摘要
目的 构建颈椎骨折急性脊髓颈损伤患者(ACSCI)手术预后的列线图预测模型并进行验证.方法 选择2016年1月至2022年6月在该院接受手术治疗的323例ACSCI患者资料进行回顾性分析,患者按照7:3的比例随机分为训练队列和验证队列,采用卡方检验比较队列之间的变量差异,采用单因素和多因素Logistic回归模型筛选与预后相关的变量,并利用最终筛选的变量构建预测ACSCI术后6个月预后的列线图模型,采用受试者工作曲线(ROC)和曲线下面积(AUC)评估模型的辨别能力,校准曲线(CC)评估一致性,决策分析曲线(DCA)评估模型的临床受益情况.结果 预后不良发生率33.12％.训练队列(226例)和验证队列(97例)的临床资料差异均无统计学意义(P＞0.05).单因素和多因素Logistic回归分析显示,男性、年龄≥45岁、高能量损伤、上颈椎损伤、髓内T2信号长度(IMLL)＞6 cm、ASIA-B级、ASIA-A级和晚期手术与ACSCI患者6个月预后存在相关性.校准曲线和标准曲线贴合度较好,DCA曲线显示列线图在预测ACSCI患者术后临床预后不良方面具有良好的临床适用性.ROC曲线分析显示,训练队列和验证队列的AUC分别为0.829和0.811.结论 基于性别、年龄、损伤机制、损伤节段、IMLL、ASIA分期和手术时机等风险因素构成的列线图可有效预测ACSCI患者的手术预后.
Abstract
Objective To construct a column chart prediction model for surgical prognosis in patients with acute cervical spinal cord injury(ACSCI)after cervical spine fracture and validate it.Methods A retrospective analysis was conducted on the data of 323 ACSCI patients who underwent surgical treatment in our hospital from January 2016 to June 2022.Patients were randomly divided into a training cohort and a validation cohort in a 7:3 ratio.Chi-square test was used to examine the differences in variables between the queues,and factor and multivariate logistic regression were used to screen variables related to prognosis.A column chart model was constructed to predict the 6-month prognosis of ACSCI using the final screened variables,the discriminant ability of the model was eval-uated using the receiver operating curve(ROC)and area under curve(AUC),consistency was evaluated using the calibration curve(CC),and clinical benefits were evaluated using the decision analysis curve(DCA).Results The incidence of poor prognosis is 33.12％.There was no statistically significant difference in clinical data between the training queue(226 cases)and the validation queue(97 cases)(P＞0.05).There was no statistically significant difference in clinical data between the training queue(226 cases)and the validation queue(97 cases)(P＞0.05).Univariate and multivariate logistic regression analysis showed that there was a corre-lation between male,age≥45 years,high-energy injury,upper cervical spine injury,IMLL＞6 cm,ASIA-B grade,ASIA-A grade,and advanced surgery and the 6-month prognosis of ACSCI patients.The calibration curve and standard curve had a good fit,and the DCA curve showed that column chart had good clinical applicability in predicting poor postoperative clinical prognosis in ACSCI pa-tients.The AUC of the ROC analysis results for the training queue and validation queue were 0.829 and 0.811,respectively.Conclu-sions A column chart based on risk factors such as gender,age,injury mechanism,injury segment,IMLL,ASIA stage,and surgical timing can effectively predict the clinical prognosis of ACSCI patients.
NETL
NSTL
万方数据