Visual field defects and accident risk, a retrospective study

1
Visual field defects and accident risk, a
retrospective study

• Gunnar D. Jenssen
• SINTEF

2
Scope

• Evaluate how drivers passing the SINTEF simulator
  test the last 8 years have performed as drivers
  (accident history).
• The SINTEF database includes prior medical
  evaluations, neuropsychological evaluations and
  on-road test results.
• Find out which method, combination of methods and
  set out outcome parameters within each method,
  that is most predictive of driving ability and
  the associated risk for drivers with visual field
  defects.

3
Previous studies

• Johnson Keltner (1983)
• performed an automated visual field screening of
  10000 volunteers.
• accident and conviction rates were more than
  twice as high as the control group.
• Results of subjects with visual field loss in one
  eye did not differ from their controls.
• Elvik, Mysen Vaa (1997)
• A reduction of the active visual field with more
  than 40 increases the accident rate
  dramatically.
• It is important to note that these results are
  for elderly drivers. Many of these elderly
  drivers had additional mental impairments (e.g.
  Alzheimer) which can increase accident risk in
  itself.

4
Previous studies (accidents behaviour)

• Coeckelbergh, (2001).
• Severe visual field defects have an impact on
  safety, and less severe visual field defects
  result in slower driving but otherwise unimpaired
  driving performance.
• Johnson Keltner, (1983)
• Severe defects result in impaired driving
  performance defined by conviction and accident
  rates
• Wood Troutbeck, (1992)
• Severe defects gave more manoeuvring errors,
  obstacle avoidance errors or less peripheral
  awareness.
• Szlyk, Seiple Viana (1995)
• Reported that degree of central fields loss
  marginally predicted accidents.
• Szlyk et al, (1992).
• Assessment of driving performance in patients
  with retinitis pigmentosa and visual field
  defects demonstrate an increase in both accidents
  and driving performance errors in a simulator

5
Compensation - (Pro)

• Szlyk et al, 1995 Wood Troutbeck, 1992, 1994
  
• It is assumed that persons with less severe field
  defects have no impact on accidents because of
  compensatory mechanisms . These strategies can
  allow safe driving despite vision impairments.
• Szlyk et al, 1995
• Compensatory mechanisms may be speed reduction,
  reduced risk-taking and
• Lövsund et al, 1991 Szlyk et al, 1995
• Increase of eye movements

6
Compensation - (Con)

• Hedin, (2001) Individuals with visual field
  defects usually deny having driving problems and
  argue that this is due to compensation with
  frequent eye and head movements
• Hedin Lövsund (1987) concluded that they
  normally did not compensate. The simulator study
  compared drivers with visual field defects with
  normal drivers. The results indicate an increase
  in detection time when objects were presented in
  the blind areas. And in some cases the object was
  not discovered at all.
• Lövsund, Hedin Törnros (1991) reported that
  only 4 out of 31 subjects with visual field
  defects compensated. The study was based on a
  simulator study. The results indicate that the
  subjects who compensate concentrate much more
  fixations to the affected side of their field
  while driving.

7
Purpose

• Provide evidence of accident risk associated with
  visual field defects.
• Establish better decision criteria for doctors in
  licensing assessment of drivers with visual field
  defects
• Assess the effectiveness of screening devices to
  test for visual field defects.
• Support future development of policy governing
  driver impairment with visual field defects.

8
Method

• Inclusion criteria
• Tested in SINTEF simulator.
• Documentation of prior test results
• Goldman perimeter, medical evaluation,
  neuropsychological evaluation as well as on-road
  testing in traffic.
• Exclusion criteria
• Subjects showing indication of dementia or
  spatial hemi-neglect are excluded.

9
Method Subjective and objective measures

• QuestionnairesHealth, travel habits,,
  accidents, self perceived driving ability,
  driving problems, compensatory strategies
• In-depth interviewsTo confirm answers and the
  course of possible accidents in order to reveal
  accident factors, which may be attributed to the
  visual field defect or insufficient compensatory
  behavior.
• Insurance reportsobtained to support analysis
  and confirm self reported accident history of the
  selected drivers.

10
MethodSubjective and objective measures

• In addition all the 50 drivers were subject to a
  new
• Medical examination
• Optometrical examination (Goldman Ring
  perimeter etc)
• Test of visual functioning (acuity, contrast
  sensitivity, glare, etc)
• Neuropsychological tests
• Purpose
• To document their current status,
• Check for possible changes in visual perceptual
  functions
• Control for conditions possibly affecting
  accident risk in the period between test and
  retest.

