Abiodun Emmanuel Akinwuntan,
Hilde Feys, Willy De Weerdt
Department of Rehabilitation
Sciences
Faculty of Physical
Education and Physiotherapy
Katholieke Universiteit
Leuven
Belgium
Jan Pauwels
Department of Kinesiology
Katholieke Universiteit
Leuven
Belgium
Guido Baten, Emmanuel
Strypstein
Brussels
Belgium
Summary: Objective: To identify from a
predriving assessment the tests that best predict the driving ability of stroke
patients. Design: Examination of records of 104 first ever stroke patients who
visited the Belgian Institute for Road Safety in 1998 and 1999 and performed
the medical, visual, perceptual and the ‘on the road’ assessments. Measures: Variables such as age, sex,
side of lesion, driving experience, visual tests, neuropsychological
assessments most of which are from the Test for Attentional Performance (TAP)
battery and an ‘on the road’ test were considered. Based on performances of
subjects on the tests, a suitable, not immediately suitable or not suitable to
drive decision was jointly taken by the team of assessors. Results: Following a
logistic regression analysis, a combination of side of lesion, kinetic vision,
visual scanning and the ‘on the road’ tests led to the best model in predicting
the final group decision (R2=0.53). The ‘on the road’ test was the
most important determinant (R2=0.42). Conclusion: The
predictive accuracy of the decision
is moderate. Inclusion of assessments more closely related to real road
situations should be considered.
INTRODUCTION
Stroke was reported to be the third major cause of
death and a major source of disability in the United States (Duncan et al,
1992). The incidence of primary stroke in Britain is around 2 per 1000
population per year, a quarter occurring in those aged under 65 years (Bramford
et al, 1988). In Belgium, the incidence of stroke is estimated at 2.62 per 1000
population annually (Van Casteren et al, 1988). A neurologic event such as
stroke affects driving ability (Mazer et al, 1998). Victims present various
physical and neuropsychological impairments such as motor, speech, visual,
cognitive and perceptual problems of which some represent definite symptoms
contrary to driving (Sundet et al, 1995). A well-configured predriving
assessment should identify stroke patients with potential to resume safe
driving (Kumar et al, 1991). The aim of this study was to identify the variable
or combination of variables from the predriving assessment as performed in
Belgium, which best predict the decision of driving ability in stroke patients.
DESIGN
The records of all patients who visited the CARA
section of the Belgian Institute for Road Safety in the year 1998 and 1999 for the first time were reviewed.
Patients were selected if they had a history of first stroke, possessed a valid
driver’s license and had been driving before the stroke onset. Patients with
visual field problems before October 1 1998 were excluded from the study because
the law in Belgium (Van Den Meerschaut, 1998), until 1st of October
1998, forbade patients with a visual field problem such as hemianopia after
brain damage to resume driving. In all, one hundred and four subjects were
included in the study.
MEASURES
Variables such as age, sex, side of lesion, driving
experience, visual acuity, kinetic vision, stereoscopy, neuropsychological
assessments most of which are from the Test for Attentional Performance (TAP)
battery and an ‘on the road’ test were considered. Based on the performances of
subjects in the tests, a suitable, not immediately suitable or not suitable to
drive decision was jointly taken by a team of professional assessors. The team
was made up of neurologists, occupational therapists and neuropsychologists in
the CARA (Center for determination of fitness to drive and car adaptations)
unit of the Belgian Institute for Road Safety.
RESULTS
Analysis of general data revealed that the variables of age, interval
between onset of stroke and examination, driving experience and visual field
problem (p<0.005) were
significantly different between the three groups of decision. All aspects of
the visual test battery and most variables of the neuropsychological test (p-values varying between 0.04 to
<0.0001) were also significantly different between the three groups.
Further comparisons revealed that the variables of age and driving
experience (Tukey/Kramer test, p<0.05)
and interval between onset of stroke and examination (Wilcoxon Rank Sum test
with Bonferroni correction, p<0.0167)
differentiated between the subjects found either suitable or not immediately
suitable to drive from subjects found not suitable to drive. For all
neuropsychological tests, similar results were found i.e. significant
differences were found between the patients who were either suitable to drive
or not immediately suitable to drive and those found not suitable to drive
(Tukey/Kramer test, p<0.05 and
Wilcoxon Rank Sum test with Bonferroni correction, p<0.0167). The visual tests with the exception of acuity of the
left and right eyes, differentiated only between subjects found suitable to
drive versus those not suitable to drive (Wilcoxon Rank Sum test with
Bonferroni correction, p<0.0167).
