Purchase this article with an account.
C. Rothlander, E. Papageorgiou, G. Hardiess, H. Wiethoelter, F. Schaeffel, H.-O. Karnath, H. Mallot, R. Vonthein, B. Schoenfisch, U. Schiefer; Size of Homonymous Scotomas and Traffic Accidents in Virtual Reality Driving. Invest. Ophthalmol. Vis. Sci. 2007;48(13):939.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess (i) the performance of patients with homonymous visual field defects (HVFDs) and normal-sighted control subjects in a driving simulator, and (ii) factors predicting the frequency of traffic accidents.
Twenty-nine patients with homonymous visual field defects due to cerebro-vascular lesions (mean age: 46.9 years, SD + 15.3 years) and 15 control subjects (mean age 42.4 years, SD + 14.3 years) were tested in a driving simulator. Absolute HVFDs were assessed by binocular semi-automated kinetic perimetry within the 90° visual field (stimulus III4e, angular velocity 3°/s, background luminance 10 cd/m²) using the Octopus 101 Perimeter (Haag-Streit Inc., Koeniz Switzerland) and were characterized by the area of sparing within the affected hemifield (AS). Two traffic density levels were presented in the driving simulator: 50% and 75%, each representing the crash probability when driving without any visual input (i.e. with closed eyes). In order to identify factors important for the frequency of accidents, a multiple regression model including individual as a random factor was formulated. Major factors were identified by selecting the most appropriate of all models including the factors traffic density, age, AS, perimetric reaction time (RT), minimum linear distance between central fixation point and defect border (DB), and mean time to perform the driving task by Mallow’s Cp.
The mean frequency of accidents was 18% (range 0 to 67%) at low traffic density and 56% (20 to 93%) at high traffic density. Frequency of traffic accidents was explained best by traffic density, age of individuals and the area of sparing within the affected hemifield (all p < 0.05). This model explained 73% of the total variability (Radj=0.73). Adding one of the factors RT, DB or mean time lowered Radj and none of these factors was significant. Traffic density explained 55% of total variance, age 2.9% and AS 2.4%.
Traffic density is the most important factor for the frequency of traffic accidents. The area of sparing determines driving performance under virtual reality conditions independently of age and to a similar (low) extent as age.
This PDF is available to Subscribers Only