出 处:《肿瘤研究与临床》2025年第2期87-92,共6页Cancer Research and Clinic
摘 要:目的探讨子宫颈癌患者术后下肢淋巴水肿发生的危险因素,并构建列线图预测模型。方法回顾性病例对照研究。采用便利抽样法,选择资阳市人民医院2014年5月至2022年5月750例子宫颈癌患者。将纳入患者按照7∶3的比例分为建模组(525例)及验证组(225例),建模组根据术后是否发生下肢淋巴水肿将患者分为发生组和未发生组。自行设计基线资料调查表统计患者临床资料。采用logistic逐步回归分析法分析子宫颈癌患者术后下肢淋巴水肿发生的影响因素;采用R 4.0.2软件,基于独立危险因素建立预测子宫颈癌患者术后下肢淋巴水肿发生风险的列线图模型;采用受试者工作特征(ROC)曲线评价列线图预测模型的区分度;绘制校准曲线评价列线图预测模型的一致性。结果525例建模组子宫颈癌患者中,109例(20.76%)发生术后下肢淋巴水肿,下肢淋巴水肿发生组和未发生组不同年龄、肿瘤分期、手术方式、术后引流时间、持续站立时间、淋巴结清扫数目,以及是否合并糖尿病、术后是否化疗、术后是否放疗、是否剧烈运动或体力劳动的患者比例比较,差异均有统计学意义(均P<0.05)。多因素logistic回归分析结果显示,年龄≥60岁(OR=2.548,95%CI:1.253~5.181)、术后放疗(OR=8.048,95%CI:3.867~16.749)、术后引流时间≥5 d(OR=2.664,95%CI:1.043~6.803)、持续站立时间≥1 h(OR=3.024,95%CI:1.252~7.305)、淋巴结清扫数目>20个(OR=7.063,95%CI:1.595~31.281)、合并糖尿病(OR=2.617,95%CI:1.200~5.711)、剧烈运动或体力劳动(OR=3.518,95%CI:1.841~6.721)均是影响子宫颈癌患者术后下肢淋巴水肿发生的独立危险因素(均P<0.05)。基于以上危险因素构建预测子宫颈癌患者术后下肢淋巴水肿发生风险的列线图模型,并进行内外部验证,结果显示建模组和验证组校准曲线和理想曲线拟合度均较好(建模组:χ^(2)=7.87,P=0.446;验证组:χ^(2)=7.92,P=0.441);ROC曲线的曲线下Objective To explore the risk factors for postoperative lower extremity lymphedema in patients with cervical cancer and construct a nomogram prediction model.MethodsA retrospective case-control study was conducted.Using convenience sampling method,750 cervical cancer patients in Ziyang People's Hospital from May 2014 to May 2022 were selected.The included patients were divided into a modeling group(525 cases)and a validation group(225 cases)according to a ratio of 7∶3.In the modeling group,patients were divided into a developing group and a non-developing group according to the incidence of postoperative lower extremity lymphedema.Self-designed baseline data questionnaire was used to collect clinical data of patients.Logistic stepwise regression analysis was used to analyze the influencing factor for the incidence of postoperative lower extremity lymphedema in cervical cancer patients.R 4.0.2 software was used to construct a nomogram model for predicting the incidence risk of postoperative lower extremity lymphedema in cervical cancer patients based on the independent risk factors.Receiver operating characteristic(ROC)curve was used to evaluate the differentiation of the nomogram prediction model.Calibration curves were drawn to evaluate the consistency of the nomogram prediction model.ResultsAmong 525 cervical cancer patients in the modeling group,109(20.76%)had postoperative lower extremity lymphedema.There were statistically significant differences in the proportions of patients with different ages,tumor stages,surgical methods,postoperative drainage time,continuous standing time,number of lymph node dissection,and with or without diabetes mellitus,postoperative chemotherapy,postoperative radiotherapy,strenuous exercise or physical labor between the developing group and the non-developing group(all P<0.05).Multivariate logistic regression analysis showed that age≥60 years old(OR=2.548,95%CI:1.253-5.181),postoperative radiotherapy(OR=8.048,95%CI:3.867-16.749),postoperative drainage time≥5 d(OR=2.664,95%CI
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...