485 Identifying the Minimum Clinically Important Differences (MCID) in ISNCSCI Score for Patients with Traumatic Spinal Cord Injury: an Analysis from the National Spinal Cord Injury Model Systems (SCIMS) Database

INTRODUCTION: Spinal Cord Injury (SCI) is a major cause of disability, mostly presenting due to a traumatic event such as a motor vehicle accident or a fall. To evaluate and follow-up the neurological recovery for patients with SCI, several scales, and scoring systems have been published with the collaborative multidisciplinary efforts, including the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI). Identifying an MCID for this scoring system is crucial to evaluate patient improvement and provide hints about the success of any intervention. METHODS: National Spinal Cord Injury Model Systems Database was queried from 2001-2016 for patients admitted to rehabilitation centers following traumatic spinal cord injury. The following methods were used to identify MCIDs; (i) anchor-based methods where an outcome measure is selected to be the “anchor” (Functional Independence Measure) and defined changes in this anchor is linked to the changes in the investigated outcome measure (ISNCSCI motor scores) and (ii) distribution-based methods where several different parameters (Standard error of measurement, half standard deviation, etc.) are used determine one MCID. RESULTS: A total of 1,521 patients were identified. MCIDs for the overall cohort ranged from 3.25 to 6.6 for the anchor-based methods and 0.93 to 5.30 based on the distribution-based methods. Although slight changes were seen, the range of MCIDs was found to be similar to the overall cohort when cohorts were stratified by sex, the level of injury, complete/incomplete injury, and different age groups. The highest MCID was detected for patients with cervical injury in both anchor-based (9.03) and distribution-based approaches (6.06), while the lowest MCID was detected for patients with thoracic injury in both anchor-based (1.0) and distribution-based approaches (0.23). CONCLUSION: Our findings demonstrated similar results with regards to MCIDs for ISNCSCI scoring between different patient profiles (stratified by sex, the level of injury, complete/incomplete injury, and age groups) with scores ranging from 0.23 to 9.09 based on different approaches. Anchor-based methods provided higher MCIDs overall when compared to distribution-based methods.