A Bedside Score Predicts One-Year Walking Recovery in AIS C Spinal Cord Injury: A Multisite Cohort Study
In brief
Bedside score classifies AIS C spinal cord injury walking recovery with 81% accuracy
Among 1,414 adults with AIS C spinal cord injury, 40% regained walking by one year. A score combining leg strength and age separated a small group likely to walk from a small group unlikely to walk, with 81% accuracy; many patients remained uncertain. Prospective validation is needed before clinicians can rely on it for individual predictions.
- Journal
- Topics in spinal cord injury rehabilitation (Q1)
- Published
- 28 September 2026
- Study design
- Prospective / inception cohort
- Evidence level
- Level 2, Moderate (CEBM 2b)
- Authors
- Andrew C Smith, Jeffrey Kepple, Kenneth A Weber, Wesley A Thornton, Jordan R Connor, Jesse Villines, et al.
- PMID
- 42812979
- DOI
- 10.46292/sci26-00004
Why clinicians should know about it
- Picked for Rehabilitation (paper of the day, 3 October 2026): Bedside score predicts walking recovery in AIS C SCI
Abstract
BACKGROUND: Walking recovery after American Spinal Injury Association Impairment Scale (AIS) C spinal cord injury (SCI) remains difficult to predict. OBJECTIVES: This study aimed to validate the prognostic value of quadriceps strength for future walking and to develop and test a simplified bedside clinical prediction rule (CPR) for walking recovery in AIS C SCI. METHODS: We conducted a retrospective cohort study using Spinal Cord Injury Model Systems data. Neurological examinations were performed at rehabilitation admission. Predictors included binary indicators of ≥3/5 muscle strength for key myotomes (L2, L3, L4, S1) and age cutoffs (≥50, ≥65 years). Logistic regression coefficients were scaled into an integer-based bedside score with 3 categories: likely walk, likely no walk, and uncertain. Model performance was evaluated using a random 20% holdout test set. A total of 1414 adults initially classified as AIS C were included; 40% achieved walking recovery at 1 year. RESULTS: Quadriceps strength ≥3/5 was significantly associated with walking recovery (P < .0001) but demonstrated modest diagnostic accuracy when considered alone. In the CPR training set, 9% were categorized as likely walk and 7% as likely no walk, with accuracy 81%, sensitivity 92%, specificity 72%, positive predictive value 72%, and negative predictive value 92%, with similar performance in the test set. A large proportion of patients remained uncertain. CONCLUSION: A simple bedside score combining lower extremity strength and age offers clinically useful early risk stratification for individuals with AIS C SCI. Prospective validation and refinement are needed to improve classification of indeterminate cases. The CPR is available at: aiscwalkingcpr.org.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.