Predicting Driving Performance in Older Adults: We Are Not There Yet!

Abstract

Statistical associations are not sufficient to infer adequate predictive value, especially when crucial decisions such as whether one can continue driving are at stake. The predictors we examined have limited predictive value if used as stand-alone screening tests.

Predicting Road Test Performance in Drivers With Stroke

“…Additional studies are needed with larger sample sizes and in a variety of settings to validate these results; no test or set of tests has been studied with sufficient rigor to warrant adoption (Bédard, Weaver, Darzins, & Porter, 2008). Although we report on specific cutpoints, more debate is needed on the placement of cutpoints; the level of test uncertainty that is acceptable to clients, families, and professionals; and an acceptable correct classification rate to consider adopting such a test paradigm.…”

Dynamic Attention as a Predictor of Driving Performance in Clinical Populations: Preliminary Results

“…Particularly successful has been the Useful Field of View test (UFOV, Ball et al, 1988;Hoffman, McDowd, Atchley, & Dubinsky, 2005). However, while there are significant associations between UFOV and driving performance (Clay et al, 2005), ROC analysis (Bedard, Weaver, Darzins, & Porter, 2008) suggests that as a screening test it would produce unacceptable numbers of false positives (unsafe drivers categorized as safe) and/or false negatives (safe drivers categorized as unsafe).…”

Evaluating the SIMARD MD a New Screening Tool to Identify Cognitively Impaired Drivers

“…The discrepancy between their text and the ROC plots is not clear.) The authors base their conclusion, in large part, on a comparison of ROC curves from two data points they indicate are derived from Dobbs and Schopflocher's Screen for the Identification of Cognitively Impaired Medically At-Risk Drivers, A Modification of the DemTect (SIMARD MD) findings 2 to two data points from an incompletely described combination of data from the standardized Mini-Mental Status Evaluation (MMSE) and Trails A from a previously published study (see Bédard et al 2008) 3 . There are a number of methodological flaws with the Bédard et al 1 comparison that invalidate their conclusions.…”

“…2 However, the two studies used different criteria (road outcomes from what appears to be a standard novice driver road test in the Bédard et al study 1 vs pass/fail outcomes from an on-road driving assessment specifically developed to assess cognitively impaired drivers in the Dobbs and Schopflocher studies 2 ). In addition, the two tests were evaluated using different populations (a sample of drivers who "volunteered to participate" [p 337] in the 2008 Bédard et al study, 3 compared to a mix of drivers diagnosed with cognitive impairment with or without dementia [about twothirds of the sample] and healthy controls [about one-third of the sample] in the Dobbs and Schopflocher studies 2 ). Because these differences in the criteria and populations invalidate all of their test comparisons, the differences in the criteria and populations are detailed below.…”