The on the road test differentiated between all the three groups (Tukey/Kramer
test, p<0.05). Subjects found
suitable to drive performed markedly better compared to the other two
categories (not immediately suitable to drive and not suitable to drive).
Relationships between the predriving assessment and the decision to
drive as well as the on the road test were further explored using correlation
coefficients. General parameters with the exceptions of sex, side of lesion and
aphasia had small but significant associations with the group decision (rs
varying between -0.24 and -0.36). Correlation coefficients of all items of the
visual test battery were found to be significant. Kinetic vision showed the
highest association with group decision (rs=0.43). Most of the
neuropsychological tests correlated significantly with the decision to drive.
The on the road driving test showed the highest association with the final
group decision (rs=-0.67).
Following a logistic regression analysis of 93
complete data sets, a combination of side of lesion, kinetic vision, visual scanning
and the ‘on the road’ test led to the best model in predicting the final group
decision (R2 = 0.53). The on the road driving test was the most
important determinant (R2 = 0.42). The odds ratio indicates that
subjects who improve by five points in the on the road test are two times more
likely to be found suitable to drive.
Table 1: Selected model by logistic regression in predicting the final
group decision
|
Variables |
Parameter estimate |
Standard error |
Wald Chi-square |
P - value |
Odds ratio |
|
Side of lesion Kinetic vision Scanning On the road test (5 units) |
-1.23 0.81 -0.16 -0.14 |
0.52 0.32 0.08 0.03 |
5.55 6.51 4.15 28.55 |
0.02 0.01 0.04 0.00 |
0.29 2.25 0.85 2.02 |
N=93, R2=0.53
DISCUSSION
Many variables had individual predictive values for the final group decision
as shown by the comparison of means, medians or frequencies and the correlation
coefficients. Further examination of these results showed that the
neuropsychological tests differentiated mainly between those found not suitable
and the two other categories. On the other hand, the subjects found suitable to
drive performed much better in the on the road test. These findings suggest
that subjects in the not immediately suitable to drive category performed well
in the neuropsychological tests but rather bad in the on the road driving test.
This was probably due to poor compensatory mechanisms during driving.
CONCLUSION
The on the road driving test was the most important
determinant of the decision to drive. The predictive accuracy of the decision
to drive was modest. It was therefore suggested that other neuropsychological
tests more closely related to road situations be included as part of the
pre-driving assessment. A training program for subjects in the not immediately
suitable to drive category aimed at teaching compensatory techniques during
driving is currently been developed.
The primary author expresses profound appreciation to the Belgian
Institute for Road Safety, Brussels, Belgium particularly the CARA section, for
funding support to participate at the conference. The author also acknowledges
the Interfaculty Council for Development Co-operation Scholarships Programme of
the Katholieke Unversiteit Leuven, Belgium for funding support of his doctoral
education.
REFERENCES
1. Bramford J, Sandercock P, Dennis M. A prospective study of acute cerebrovascular disease in the community: the oxfordshire community stroke project. 1. Methodology, demography and incident cases of first-ever stroke. J Neurol Neurosurg Psychiatry 1988;51:1373-80.
2. Duncan PW, Goldstein LB, Matchar D, Divine GW, Feussner J. Measurement of motor recovery after stroke: outcome assessment and sample size requirements. Stroke 1992;23:1084-9.
3.
Kumar
R, Powell B, Tani N, Naliboff B, Metter EJ. Perceptual dysfunction in hemiplegia
and automobile driving. The Gereontologist 1991;31:807-10.
4.
Mazer
BL, Korner-Bitensky NA, Sofer S. Predicting ability to drive after stroke. Arch
Phys Med Rehabil 1998;79:743-50.
5.
Sundet
K, Goffeng L, Hofft E. To drive or not to drive: neuropsychological assessment
for driver’s license among stroke patients. Scan J of Psych 1995;36:47-58.
6. Van Casteren V, Stroobant A, Lobet MP, Cornelis R. Cerebrovasculaire accidenten in België: een epidemiologische studie. Tijdschrift voor geneeskunde 1988;44:1065-70.
7.
Van
Den Meerschaut C. Rijgeschiktheid. Belgium: Semper October 1998;226:1-35.