471 Using the QOD to Identify Minimum Clinically Important Differences for Patients With Cervical Spondylotic Myelopathy

Abstract

INTRODUCTION: Minimum clinically important differences (MCID) for quality-adjusted life years (QALYs) and arm pain patient-reported outcome metrics (PROMs) in cervical spondylotic myelopathy (CSM) have yet to be determined, while data for myelopathy severity, impairment, and neck pain remain limited by small series and low rates of follow-up. METHODS: The CSM database of the QOD Study Group was utilized for this analysis. Patients were prospectively enrolled and followed for 24 months, with a follow-up rate of 85.2%. The MCIDs were computed for five standardized instruments: Neck Disability Index (NDI), five-dimension Euro-QoL (EQ-5D) in QALYs, neck and arm pain numeric rating scale (NRS), and modified Japanese Orthopedic Association score (mJOA) at both 3 and 24-month follow-up. Seven previously validated measures of MCID were compared using the area-under-the-curve (AUC). The North American Spine Society (NASS) satisfaction scale was the anchor, with scores of 1-2 meeting criteria for satisfaction. RESULTS: A total of 1,141 patients underwent surgery for CSM. Improvement of =30% from baseline was the optimal MCID across time points for NDI and neck/arm NRS. At 24 months, 30% improvement for these three instruments resulted in AUCs of 0.73, 0.71, and 0.67, respectively. Numeric cut-offs (0.065 at 3 months and 0.149 at 24 months) were the superior MCIDs for EQ-5D-derived QALYs. For mJOA, a severity-adjusted MCID outperformed the other methods, yielding an AUC of 0.67 at 24 months. CONCLUSIONS: We used the QOD dataset to define previously unknown MCIDs for several outcome measures. A =30% improvement from baseline was a consistent, high-performing MCID in NDI and arm/neck NRS. However, an absolute numeric cut-off (0.149) was the superior MCID for EQ-5D QALYs. mJOA exhibited the lowest predictive capacity for patient satisfaction.