Application of Rasch Analysis for Measurement of Disability in Multiple Sclerosis
Nadia Azad1, Andrew Tarr2, Christina Fournier1, William Tyor2
1Emory University, 2Atlanta VA Medical Center
Objective:
To use Rasch analysis to create a linear, validated scale of patient-reported outcomes to measure disability in multiple sclerosis (MS).
Background:
Measuring disability progression in MS is vital to patient management and clinical research. The gold-standard in MS is the Expanded Disability Status Scale (EDSS). However, it has significant limitations. Patient-reported outcomes (PROs) are a promising additional measurement of MS disability progression. Data on validity are lacking. We aim to create a linear, validated scale measuring MS disability by analyzing patient responses to a preliminary PRO using Rasch analysis. We compare our findings to the EDSS and MS functional composite (MSFC).
Design/Methods:
Participants aged 18 - 65 with all MS subtypes are recruited from the Atlanta VA and Emory MS Clinics. A preliminary questionnaire with 140 questions, EDSS, and MSFC are administered. Participants complete a post-visit questionnaire within 7 days for test-retest reliability. Rasch analysis will be performed on questionnaire data and items will be removed based on fit. Correlational analyses with EDSS and MSFC are performed. We will recruit 250 participants.
Results:
44 participants have completed the preliminary questionnaire. The mean (SD) age is 50.1 (10.2) with 68.2% females and 31.8% males. 54.6% of participants self-identified as black, 43.2% white and 2.3% mixed. 50% receive anti-CD20 therapy. Median EDSS score is 3.75 with a range of 1.0-7.0 and mean MSFC score is –1.75 (2.7). Total questionnaire score correlated moderately with EDSS (r=0.69) and less with MSFC (r=-0.34). Test-retest reliability studies strongly correlate (n=27, r= 0.95). Rasch analysis data will follow.
Conclusions:
Our sample represents a wide range of demographics, disability, and treatments and is similar to published MS cohorts. Preliminary analysis demonstrated good construct validity compared to EDSS and reliable test-retest. This linear scale may aid in monitoring disability progression in MS patients.