Driving performance in patients with homonymous visual field defects and healthy subjects in a standardized virtual reality environment

Abstract

Abstract Purpose The aim of this study was to assess whether visual field‐related parameters per se are able to predict driving performance of patients with homonymous visual field defects (HVFDs) and healthy controls in a standardized virtual reality (VR) environment. Methods Thirty‐two patients with HVFDs due to cerebro‐vascular lesions and 32 healthy control subjects underwent testing on a driving simulator under two levels of traffic density. Driving performance was defined as the frequency of accidents under VR‐conditions. From the binocular visual field, as measured with kinetic perimetry, we assessed the area of sparing within the affected hemifield (A‐SPAR) and the perimetric reaction time (RT). Results Frequency of traffic accidents was best explained by traffic density, age of individuals, A‐SPAR and RT. This model explained 78% of the total variability (Radj2 = 0.78). All factors were significant. The effect of traffic density explained 63.4 % of the total variance. Age, A‐SPAR and RT were minor effects explaining only 2.7%, 2.0% and 0.9% respectively. Conclusion The extent of the visual field loss is weakly related to driving performance. This finding suggests that – at least for this group of patients – the visual field‐related parameters should not be taken as the sole indicator of fitness to drive. Some patients with HVFDs demonstrated sufficient compensatory driving behavior during the simulated test ride, presumably due to eye and head movements. Driving performance declined slightly with age; these changes were exacerbated in the presence of cerebro‐vascular disease.