The driving population of the USA is aging and faces increased risk for injury and death from motor vehicle crashes. Clinicians are often asked to judge the driving safety of their patients without adequate guidelines. This article describes the development of a systematic performance-based road test for measuring driving skills of elderly drivers and its correlation with cognitive measures. A prospective, masked, observational study was used, in which a driving instructor's global scores ("criterion standard") and cognitive test scores were correlated with research driving scores created by two independent research raters sitting in the back seat of the car during each driving test. A convenience sample of 30 licensed drivers with a broad range of cognitive skills, over age 60, were studied on a closed course and in traffic. Statistically significant correlations were observed between the "criterion standard" and closed course scores (r = .35, p<.05) and between the "criterion standard" and in-traffic-scores (r = .64, p <.01). Significant correlations were obtained between in-traffic and cognitive test scores, e.g. Mini-Mental State Exam (r = .72, p <.01). Inter-rater reliability on the closed course was .84 and on the in-traffic component was .74. Internal consistence for the closed course was .78 and for in-traffic .89. This study documented the safety, reliability, and vaiditity of a systematic road test for elderly drivers with a range of cognitive skills. Larger studies are needed to determine the cognitive factors that independently predict driving performance.
Abstract