Principal Data Scientist Genentech, a Member of the Roche Group South San Francisco, United States
Background: Expanded disability status scale (EDSS) is the gold standard for measuring multiple sclerosis (MS) disability in clinical trials, but it is not routinely collected in clinical practice and thus not reliably available in administrative claims or health records, presenting a major challenge for real-world data studies. Ambulatory status has been proposed as a proxy measure for MS disability and has shown moderate correlation with EDSS in linked claims-electronic health record data.
Objectives: To validate ambulatory status as a proxy measure for disability in MS using digitized health records from the FlywheelMS cohort.
Methods: This study used data (for the period January 1, 2019 to December 31, 2019) from the FlywheelMS cohort, a patient-centered study that digitizes health records from various sites of care for participating MS patients in the US. The sample included patients with MS who had ≥1 diagnosis for MS before the study period, continuous database activity during the study period, and ≥1 EDSS score recorded during the study period. Ambulatory status was derived as a categorical variable (fully ambulatory, walks with assistance, non-ambulatory) and defined based on diagnosis and procedure codes/terms. EDSS constituted the reference standard and was categorized as EDSS < 4 (without walking impairment) or EDSS≥4 (with walking impairment). Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and agreement with EDSS (using Spearman’s rank correlation coefficient) were determined. Sensitivity analyses were performed using data from earlier periods.
Results: A total of 381 patients with MS were included in this study. The mean age was 48.8 years (standard deviation, 11.3) and 74.3% were females. Most patients had relapsing remitting MS (52.8%), followed by primary progressive MS (15.2%) and secondary progressive MS (7.3%). The median EDSS was 4.0 (minimum-maximum, 0.0-9.0). This study found moderate correlation between ambulatory status and EDSS (⍴=0.61, p< 0.001). When categorized as without walking impairment (EDSS < 4) or with walking impairment (EDSS≥4), ambulatory status had a sensitivity of 87%, specificity of 68%, PPV of 77% and NPV of 82% in identifying walking impairment. Similar results were observed in sensitivity analyses.
Conclusions: This study validated ambulatory status as a proxy measure for disability in MS using an independent dataset, showing that the proxy measure may support pharmaco-epidemiological studies that wish to account for disability but do not have access to EDSS. Future studies are needed to refine and apply the proxy measure in different data sources and populations.
