Clinical validation of movement disorder society–recommended diagnostic criteria for Parkinson's disease with dementia

Abstract

The objective of this work was to evaluate the Movement Disorders Society (MDS) Task Force-proposed screening checklist for detecting Parkinson's disease dementia (PD-D) in relation to full neuropsychological testing. An MDS Task Force has proposed diagnostic procedures for PD-D, which have not been fully validated against more extensive neuropsychological testing. PD subjects were recruited from 2 specialty centers. A neuropsychologist evaluated them for dementia as part of routine clinical care. Independent clinical neurologists administered the MDS PD-D screening checklist. Diagnosis of PD-D by the 2 methods was compared. Ninety-one PD subjects had a mean age of 66.3 (SD = 9.7) years and a mean PD duration of 8.8 (SD = 6.1) years. Seven subjects (7.7%) met all 8 screening checklist criteria from the MDS PD-D screening tool and were classified as probable PD-D. Fifteen (16.5%) subjects were classified as PD-D by full neuropsychological assessment. The screening checklist showed 100% specificity, but only 46.7% sensitivity, for diagnosing PD-D compared to the full neuropsychological assessment. PD-D cases missed by the PD-D screening tool were largely due to 2 checklist items that were not endorsed (absence of depression and Mini-Mental State Examination [MMSE] scores <26). There was moderate agreement between these 2 methods for determination of PD-D (kappa = 0.59, P < .001). The MDS-PD-D screening checklist is highly accurate for detecting PD-D if all items are endorsed. However, for cases that do not meet these criteria, full neuropsychological testing is needed to differentiate PD-D from milder cognitive impairment. Revision of the checklist by altering or eliminating the 2 problematic checklist items may improve sensitivity.