Abstract
Purpose :
Screening vision for licensure is intended by policy makers to ensure that drivers have adequate vision to support safe driving. Visual acuity, a common screening test, assesses if road signs can be read, but there is no evidence that impaired acuity increases crash risk. Contrast sensitivity, visual field light sensitivity, visual processing speed, and visual-spatial skills have been identified as visual functions associated with motor vehicle collision (MVC) involvement by older drivers. Whether these factors distinguish drivers that will vs. will not be involved in MVCs in subsequent years is unknown. The purpose of this study is to examine the sensitivity and specificity of vision tests in older drivers for identifying those who have ≥1 MVC in the subsequent 3 years.
Methods :
Visual acuity, contrast sensitivity, driving visual field sensitivity, UFOV (subtest 2), Trails B, and Spatial Closure Subtest of the MVPT were assessed at baseline in a population-based sample of 2000 licensed drivers aged ≥70 years from north central Alabama. Information on MVC involvement occurring 3 years after enrollment was obtained from the Alabama Department of Public Safety. Clinically meaningful cut-points were used to define impairment for each measure of visual function. Sensitivity and specificity were calculated for each visual function separately and in combination with visual acuity. Area under the curve (AUC)/receiver operator characteristic (ROC) curves using logistic regression models were generated.
Results :
Over the study period, participants were involved in 278 MVCs, yielding a rate of 5.59 MVCs per million miles traveled. The sensitivity of visual acuity impairment for identifying any MVC involvement was 10.4% and the specificity was 92.3%. Impaired visual field status had the highest sensitivity for any MVC involvement at 26.5% (74.4% for specificity). ROC analysis showed that Trails B had the highest AUC of 0.53 for any MVC involvement. Combinations of visual functions did not improve sensitivity and specificity. The pattern of results was similar for at-fault MVCs.
Conclusions :
None of the vision tests demonstrate high sensitivity and specificity, either alone or in combination for identifying drivers who were crash involved in the subsequent 3 years. While these vision measures are poor screening tests, they may remain important risk factors for MVC involvement among older drivers.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.