目的 构建宫颈癌术后患者下肢淋巴水肿（lower limb lymphedema，LLL）风险预测模型，并验证其预测性能，为医护人员早期识别LLL发生风险提供依据。方法 回顾性选取2010-2020年某院759例宫颈癌术后患者为研究对象，通过单因素和多因素Logistic回归分析筛选危险因素，建立Bayes判别分析及列线图模型；受试者工作特征曲线（receiver operating characteristic curve，ROC）和曲线下面积（area undercurve，AUC）评估模型预测能力，敏感度和特异度评估模型预测价值。结果 Bayes判别分析和列线图模型的AUC分别为91.30%、95.00%，敏感度分别为83.05%、87.60%，特异度分别为85.02%、90.01%。结论 构建的宫颈癌术后LLL风险预测列线图模型预测性能较好，能为临床识别高风险患者提供参考。

Objective To construct a risk prediction model of lower limb lymphedema (LLL) after cervical cancer surgery and validate its prediction performance,so as to provide a basis for healthcare workers to recognize the risk of LLL at an early stage.Methods A selection of 759 postoperative cervical cancer patients in a hospital from 2010 to 2020 were retrospectively selected as the study subjects,and the risk factors were screened by unifactorial and multifactorial logistic regression analyses,and Bayes discriminant analysis and nomogram models were established;the predictive ability of the model was assessed by the receiver operating characteristic curve（ROC）of the subjects and the area undercurve（AUC）,and the predictive value of the model was assessed by the sensitivity and the specificity.Results The AUC for the Bayes discriminant analysis and the nomogram model was 91.30% and 95.00%,the sensitivity was 83.05% and 87.60%,and the specificity was 85.02% and 90.01%.Conclusions The constructed column-line graph model for predicting the risk of LLL after cervical cancer surgery has good predictive performance and can provide a reference for the clinical identification of high-risk patients.

R473.73

国家自然科学基金项目（82003313）；湖南省2023度卫生健康适宜技术推广项目（202319010063）；湖南省中医药科研课题项目（B2023071）