传统开放腰椎椎体间融合术后切口并发症列线图预测模型的建立

Establishment of a line chart prediction model for incisional complications after traditional open lumbarinterbody fusion

目的建立传统开放腰椎椎体间融合术(lumbar interbody fusion,LIF)后发生切口并发症的列线图预测模型,以期为临床围术期管理提供参考。方法回顾性分析2015年9月至2020年8月于我院住院并接受传统开放LIF患者的临床资料,分析相关危险因素,通过单因素、多因素logistic回归分析结果建立列线图预测模型。对预测模型的预测效能、特异度、灵敏度进行评价。结果本研究426例患者共出现术后切口并发症65例,发生率为15.26%。单因素分析及多因素非条件logistic回归分析结果显示,年龄、BMI、术前白蛋白、手术时间、手术节段数是传统开放LIF术后切口并发症的独立危险因素(OR=2.290、2.679、3.524、2.740、3.330,P<0.05)。建立列线图预测模型,计算ROC曲线下面积0.809(95%CI:0.768~0.845,P<0.001),约登指数为0.549,特异度为0.795,灵敏度为0.754。结论本研究建立的列线图模型预测效能较好,为临床早期识别高危人群、进行早期干预提供有效参考。

Objective To establish a line chart prediction model of incision complications after traditional open lumbar interbody fusion(lumbar interbody fusion,LIF),in order to provide reference for clinical periopera⁃tive management.Methods The clinical data of patients with traditional open LIF hospitalized in our hospital from September 2015 to August 2020 were analyzed retrospectively,and the related risk factors were analyzed.The line chart prediction model was established by the results of univariate and multivariate logistic regression analysis.the prediction efficiency,specificity and sensitivity of the prediction model are evaluated.Results In this study,65 cases of postoperative incision complications occurred in 426 patients,with an incidence of 15.26%.The results of univariate analysis and multivariate unconditional Logistic regression analysis showed that age,BMI,preoperative albumin,operation time and the number of operative segments were independent risk factors for incisional compli⁃cations after traditional open LIF,and their odds ratios were 2.290,2.679,3.524,2.740 and 3.330,respectively(P<0.05).The prediction model of line chart was established,and the area under ROC curve was calculated(AUC=0.809,95%CI:0.768~0.845,P<0.001).The Jordan index,specificity and sensitivity were 0.549,0.795 and 0.754 respectively.Conclusions The prediction efficiency of the line chart model established in this study is good,which provides an effective reference for early clinical identification of high⁃risk groups and early intervention.