A Study on Interobserver and Intraobserver Reliability of the Huashan Radiologic Classification System for Cervical Spinal Cord Injury Without Fracture and Dislocation
Chi Sun,Guangyu Xu,Hongwei Wang,Cong Nie,Guanhua Xu,Dayong Liu,Yong Yang,Xiandi Wang,Lin Xie,Linli Li,Xiaosheng Ma,Feizhou Lu,Jianyuan Jiang,Hongli Wang
DOI: https://doi.org/10.1097/bsd.0000000000001621
2024-01-01
Abstract:Study Design: Observational study. Objective: To assess the reproducibility and reliability of the system. Background: The Huashan radiologic classification system for cervical spinal cord injury without fracture and dislocation (CSCIWFD) was recently proposed and found useful for clinical practice. Patients and Methods: Patients diagnosed with CSCIWFD between 2015 and 2021 were recruited. Six spine surgeons from different institutions, three experienced and other inexperienced respectively, were trained as observers of the system, and these surgeons classified the recruited patients using the system. Then, 8 weeks later, they repeated the classification on the same patients in a different order. The interobserver and intraobserver agreement between the results was analyzed using percentage agreement, weighted kappa, and Cohen kappa (κ) statistics. Results: A total of 60 patients were included in the analysis. Type I was the most frequent type (29 cases, 48.3%), followed by type II (13 cases, 21.7%), type III (12 cases, 20%), and type IV (6 cases, 10%). For all the observers, experienced observers, and inexperienced observers, the overall agreement percentages were 77.6% (κ = 0.78), 84.4% (κ = 0.84), and 72.8% (κ = 0.74), respectively, indicating substantial to nearly perfect interobserver reproducibility. A higher level of agreement was found for differentiating type I from other types, with the percentage agreement ranging from 87.8% to 94.4% (κ= 0.74–0.88). For distinguishing compression on the spinal cord (types I and II vs types III and IV) among the different groups of observers, the percentage agreement was 97.8% (κ = 0.94), indicating nearly perfect reproducibility. As for intraobserver agreement, the percentage agreement ranged from 86.7% to 96.7% (κ = 0.78–0.95), indicating at least substantial reliability. Conclusions: The Huashan radiologic classification system for CSCIWFD was easy to learn and apply in a clinical environment, showing excellent reproducibility and reliability. Therefore, it would be promising to apply and promote this system for the precise evaluation and personalized treatment strategy.