Validation and attempts of revision of the MDS-recommended tests for the screening of Parkinson's disease dementia

The Movement Disorders Society (MDS) formulated diagnostic criteria and assessment guidelines for the screening of dementia in Parkinson's disease (PD). We carried out a validation of the cognitive measures suggested in the screening algorithm (i.e. the Mini Mental State Examination - MMSE - total score, serial 7s subtraction, 3-word recall, pentagons copy, and one minute letter fluency) in 86 patients with PD. Thirty-six percent of participants were diagnosed with dementia using the MDS algorithm, but with the Dementia Rating Scale instead of the MMSE. The original MDS procedure misclassified 11 patients (12.8%) as false negatives and 3 (3.5%) as false positives, leading to 65% sensitivity and 95% specificity. The main reason for misdiagnoses was insensitivity of the MMSE total score. Three attempts were made to reach a better screening performance, which warrants high sensitivity more than high specificity: 1. exclusion of the MMSE total score as a diagnostic requirement; 2. determination of a better cut off through Receiver Operating Characteristic curve analysis; 3. replacement of the MMSE with the equally undemanding, but more PD-specific, Mini Mental Parkinson. The first two strategies generally yielded high sensitivity, but poor specificity. The best outcome was achieved using a Mini Mental Parkinson total score <27 as cognitive criterion: sensitivity was 87% and negative predictive value was 90%; however, specificity was only 67%. Our findings seem to suggest that MDS practical guidelines are specific, but might benefit from the use of more PD-oriented tools than the MMSE in terms of sensitivity. © 2013 Elsevier Ltd.