Attentional orienting deficits as a predictor of driving impairment in older adults.

摘要

Adults over the age of 65 are among the highest risk groups for automobile accidents per mile driven. Previous attempts to predict accident risk within this population from basic physical or visual impairments have been largely unsuccessful. The Useful Field of View (UFOV), a measure of preattentive processing speed, is the most promising to date as a predictor of accident risk in older adults. The possibility that deficits in attentional orienting might also be related to driving impairment in older adults was examined in a sample of 155 current drivers age 63--87.; There were two primary aims. The first aim was to develop and evaluate DriverScan, a measure of attentional orienting, via item response theory techniques. The difficulty values of the DriverScan change detection items were predominantly located such that persons with lower orienting abilities (deficits) were measured most precisely. Item difficulties were significantly correlated with those computed from weighted combinations of greater visual clutter, less change relevance to driving, and less change brightness. DriverScan was correlated only with the attention-related subtests of the UFOV, as hypothesized, although the influence of context on attentional functioning within this task distinguishes the instrument from the UFOV.; The second aim was to examine the relationships between visual impairment, visual attention deficits (UFOV; DriverScan), and driving impairment (real-life accidents within the previous 3 years; driving simulator performance). No measures were correlated with accident history, most likely due to restriction in range. The best predictors of simulator driving impairment (R2 = .43) were UFOV Processing Speed and DriverScan, which were each predicted by visual impairment and/or age. At the present time, however, it is recommended that these measures could only be used as screening tools with which to identify older adults whose attentional impairments warrant further testing or training, given their lack of sensitivity and specificity in predicting accident history or simulator impairment.