Reliability and reproducibility of the modified thoracolumbar injury classification and severity score and the thoracolumbar AO spine injury score for guiding surgical decision-making in thoracolumbar fractures.

ObjectiveTo evaluate the reliability and reproducibility of the modified Thoracolumbar Injury Classification and Severity Score (mTLICS) and the Thoracolumbar AO Spine Injury Score (TL AOSIS) in guiding surgical decision-making for thoracolumbar fractures, and to investigate factors associated with observer agreement.MethodsA retrospective analysis was conducted on 100 thoracolumbar injury patients treated at the Second Affiliated Hospital of Zhejiang Chinese Medical University from January 2021 to December 2023, all with complete preoperative imaging data. The cohort included 64 males and 36 females, aged 25-55 years (mean 41.3 ± 6.9 years). Six evaluators independently assessed the anonymized cases on two occasions separated by a 4-week interval. Weighted Cohen's kappa coefficients were used to evaluate interobserver reliability and intraobserver reproducibility.ResultsFor TL AOSIS, interobserver/intraobserver kappa values 0.706/0.687 for fracture morphology, 0.906/0.942 for neurological status, 0.869/0.879 for tension band injury, and 0.736/0.732 for surgical recommendation, respectively. For mTLICS, interobserver/intraobserver kappa values were 0.773/0.763 for fracture morphology, 0.878/0.894 for neurological status, 0.716/0.721 for tension band injury, 0.837/0.845 for disc injury, and 0.702/0.685 for surgical recommendation, respectively. Significant differences (P P > 0.05) were found in neurological status evaluations.ConclusionBoth TL AOSIS and mTLICS demonstrate good reliability and reproducibility in guiding surgical decision-making for thoracolumbar fractures.