
-view and react to films from Doron Driving Analyzer (e.g., Threat Recognition and Crash Avoidance).
-outcome measures were errors in braking and steering to escape danger or avoid disasters.
On-Road Evaluation
-criteria for ratings operationally defined prior to study.
-pass/fail ratings used to assess performance on individual measures.
-critical behaviors (impulsivity, distractability, confusion, anxiety,inattention, slowness, following directions, evaluation) scored as present or absent. |
Behind the wheel performance used as the criterion of fitness to drive.
Utility of measures determined by the amount of variance of the behind the wheel evaluation (street component)explained by a) the pre-driver evaluation, b) the simulator evaluation, and c) parking lot driving scores.
Pre-driver Evaluations
(Neuropsychological Tests) 64 percent of the behind-the wheel evaluation performance explained by tests that measured visual perception, visuomotor coordination, visuoconstructive abilities (planning, organizing, and executing test operations), and scanning and attention(selective and sustained).
Doron Simulator
Variance explained by simulator independently accounted for 63 percent of the variance.
When simulator measures combined with pre-driver evaluation, the best simulator items enhanced the predictive ability of pre-driver evaluation by only 6 percent.
Most of the simulator measures (e.g., braking, reaction time, steering, and acceleration) were ineffective predictors of behind-the wheel performance. Only two simulator items were significant predictors of driving outcome: appropriate use of signals on an introductory film and calculated percentage of valid attempts to steer out of potentially hazardous situations.
Parking Lot Driving Scores
Behind the wheel lot behaviors (e.g., following directions, slow response, inattention, distractability and lot index [lot behaviors scored and ranked, and then summed]) accounted for an additional 14 percent and 9 percent respectively of the outcome variance (raising the level of explained variance to 93 percent). |
n = 48 (CVA)
n = 58 (TBI) |
Neuropsychological Testing Battery, Doron Simulator, and On-road Performance
(see details in Galski et al., 1992). |
Study designed to determine the discriminative power and measurements of sensitivity of the battery described in 1992 study.
Results
Methods of evaluation sensitive in predicting outcome: off-road and on-road sensitivities of 90 percent and 92 percent with the inclusion of behavioral measures were obtained. Importantly, results revealed that residual deficits in cognition per se did not render a person unfit to drive. The research underscores the importance of considering behaviors in determining fitness-to-drive. |
n = 106
(CVA and THI) |
Neuropsychological Battery and Doron Simulator measures
(see Galski et al., 1992). |
Study objective was to determine the underlying factors of psychomotor testing and simulator evaluations useful in assessments of fitness-to-drive. Factor analysis identified 5 factors, accounting for 66 percent of the variance:
1. Higher Order Visuospatial Abilities.
2. Basic Visual Recognition and Responding.
3. Anticipatory Braking.
4. Defensive Steering.
5. Behavioral Manifestations of Complex Attention. More research needed to identify other relevant measures, as 34 percent of the variance remains unexplained. |
LH = 20
RH = 20
Mean time since stroke 33.3 + 40.7 months |
Cognitive Assessment
See battery in Nouri et al.(1987).
Tasks excluded in this study were Choice Reaction test, Stereodepth Perception test, and the Hand Sequencing Task.
Road Test
Rated on 26 items (Correct or Fault) and graded into one of three categories: Pass, Borderline, and Fail (based on overall subjective impression). Different driving instructor and different route than 1987 study. |
Purpose of study was to validate the cognitive battery developed in 1987 study.
Results of Road Test
Pass = 12
Borderline = 8
Fail = 20
Unable to validate equation from 1987 study because results of road tests between studies so discrepant (majority passed in 1987 study and majority failed in present study-may be due to change in instructor, route, or difference in severity of patient).
New algorithm developed based on three tests: Dot Cancellation, What Else Is In The Square?, and Road Sign Recognition Test (Stroke Drivers Screening Assessment).
Predictive values of algorithm ranged from 79 to 82 percent correct classification. No further details of classification provided. |
CVA = 27
(given SDSA) compared with general practitioner ratings for CVA = 25 |
All subjects given an on-road evaluation (criteria not defined).
27 subjects were given the Stroke Drivers Screening Assessment (SDSA).
27 subjects sought the advice of their General Practitioner re: driving fitness. |
Comparison of predictive value of the SDSA and Advice from General Practitioner based on clinical assessment to results from road test.
Results reveal that SDSA is better than General Practitioner assessment at predicting on-road performance. However, neither assessment is very accurate. For example, physicians misclassified 44 percent of the patients based on on-road criteria. The SDSA misclassified ~19 percent of the patients based on on-road criteria. |