11
Simulator

• SINTEF has developed a simulator test similar to
  the Attentional Field of View method (AFOV)
  which assess the drivers ability to detect
  briefly presented peripheral targets (traffic
  sign symbols) in the presence of a distracters
  and a central manoeuvring related task.
• The simulator method allows free head movement
  and thus compensatory viewing behaviour not
  possible with traditional visual field testing
  (Goldman Perimeter or Useful Field of View test).

12
The SINTEF video Simulator with graphic overlay
symbols
13
Response buttons
14
Diagnosis and recommendation for drivers tested
in the SINTEF driving simulator
15
The visual system and associated visual field
defects
16
Sample

• Sample from SINTEF database
• 203 Total number fitness to drive 1992-2000,
• 98 Visual disturbance, without neglect and
  cognitive failure
• 34 Not recommended
• 62 Permission to drive

• Response rate
• Potential participants 62
• 4 deceased
• 8 unknown address
• Consent to study 50
• 2 Illness (4)
• 8 work (16)
• 5 family (10)
• 35 participants (70)

17
Sample

• N Age SD Sex
• Group 1 35 49 yrs 17 yrs 89 Male - 11 Female
• Exempted
• Licence
• Group 2 12 57 yrs 13 yrs 83 Male - 17
  Female
• Not approved
• drivers
• Control group 143 47 yrs 25 yrs 52 Male - 48
  Female
• Normal health

18
Graphic sign symbols projected on the video in
six different quadrants
19
Neuropsychological Tests

• Trail Making Test A B
• Similarities (item from WAIS)
• Block Design (item from WAIS)
• Wisconsin Card Sorting Test (WCST)
• CALCAP

20
Driving Test under Supervision

• Motor Skills Ability to handle interior
  interface
• Movement Tempo Pace of actions
• Observational SkillsSituation awareness and
  attention
• Perceptual Skills Detection of stimuli
• Decision Making Skills Making and executing
  correct judgements
• Vehicle HandlingNavigating chosen course -
  adapting vehicle position

• Speed Regulation Adapting speed to environment
• Distinctness Communicating own driving behaviour
  to other vehicles
• Risk Perception Ability to perceive, evaluate
  and handle risk
• Empathic Reasoning Ability to take other drivers
  perspective in a situation
• Self Monitoring Insight and management in own
  strengths and weak points

21
Visual Status

• Goldmann Perimeter Test
• Ophthimus automatic Ring perimeter.
• Psycho-physical assessments
• Ophtalmological and optometric evaluations

22
(No Transcript)
23
(No Transcript)
24
Exposure
25
Travel exposure

• Light
• Daylight Darkness
• 62 38
• Roads
• Rural roads Urban roads
• 39 61
• Season
• Summer Winter
• 55 45

26
Travel habits
27
Comparison of travel patterns by season
28
Visual acuity

• The monocular and binocular visual acuity data
  were for most subjects within normal values. One
  subject had a maximum acuity of 0.12 at 40 cd/m2,
  the rest had acuity's above 0.6.

29
Maximum contrast sensitivity (MCS) and integrated
contrast sensitivity (ICS) were for most subjects
within the normal range.
30
Reaction time for different groups by visual
square
31
Driver performance

• The group of drivers not approved for holding a
  licence has longer reaction times than the driver
  with exempted licence
• Control group has shorter reaction times than the
  two patient groups.
• Drivers with exemption lay closer to the
  performance of the control group.

32
Driver Performance Neuropsychological Tests

• Group 1 consisting of drivers exempted licence
  was divided into 2 clusters by way of K-Means
  Cluster analysis of results on neuropsychological
  tests.
• Cluster A performed better on neuropsychological
  tests than Cluster B (Fgt6.9, plt.05).
• Comparing the two groups in terms of their driver
  performance showed no significant difference in
  reaction times.

33
Driver performance - Age and gender

• The effect of age and gender on reaction times in
  each square was investigated by way of a General
  Linear Model (GLM).
• The results indicated that gender did not have en
  effect on reaction times in any of the squares
  (Flt1.8, pgt.05).
• Age proved to have an effect on all squares
  (Fgt5.6, plt.05), except upper left (F2.775,
  pgt.05) and lower right (F.758, pgt.1).

