Objective: To predict the need for dopaminergic therapy initiation in early PD patients, we sought to identify a short item bank focused on the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) patient-reported items derived from Parts 1B and 2, using longitudinal item response theory (IRT) modeling.

Background: The MDS-UPDRS Parts IB (non-motor experiences) and II (motor experiences) represent the patient voice and are critical for assessing PD severity. A previous baseline cross-sectional analysis in early PD demonstrated the superior performance of a six-item bank over the full items bank in predicting the initiation for dopaminergic therapy; however, the longitudinal performance of this item bank remains unexplored.

Method: We employed longitudinal data from 1,209 early PD patients from the Parkinson’s Progression Marker’s Initiative (PPMI), SURE-PD3, and Oxford Discovery cohorts. A unidimensional longitudinal IRT model was fitted to 20 patient-reported MDS-UPDRS items (7 from Part IB, 13 from Part II). Items were ranked based on standardized discrimination and information function to determine predictive strength for initiating dopaminergic therapy.

Results: Seven MDS-UPDRS items (Getting Out of Bed/Car/Deep Chair, Doing Hobbies/Other Activities, Hygiene, Dressing, Turning in Bed, Eating Tasks, and Walking and Balance) provided the best predictive sensitivity. This seven-item longitudinal model significantly outperformed the full 20-item set in predicting dopaminergic therapy initiation (C-index = 0.637 vs. 0.619; P = 0.003). Recognizing that six of these seven items matched the previously identified baseline cross-sectional item bank, we conducted an additional analysis restricted to these six items, which demonstrated similar predictive accuracy to the seven-item model (C-index = 0.638 vs. 0.637), yet still significantly outperformed the full 20-item set (P = 0.002).

Conclusion: Longitudinal IRT modeling of patient-reported MDS-UPDRS items markedly enhances the sensitivity of predicting the key clinical outcome of dopaminergic therapy initiation in early PD. We recommend adoption of the reduced six-item bank, given its concise nature and superior predictive accuracy compared to the full 20-item bank, to streamline clinical assessment and optimize patient management and clinical trial design.

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/a-shortened-mds-updrs-patient-based-item-bank-increases-sensitivity-for-predicting-dopaminergic-therapy-initiation-a-longitudinal-irt-analysis-in-early-pd/