Validation of a Cognitive Screening Battery to Predict Fitness-to-Drive in Individuals with Multiple Sclerosis: A Preliminary Report

In a previous study, the authors identified five cognitive tests that together predicted the outcome of a comprehensive driving evaluation of 44 individuals with multiple sclerosis (age = 46 ± 11 years, 84% females) with 91% accuracy, 70% sensitivity, and 97% specificity. In this study, the authors sought to validate the predictive accuracy of the five tests in a different cohort of individuals with multiple sclerosis. Sixty-three participants (age = 49 ± 9 years, 89% females) were administered the five cognitive tests. Participants were also administered a standardized practical on-road driving test. Performance on the road test was judged by completing a 16-item checklist of very important driving skills. A raw score of 45 or more out of 50 maximum points was classified as “pass” and below 45 as “fail”. Performance on the five cognitive tests was used to predict the pass/fail outcome of the on-road test. Study results showed that all five variables each had significant association with the on-road test raw score. The five tests together explained 44% of the variance of the pass/fail classification. Participants’ “pass” or “fail” performance on the road test was predicted with 83% accuracy, 67% sensitivity, and 85% specificity. The short battery of five tests appears to be a valid predictor of fitness-to-drive of individuals with multiple sclerosis and more accurate at predicting individuals who will pass the on-road evaluation (85% specificity) than those who will fail (67% sensitivity).