摘要
目的构建并验证经尿道输尿管镜取石术难度预测列线图,为临床医生术前评估和术中决策提供依据。方法回顾性纳入2021年6月至2023年6月该院收治的经尿道输尿管镜取石术患者300例,通过logistic回归模型筛选影响手术难度的独立危险因素,并基于回归模型利用R语言构建列线图。内部验证采用自助法估计模型的区分度(C-指数)与校正度。并利用临床决策曲线验证模型的临床获益性。随后于2023年7月至2024年7月前瞻性纳入100例患者作为外部验证队列,评估经尿道输尿管镜取石术难度预测列线图的预测性能。结果300例患者中困难手术80例(26.7%),常规手术220例(73.3%)。结石大小、输尿管壁最大厚度、严重肾积水、既往输尿管手术史、血小板计数、白细胞等是影响手术难度的独立危险因素。基于多因素回归模型构建的经尿道输尿管镜取石术难度预测列线图,回顾性队列内部验证C-指数为0.88,校正曲线显示拟合度良好,受试者工作特征曲线下面积为0.88;前瞻性队列外部验证C-指数为0.85,受试者工作特征曲线下面积为0.82,模型依旧具有良好区分度与校正度。临床获益曲线显示2个模型均具有较高的临床获益性。结论构建的经尿道输尿管镜取石术难度预测列线图可较准确地预测经尿道输尿管镜取石术难度,有助于医生术前识别高风险患者并指导术中决策,后续可在多中心大样本研究中进一步验证与完善。
Objective To develop and validate a nomogram for predicting the difficulty of transurethral ureteroscopic lithotripsy,providing clinicians with a reference for preoperative evaluation and intraoperative decision-making.Methods A total of 300 patients who underwent transurethral ureteroscopic lithotripsy at our hospital from June 2021 to June 2023 were retrospectively included in the development cohort.Independent risk factors for surgical difficulty were screened through logistic regression analysis,and a nomogram was constructed based on the regression model using R.Internal validation was performed using bootstrap resampling to assess the model′s discrimination and calibration.Decision curve analysis was employed to verify the model′s clinical benefit.Subsequently,a prospective external validation cohort of 100 patients from July 2023 to July 2024 was enrolled to evaluate the predictive performance of the nomogram.Results Among the 300 patients,80(26.7%)were classified as having difficult surgeries,while 220(73.3%)underwent conventional procedures.Univariate and multivariate analyses identified that stone size,maximum ureteral wall thickness,severe hydronephrosis,history of previous ureteral surgery,platelet count,and white blood cell(WBC)count were independent risk factors affecting surgical difficulty.Based on the multivariate logistic regression model,a nomogram was developed.In the retrospective cohort,internal validation yielded a C-index of 0.88,with a calibration curve showing good agreement and the area under the receiver operating characteristic curve(AUC)of 0.88.In the prospective external validation cohort,the C-index was 0.85 and the AUC was 0.82,indicating that the model maintained good discrimination and calibration.The clinical benefit curves showed that both models offer substantial clinical net benefit.Conclusion The constructed nomogram can accurately predict the difficulty of transurethral ureteroscopic lithotripsy,aiding in the preoperative identification of high-risk patients and guiding intraoperative decision-making.Further validation and refinement in multicenter studies with larger sample sizes are warranted.
作者
汪有彪
江兴华
张细初
WANG Youbiao;JIANG Xinghua;ZHANG Xichu(Department of Urology Surgery,the Second People′s Hospital of Jingdezhen,Jingdezhen,Jiangxi 333099,China)
出处
《现代医药卫生》
2026年第2期343-348,共6页
Journal of Modern Medicine & Health
基金
江西省中医药管理局科技项目(2023B0291)
江西省景德镇市自筹经费类医药卫生科技项目(20231SFZC052)。
