It is premature to test older drivers with the SIMARD-MD

Abstract

Background: A new tool, the SIMARD-MD, has been proposed to help physicians identify cognitively impaired drivers who may be unfit to drive, but little empirical evidence is available to justify its use. We analyzed data from a cohort of older Canadian drivers who had undergone cognitive testing to: 1) correlate the SIMARD-MD with other tools that measure cognition (e.g., Trail-Making Test), 2) identify how many drivers, using published cut-offs on the SIMARD-MD, would be recommended to lose their license, or be considered fit to drive, or be required to undergo further driving assessment, and 3) determine if the SIMARD-MD is biased by level of education as many cognitive tools are. Methods: Cross-sectional data from 841 drivers aged 70 and over from seven Canadian sites who are enrolled in a 5-year cohort study were used for the analyses. Scores on the SIMARD-MD were correlated with scores on the other cognitive measures. The recommendations that would be made based on the SIMARD-MD scores were based on published cut-off values suggested by the authors of the tool. The impact of education status was examined using linear regression controlling for age. Results: Correlations between the SIMARD-MD and other cognitive measures ranged from .15 to .86. Using published cut-off scores, 21 participants (2.5%) would have been recommended to relinquish their licenses, 428 (50.9%) would have been deemed fit to drive, and 392 (46.6%) would have been required to undergo further testing. We found a difference of 8.19 points (95% CI = 4.99, 11.40, p < .001) in favour of drivers with post-secondary education versus those without, representing over 11% of the mean score.