Risk factors for failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction
10.3760/cma.j.cn115530-20240130-00049
- VernacularTitle:下颈椎骨折脱位合并关节突关节绞锁经颅骨牵引复位失败的危险因素分析
- Author:
Ziqiang ZHU
1
;
Zeyu SHANGGUAN
;
Xuexing SHI
;
Chunqing WANG
;
Jingming HE
;
Yuekui JIAN
;
Qing LI
Author Information
1. 贵州医科大学附属医院创伤骨科，贵阳 550004
- Keywords:
Cervical vertebrae;
Zygapophyseal joint;
Dislocations;
Spinal cord injuries;
Nomograms
- From:
Chinese Journal of Orthopaedic Trauma
2024;26(7):575-582
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To develop a nomogram predictive model on the basis of identification of the risk factors associated with failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction.Methods:A retrospective study was conducted of the clinical data of the patients who had been treated for dislocation of the subaxial cervical spine with locked facets at Department of Orthopaedic Trauma, The Hospital Affiliated to Guizhou Medical University and Department of Spine Surgery, The People's Hospital of Guizhou Province from January 2014 to December 2022. The clinical data from The Hospital Affiliated to Guizhou Medical University were used as a training set (156 cases) and those from The People's Hospital of Guizhou Province as an external validation set (54 cases). Univariate analysis and multi-variate logistic regression analysis of the training set were conducted to screen out independent risk factors associated with the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction. A nomogram predictive model was thus constructed and assessed by the receiver operating characteristic (ROC) curve, calibration curve, and decision curve. Internal validation of the training set and external validation set was used to evaluate and validate the model.Results:The multivariate logistic regression analysis revealed that cervical Ⅰ grade dislocation ( P=0.002), cervical Ⅱ grade dislocation ( P=0.007), low segment affected ( P=0.042), unilateral facet locked ( P=0.027), and the ASIA grading of spinal cord injury ( P=0.008) were the independent risk factors associated with the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction, based on which the nomogram model with a C-index of 0.88 was constructed to predict the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction. Analysis of the ROC curve of the training set showed an area under the curve (AUC) of 0.88, indicating good accuracy of the nomogram model. Analysis of the calibration curve showed high consistency between the probability of the nomogram model predicting the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction and the actual probability of traction reposition failure. Analysis of the decision curve showed that application of the nomogram model led to good benefits when the net benefit threshold for the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction was 0.03 to 0.84. Analysis of the ROC curve of external validation set showed an AUC of 0.79, indicating good accuracy of the nomogram model. The training set showed a C-index of 0.87 after 1,000 internal verifications by the Bootstrap method, indicating good discrimination of the nomogram model. Conclusions:Cervical Ⅰ grade dislocation, cervical Ⅱ grade dislocation, low segment affected, unilateral facet locked, and incomplete spinal cord injury are independent risk factors associated with failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction. A nomogram model has been successfully constructed which can predict the failure in repositioning the dislocation of the subaxial cervical spine with locked facets by skull traction. Validation and evaluation of the nomogram model have demonstrated its good predictive value.
