Driving with Alzheimer disease: the anatomy of a crash

Alzheimer disease (AD) affects several abilities that are crucial to automobile driving and can thereby increase the risk for a crash. The specific driver safety errors that lead up to crashes in this driver population, however, remain largely unknown. We developed a graphical means for dissecting these safety-related behaviors for application to car crashes of AD drivers. These crashes occurred in collision avoidance scenarios implemented on the Iowa Driving Simulator. Crash plots were produced by "rewinding" the data stream, then graphing the vehicle speed, path, steering wheel position, pedal positions, and driver gaze up to the very moment of impact. In this way, we were able to display in detail the patterns of vehicle and driver control to determine exactly how a crash occurred. Several different types of antecedent safety errors were noted, ranging from responding inappropriately with the hand and foot controls to not responding at all. In this paper we focus on the performance of a single impaired driver to demonstrate our technique and findings. Understanding the "anatomy" of crashes in cognitively disabled drivers with AD can help guide future efforts at injury prevention and control.