Psychometric Validation of a Modified United Multiple System Atrophy Rating Scale

Congratulations to Dr. Nikolaus McFarland on the publication of “Psychometric Validation of a Modified United Multiple System Atrophy Rating Scale,” which appears in the April 26, 2023 edition of Neurology.



Objective: To evaluate the measurement properties of a 9-item modified Unified Multiple System Atrophy Rating Scale (UMSARS).

Background: The UMSARS was designed to evaluate the signs and symptoms of multiple system atrophy (MSA) that progress over a patient’s course of disease. A subset of 9-items from the UMSARS was proposed for use as a clinically meaningful and sensitive measure of disease progression in a clinical trial conducted over 1-year. Psychometric validation was conducted to inform measurement properties of this 9-item selection (speech, cutting food and handling utensils, dressing, hygiene, walking, urinary function, arising from chair, posture, gait).

Design/Methods: The data used for this analysis was obtained from the M-STAR phase 3 clinical trial in participants with MSA (NCT03952806). The psychometric properties evaluated were data acceptability (floor and ceiling effects), internal consistency reliability, test-retest reliability, convergent validity, and responsiveness (cross-sectional evaluation).

Results: Floor or ceiling effects were not observed. Strong internal consistency reliability was demonstrated (αtotal = 0.79) and item-to-total relationships were adequate (r ≥ 0.50) for all items, except urinary function item [r=0.31]). The intra-class correlation coefficient (ICC) indicated strong test-retest reliability (ICC=0.93). Correlations between the modified UMSARS and other validated MSA scales were as follows: r=0.64 with MSA-QoL motor domain p<0.0001; r=−0.41 with SF-36 physical component summary p<0.0001; r=0.27 with Zarit Burden Interview p<0.0001; r=0.23 with PQoL carers p<0.0001; r=0.07 with COMPASS 5-Domain SELECT, p=0.2217. Mean scores for each item increased with disease severity (as assessed by UMSARS-IV) (ANOVA, p<.05), except for the urinary function item.

Conclusions: This 9-item modified UMSARS demonstrated adequate evidence of psychometric validity. While removal of the single non-motor item (urinary item) resulted in a measure with higher responsiveness, consistency, and reliability, it should be considered that this modification eliminates a core non-motor dysfunction within this heterogenous disease.