目的 探讨椎间孔镜TESSYS技术治疗腰椎间盘突出症术后恢复不良的危险因素,并构建相关预测模型及验证。方法 回顾性分析2019年1月—2021年12月我院收治的行椎间孔镜TESSYS技术治疗的腰椎间盘突出患者,收集患者的所有资料。采用LASSO回归对腰椎间盘突出患者椎间孔镜TESSYS术后恢复不良的危险因素进行筛选,LASSO回归中的调节参数λ采用10折交叉验证方法进行验证,选择λ不为0的变量纳入多因素Logistic回归模型,建立腰椎间盘突出患者椎间孔镜TESSYS术后恢复不良的危险因素预测模型。根据预测模型结果,绘制受试者工作特征（receiver operating characteristic,ROC）曲线,并计算曲线下面积（area under the curve,AUC）。利用Bootstrap重复取样（1000次）对该模型进行内部验证,计算可获得一致性指数（consistency index,C-index）。结果 根据纳入以及排除标准,入组100例患者,术后恢复不佳33例,恢复良好67例。单因素分析提示了8个显著差异的指标,使用LASSO回归降维处理,提示了7项最佳建模指标,并将其进行Logistic回归分析,最终得到了3项指标：年龄、合并糖尿病、合并骨质疏松为恢复不良的独立危险因素,并构建危险因素模型：Logi（P）=-1.159+0.681×年龄+0.291×合并糖尿病+0.428×合并骨质疏松,评估显示该模型的C-index为0.922,AUC为0.901（95% CI为0.863~0.945）,模型预测能力较好。结论 本研究基于LASSO-Logistic建立了椎间孔镜TESSYS技术治疗腰椎间盘突出症术后恢复不良的预测模型,且预测能力良好,临床中具有较好的应用价值。

Objective To explore the risk factors of poor postoperative recovery of lumbar intervertebral disc protrusion treated by transforaminal endoscopic TESSYS technology, and to construct and validate the relevant prediction model. Methods A retrospective analysis was conducted on patients with lumbar intervertebral disc protrusion who underwent transforaminal endoscopic TESSYS technology treatment in our hospital from January 2019 to December 2021, and all the data of the patients were collected. LASSO regression was used to screen the risk factors for poor recovery after endoscopic TESSYS in patients with lumbar intervertebral disc protrusion. The mediating parameter λ in LASSO regression was verified by the 10-fold cross-validation method. Variables with λ not being 0 were selected and included in the multivariate Logistic regression model. To establish a risk factor prediction model for poor recovery after transforaminal endoscopic TESSYS in patients with lumbar intervertebral disc protrusion. Based on the results of the prediction model, the receiver operating characteristic (ROC) curve was plotted, and the area under the curve (AUC) was calculated. The model was internally verified by using Bootstrap for repeated sampling (1000 times), and the consistency index (C-index) could be obtained through calculation. Results According to the inclusion and exclusion criteria, 100 patients were enrolled. Among them, 33 cases had poor postoperative recovery and 67 cases had good recovery. Univariate analysis indicated 8 significantly different indicators. Using LASSO regression dimensionality reduction processing, 7 optimal modeling indicators were indicated, and Logistic regression analysis was conducted on them. Eventually, 3 indicators were obtained: age, combined diabetes, combined osteoporosis, and a risk factor model was constructed: Logi(P)=-1.159+0.681×age+0.291×combined diabetes+0.428×combined osteoporosis. The assessment showed that the C-index of this model was 0.922. The AUC was 0.901(95% CI: 0.863-0.945), and the model had a good predictive ability. Conclusion In this study, a prediction model for poor postoperative recovery of lumbar intervertebral disc protrusion treated by transforaminal endoscopic TESSYS technology was established based on LASSO-Logistic, and it has good predictive ability and good application value in clinical practice.