The assessment of fitness to drive in people with dementia.

Objective: To determine whether cognitive tests predict fitness to drive in patients with dementia.

Design: Two group comparison of patients with dementia and healthy elderly volunteers, and comparison of patients with dementia who were found safe to drive and those found unsafe, followed by a validation study.

Participants: Forty-two people with dementia and 33 healthy elderly volunteers with no known memory problems who were driving. Of the 42 people with dementia 37 were assessed on the road. A second sample of 17 people with dementia was also assessed on the road.

Main Measures: Stroke Drivers Screening Assessment, Mini Mental State Examination, Salford Objective Recognition Test, Stroop Test, Test of Everyday Attention, Visual Object and Space Perception Battery, Behavioural Assessment of the Dysexecutive Syndrome, Adult Memory and Information Processing Battery.

Results: All healthy elderly volunteers were safe to drive but 10 of the 27 patients with dementia were unsafe. Discriminant function analysis identified a combination of tests, which correctly classified 92% of drivers with dementia as safe or unsafe. Validation of this prediction on an independent sample had 59% accuracy using a cut-off of 0 but 88% accuracy using a cut-off of 5.

Conclusions: Safety to drive in people with dementia could be predicted from a combination of six cognitive tests. These correctly identified 67% of safe drivers in a validation sample. This assessment could be used to identify those who need evaluation of their safety on the road.