Cognitive testing in the diagnosis of parkinsonian disorders: a critical appraisal of the literature.

A number of cognitive abnormalities have been identified as putative diagnostic markers in neurodegenerative parkinsonism based on statistically significant differences between diseases. The effect sizes of these differences frequently have not been reported, making practical interpretation of the results difficult. The objective of this study was to assess the potential contribution of cognitive testing to the diagnosis of parkinsonian disorders by quantifying effect size of previously identified significant differentiating cognitive features. A Medline search identified 1038 articles. Nineteen studies directly comparing at least 2 of Parkinson's disease, progressive supranuclear palsy, multiple system atrophy, and corticobasal syndrome/degeneration were selected. Cohen's d and positive likelihood ratio were calculated as appropriate for cognitive tests showing statistically significant differences between diseases. Clinically useful differences were considered present when Cohen's d > 1 or the positive likelihood ratio > 2 and were considered high when Cohen's d > 1.94 or the positive likelihood ratio > 10. Only 16 of 141 cognitive tests were found to be highly useful. Cognitive testing was only moderately helpful in separating Parkinson's disease and multiple system atrophy. Inferior performance on phonemic (d = 1.56-2.13) and semantic (d = 1.43-2.13) verbal fluency, the Trail-Making Test (d = 1.63-1.95) and the Wisconsin Card Sorting Test (d = 1.63-2.22) were moderately to very useful in separating progressive supranuclear palsy from Parkinson's disease and multiple system atrophy. Cognitive testing could not differentiate corticobasal syndrome from other parkinsonian disorders, although sequential orobuccal apraxia was very useful (d = 2.01-2.23). Few of the cognitive tests separating parkinsonian disorders identified from previous studies have sufficient effect size to be practically useful. Even these features must be interpreted in conjunction with other clinical characteristics to be helpful diagnostically. © 2012 Movement Disorder Society.