We read with interest the article by Dujardin et al., in which the investigators aimed to design a simple neuropsychological battery able to easily and accurately screen for dementia in Parkinson's disease (PDD) in routine clinical practice. 1 Brief cognitive tests have been developed in the field of dementias for screening mild Alzheimer's disease, 2 but few efforts have been done for screening PDD. We would like to note that our group previously published, in Movement Disorders, the validation study of the PDD-Short Screen (PDD-SS), 3 a brief cognitive test that showed high diagnostic accuracy for screening PDD.
Although abundant data regarding clinimetrics may bore clinicians looking for a practical neuropsychological instrument, adequate clinimetric development is mandatory for an instrument to be recommended. 4,5 In this regard, no information on the validity, consistency, reliability, and administration time of the scale was provided, and no single cut-off score for the whole scale was given. The investigators indicated cut-off scores for some of the tests that form the battery, but no detail of the accuracy for each item was provided. Although the scale appears to have excellent specificity (95%) for screening PDD, overall sensitivity (65%) is rather low. Moreover, it was not indicated which cut-off score was obtaining this diagnostic accuracy. Instructions regarding how to add the scores of every item of the scale to obtain the final score should have also been provided.
In screening for PDD, and using a cut-off score 11, the PDD-SS showed 89.8% sensitivity, 88.5% specificity, 80.7% positive predictive value, and 93.6% negative predictive value for the diagnosis of PDD. The PDD-SS showed a similar diagnostic accuracy than the Mattis Dementia Rating Scale, with the additional benefit of having a significantly shorter administration time (4-7 vs. 17-25 minutes). 3 In our validation study, the PDD-SS also showed appropriate clinimetric characteristics, with good to excellent construct and discriminative validity, and test-retest and interrater realibility. Also, PDD-SS scores were not influenced by age or education. The robustness in the structure and the ability of the scale to detect neuropsychological differences between individual patients allowed us to set up a short final version with a single cut-off score, which simplifies and allows an easier generalization of the instrument.
Although simplicity and briefness are important virtues for a cognitive scale, its applicability as a screening instrument must be supported by appropriate clinimetric investigations that guarantee a correct diagnosis in clinical situations (i.e., diagnosis of dementia) with important individual consequences.