Exploring driving characteristics of fit- and unfit-to-drive neurological patients: a driving simulator study
Lenart Motnikar,Kristina Stojmenova,Urša Čižman Štaba,Tara Klun,Karmen Resnik Robida,Jaka Sodnik
DOI: https://doi.org/10.1080/15389588.2020.1764547
Abstract:Objective: To identify driving characteristics of fit-, unfit-, and conditionally fit-to-drive neurological patient populations using a driving simulator with three high-risk scenarios comprising rural, highway, and urban environments.Methods: The study included 91 neurological patients undergoing a multidisciplinary assessment for driver's license revalidation, consisting of a clinical, neuropsychological, functional, and on-road evaluation. The groups drove through three independent driving scenarios, during which a variety of measures describing reaction time, vehicular control, and traffic rule compliance were performed. One-way analysis of variance (ANOVA) with Bonferroni correction was used for group comparison, independently for each driving scenario, and Pearson correlations were calculated between simulator variables and neuropsychological test scores.Results: The fit- and unfit-to-drive population significantly differed (p < .05) in reaction times, regardless of the scenario. No significant differences in traffic rule compliance or vehicular control parameters were observed in the rural environment (p > .05). On the highway, the unfit group exhibited greater variability of steering wheel angle, higher steering reversal rate, and a higher rate of turn signal errors. In the urban environment, the unfit group oversped more, had more collisions, and exhibited greater lane position variability. The latter, along with reaction times in the rural and highway scenarios, was also shown to significantly differ between the conditional and unfit group (p < .05). No significant differences were observed between the fit and the conditional group (p > .05). Weak to moderate associations (range: -0.5 to 0.29) between neuropsychological tests and various simulator variables were also observed.Conclusions: Our results show that driving simulators are able to capture differences between (fit- and unfit-to-drive) neurological patient populations and therefore bear the potential for being used as a deficit-independent screening, assessment, or rehabilitation tool. The conditionally-fit-to-drive group exhibited less discriminative features, which points to greater importance of human judgment for this population. The observation that differences in most of the parameters were environment-dependent suggests that developers of future driver simulation tools should carefully design scenarios in order to fully exploit their assessment potential.