34
What others think about my driving?
35
Self reported accidents
36
Accident risk - Insurance data

• Observed number of accidents
• Visual problems 15
• Norm data 25
• Observed number per 100.000 km
• Visual problems 0.89
• Norm data 1.47
• plt0.5, two-tailed t-test

37
Visual Defects
38
(No Transcript)
39
(No Transcript)
40
Defects
41
Compensation strategies
42
Conclusions

• The results of this retrospective study of
  accident involvement of 35 visually impaired
  drivers exposed to normal traffic over a period
  varying from 3-8 years, shows that the selected
  subjects, drivers with visual field defects

43
Conclusions

• Drivers with visual field defects (Homonym
  Hemianopia, Quadrantanopia Scotoma) do not
  differ significantly from the norm group in terms
  of insurance reported accidents or self reported
  accidents.
• They do not compensate for their visual field
  defects by limiting their traffic exposure.
• The group of drivers not approved for holding a
  licence has longer reaction times than the driver
  with exempted licence, and the control group has
  shorter reaction times than the two patient
  groups. But drivers with exemption lay closer to
  the performance of the control group.
• Drivers not approved have trouble compensating
  for their visual impairment in the driving
  simulator test.
• Drivers approved do compensate well for their
  visual impairment and have developed several
  conscious compensation strategies.
• Drivers approved are conscious about the
  consequences of their visual impairment and are
  motivated to drive safely.

44
Conclusions

• This study shows drivers with even quite severe
  visual field defects e.g. homonymous heminopsia
  can drive with the same risk level as age, gender
  mileage and geographical area matched norm
  drivers without limiting neither mileage nor
  traffic exposure to potentially challenging
  traffic environments and traffic conditions.

45
Conclusions

• This study supports previous findings that
  drivers with visual field defects may develop
  sufficient compensatory strategies and behaviour
  to drive safely (Coeckelbergh, 2001 Szlyk et al,
  1995 Wood Troutbeck, 1992, 1994,)

46
The results from the SINTEF driving simulator,
on road trials and interviews indicate that
patients with visual field defects may
consciously or unconsciously compensate with

• Active search strategies and recheck in "blind"
  area before driving on.
• Use of mirrors
• Increased attention and situational awarenessb
• Observational Skills (situation awareness and
  attention)
• Perceptual Skills (detection of stimuli),
• Decision Making Skills (making and executing
  correct judgements)
• Vehicle Handling (navigating chosen course and
  adaptng vehicle position)
• Distinctness (communicating own driving behaviour
  to other vehicles)
• Risk Perception (ability to perceive, evaluate
  and handle risk)
• Empathic Reasoning (ability to take other drivers
  perspective in a situation)
• Self Monitoring (insight and management in own
  strengths and weak points)

47
Conclusions

• Driving is a complex, multilevel task
• Neuropsychological tests tap into more basic and
  isolated mental capacities,
• The task in the driving simulator is a more
  complex and compounded one.
• Driving is not affected directly by the status of
  any specific ability (e.g. visual field defects,
  cognitive deficits, mental processing), but more
  an expression of a holistic organization of
  separate skills and abilities into a goal
  directed behaviour.

48
Conclusions

• Assessment of driver skills cannot be concluded
  from neuropsychological tests or reaction times
  in simulation alone, rather, an overall appraisal
  seems more appropriate.
• Yet, some extreme scores on neuropsychological
  tests may be meaningful as contraindications of
  driver abilities (e.g. severe mental retardation,
  major defects of memory).
• The challenge in this sense will be to identify
  cut off scores for when the patient in question
  is not suited for driving. The main issue here is
  that the cut off score should reflect when the
  patient is not able to perform organizing and
  compensation strategies to outweigh his or hers
  defects, rather than driving abilities as such.

49
Conclusions

• Weakness
• sample size
• Strength
• Response rate
• Multilevel study
• Control

50
Acknowledgements
The study was funded by a grant from the
Swedish Public Roads Administration Project
partners Per Fosse, Low Vision Therapist,
Tambartun National Resource Centre of the
Visually Impaired (TRNC), Melhus. Håvard
Arnljot, Medical Officer, Region Hospital of
Sollefteå. Christian Westad, Optometrist,
Optiker H.Iversen, Trondheim. Hanne Simensen,
Neuropsychologist, Munkvoll Rehabilitation
Centre, St. Olav Hospital, Trondheim. Kenneth
Nøstberg, Driving Instructor, KMS Traffic
Education Centre, Trondheim.