T.D. Marcotte, J.C. Scott
Driving an automobile is an activity deeply embedded in many societies. As one of the most dangerous of daily tasks, there has been increased interest in establishing methods for identifying drivers at risk for accidents, both in the normal aging population and neurologic cohorts. The aim of this paper is to review the major methodologies that have been applied for the assessment of driving abilities, and to discuss their relative benefits and limitations. Laboratory measures, such as neuropsychological tests, have been used to infer poor driving skills based upon poor test performance, but have met with mixed success. While poorer global cognitive performance is associated with impaired driving, the specificity of the neuropsychological measures has been limited. Measures of visual attention appear to be the most promising of the in-clinic measures. Accident history (often based upon accidents per million miles) has been used to examine whether certain populations are at greater risk for on-road accidents. While this provides a valid measure of “real world” risk, it does not inform researchers and clinicians regarding individual risk and may over- or underestimate risk depending upon the driving environs. On-road evaluations, either on public streets or a closed course, are often considered the gold standard for determining driving impairment. These assessments have the best face validity, most closely approximating true driving, but there remains some question as to whether they adequately assess all important driving skills (e.g., judgment and executive functioning). Lastly, the emergence of driving simulators has provided an exciting opportunity to evaluate underaddressed skills such as accident avoidance and navigational abilities, and to experimentally delineate the components of driving performance. Concerns remain regarding their realism and usability in various populations. Given the limitations of each method in detecting impaired drivers, we recommend a multi-modal approach to assessing risks at both the group and individual level. Future directions include application of new technologies such as eye trackers to better evaluate aspects of the driving process, such as attentional allocation. Lastly, we briefly review the development of a driving research program at the HIV Neurobehavioral Research Center in the United States.
Keywords: driving; driving assessments; driving simulations