Neurological recovery following traumatic spinal cord injury: a systematic review and meta-analysis

OBJECTIVE: Predicting neurological recovery following traumatic spinal cord injury (TSCI) is a complex task considering the heterogeneous nature of injury and the inconsistency of individual studies. This study aims to summarize the current evidence on neurological recovery following TSCI by use of a meta-analytical approach, and to identify injury, treatment, and study variables with prognostic significance. METHODS: A literature search in MEDLINE and EMBASE was performed, and studies reporting follow-up changes in American Spinal Injury Association (ASIA) Impairment Scale (AIS) or Frankel or ASIA motor score (AMS) scales were included in the meta-analysis. The proportion of patients with at least 1 grade of AIS/Frankel improvement, and point changes in AMS were calculated using random pooled effect analysis. The potential effect of severity, level and mechanism of injury, type of treatment, time and country of study, and follow-up duration were evaluated using meta-regression analysis. RESULTS: A total of 114 studies were included, reporting AIS/Frankel changes in 19,913 patients and AMS changes in 6920 patients. Overall, the quality of evidence was poor. The AIS/Frankel conversion rate was 19.3% (95% CI 16.2-22.6) for patients with grade A, 73.8% (95% CI 69.0-78.4) for those with grade B, 87.3% (95% CI 77.9-94.8) for those with grade C, and 46.5% (95% CI 38.2-54.9) for those with grade D. Neurological recovery was significantly different between all grades of SCI severity in the following order: C > B > D > A. Level of injury was a significant predictor of recovery; recovery rates followed this pattern: lumbar > cervical and thoracolumbar > thoracic. Thoracic SCI and penetrating SCI were significantly more likely to result in complete injury. Penetrating TSCI had a significantly lower recovery rate compared to blunt injury (OR 0.76, 95% CI 0.62-0.92; p = 0.006). Recovery rate was positively correlated with longer follow-up duration (p = 0.001). Studies with follow-up durations of approximately 6 months or less reported significantly lower recovery rates for incomplete SCI compared to studies with long-term (3-5 years) follow-ups. CONCLUSIONS: The authors' meta-analysis provides an overall quantitative description of neurological outcomes associated with TSCI. Moreover, they demonstrated how neurological recovery after TSCI is significantly dependent on injury factors (i.e., severity, level, and mechanism of injury), but is not associated with type of treatment or country of origin. Based on these results, a minimum follow-up of 12 months is recommended for TSCI studies that include patients with neurologically incomplete injury.