The effect of length of follow-up on substantial clinical benefit thresholds in patients undergoing surgery for cervical degenerative myelopathy

Patient-reported outcome measures are increasingly used to access pain, disability, physical function, and mental status to quantify effectiveness of surgical intervention in cervical myelopathy, yet each score has little meaning without established thresholds linked to clinical benefit. We set out to develop thresholds for substantial clinical benefit (SCB) in patients undergoing surgery for cervical myelopathy and to evaluate the effect of length of follow-up on SCB thresholds. Thirty-five patients undergoing spinal surgery for progressive cervical myelopathy were tracked from 2005 to 2015. Observations were categorized into three groups: short-term, intermediate, and long-term, corresponding to average follow-up intervals of 3.8, 9.2, and 29.0 months. SCB thresholds were calculated for neck visual analog score (VAS), Neck Disability Index (NDI), Short Form-12 physical (PCS), SF-12 mental component scores (MCS), and modified Japanese Orthopedic Association score (mJOA) using receiver operating curve analysis with a 5-level patient satisfaction index as the anchor. SCB thresholds for each outcome measure were obtained with a range of areas under the curve indicating varying degrees of discriminatory ability, reported with increasing length of follow-up. NDI and PCS were most discriminatory of SCB at any time period. Stratification of thresholds by length of time revealed a significant effect of follow-up time with NDI but not PCS. NDI and PCS thresholds have significantly strong discriminatory value in identifying patients receiving substantial clinical benefit with regard to cervical myelopathy. When NDI is used to predict outcome, choosing thresholds calibrated for follow-up time is recommended to maximize predictive power. (C) 2018 Published by Elsevier Ltd.