Systematic review and meta-analysis of association between cognitive tests and on-road driving ability in people with dementia.

Objectives: To assess the strength of association between cognitive tests and on-road driving ability in people with dementia by conducting a systematic review and meta-analysis.

Methods: A literature search strategy was developed using terms related to dementia and driving. Fifteen databases were searched. Studies were selected if they investigated the predictive ability of cognitive tests and on-road fitness to drive in people with dementia. The primary author performed the initial screening of title and abstract of each paper. Two reviewers then independently checked whether the abstracts met the inclusion and exclusion criteria. The primary author also hand searched included studies, reference lists and conducted citation searches. Studies were quality assessed using Newcastle-Ottawa Quality Assessment Scale for Cohort and Case-Control Studies. A random effect model was employed to analyse effect size.

Results: The initial search identified 6734 articles, of which 47 full-text articles were screened for inclusion. Thirty articles were included in the review and sixteen articles were included in the meta-analysis. Analysis showed statistically significant weak to moderate (0.3-0.4) relationships between cognitive measures and on-road driving for all cognitive domains except language.

Conclusion: All cognitive domains apart from language showed a statistically significant moderate association with on-road driving outcome in patients with mild dementia. A composite cognitive battery is likely to be better than testing single cognitive domains in assessing fitness to drive in patients with dementia.