This study addresses the need to measure and monitor objective, real-world driver safety behavior in at-risk drivers with age-related dysfunction.

Older drivers are at risk for age-related cognitive and visual dysfunction, which may increase errors leading to crashes.  As a result, new methods to assess risk are needed.

Patterns of real-world driver risk exposure and vehicle control in older adults with a range of cognitive and visual functional abilities were collected from passive-monitoring systems 

A total of 1,169 traces were extracted (300 feet before and 100 feet after) at 173 intersections in Omaha, Nebraska. Stopping behavior (rolling-stop/no-stop/full-stop) were recorded for each trace (dependent variable). Independent variables include intersection, environmental, driver gender, traffic conditions, glances, visual acuity scores (logMAR values) and cognitive assessment score (MoCa test). Stopping behavior was modeled using logistic regression using with forward stepwise regression. Statistical interactions were considered in the model selection process.

Drivers with multiple glances left and right are, in general, more likely to engage in a full stop at an intersection.  The likelihood changes for drivers with low MoCa score (<25) and high MoCa score (>25): drivers in the former group are 44% more likely to fully stop, while those in the latter group are 282% more likely to fully stop. Drivers with worse vision are more likely to fully stop at an intersection.  Other variables included in the final model are time of day, turning movement, crossing vehicle status, opposing vehicle status, and road type.

We found a relationship between stopping behavior, and vision and cognitive scores. In both cases, drivers with worse scores are more likely to fully stop at traffic intersections.   This suggests drivers with impairments may compensate.  Additionally, the results suggest differences in stopping behavior may be used to identify drivers experiencing cognitive or visual impairment.