妇科恶性肿瘤患者术后下肢淋巴水肿预警模型的构建与评估

Construction and evaluation of an early warning model for postoperative lower limb lymphedema in patients with gynecological malignant tumors

目的:构建妇科恶性肿瘤患者术后下肢淋巴水肿的预警模型并进行评估。方法:选取在本院进行手术治疗的565例妇科恶性肿瘤患者作为研究对象,根据7:3将其分为建模组(396例)和验证组(169例),再将建模组根据患者是否发生术后下肢淋巴水肿将其分为水肿组(78例)和非水肿组(318例)。对妇科恶性肿瘤患者术后发生下肢淋巴水肿的风险进行单因素分析,并根据单因素分析结果行多因素Logistic回归分析。使用R(R4.0.1),rms程序包以多因素Logistic回归分析结果构建妇科恶性肿瘤患者术后发生下肢淋巴水肿的列线图预警模型,使用Bootstrap重复抽样法绘制Hosmer-Lemeshow校准曲线对模型进行验证,P>0.05表示模型准确性较好。以SPSS 20.0软件绘制ROC曲线评估模型的区分度。结果:本研究中共收集396例妇科恶性肿瘤术后患者,其中78例患者发生下肢淋巴水肿,发生率19.70%。单因素分析显示,术后放疗、淋巴结清扫、TNM分期、术后引流时间、合并糖尿病为影响妇科恶性肿瘤患者术后下肢淋巴水肿发生的因素(P<0.05),下肢淋巴水肿发生与体质量指数、年龄、术后化疗、生活环境、合并高血压等因素无关(P>0.05);进一步行多因素Logistic回归分析,结果显示,术后放疗、淋巴结清扫、TNM分期、术后引流时间、合并糖尿病为影响妇科恶性肿瘤术后患者发生下肢淋巴水肿的独立危险因素(P<0.05)。建模组ROC曲线下面积为0.817,敏感性、特异性分别为66.04%、83.33%;H-L检验的结果为χ^(2)=3.215,P=0.278。验证组ROC曲线下面积为0.773,敏感性66.35%,特异性为84.62%;H-L检验的结果为χ^(2)=3.112,P=0.254,说明该预警模型准确度及区分度较好。结论:术后放疗、淋巴结清扫、TNM分期、术后引流时间、合并糖尿病为影响妇科恶性肿瘤患者术后发生下肢淋巴水肿的独立危险因素,以此建立的预警模型具有较好的区分度及准确度,可帮助临床识别风险较高人群并给予预防、及时治疗。

Objective:To construct an early warning model of postoperative lower limb lymphedema in patients with gynecological malignancies and evaluate it.Methods:A total of 565 patients with gynecological malignant tumors who underwent surgical treatment in our hospital were selected as the research objects.They were divided into modeling group(396 cases)and validation group(169 cases)according to 10:3,and then the modeling group was divided into edema group(78 cases)and non-edema group(318 cases)according to whether the patients had postoperative lower extremity lymphedema.Univariate analysis of the risk factors of lower extremity lymphedema in patients with gynecological malignant tumors was carried out,and multivariate Logisitic regression analysis was performed on the results of the univariate analysis to analyze the risk factors of lower extremity lymphedema in patients with gynecological malignant tumors.R(R4.0.1)and rms package was used to construct a nomogram early warning model for postoperative lower extremity lymphedema in patients with gynecological malignant tumors based on multivariate Logistic regression analysis results.Bootstrap repeated sampling method was used to draw Hosmer-Lemeshow calibration curve for model validation,P>0.05 indicates that the model accuracy was better.SPSS20.0 software was used to draw ROC curve to evaluate the discriminative degree of the model.Results:In this study,a total of 396 patients with gynecological malignant tumors were collected,of which 78 patients had lower limb lymphedema,accounting for 19.70%.Univariate analysis showed that postoperative radiotherapy,lymph node dissection,TNM staging,postoperative drainage time,and diabetes were the factors that affect the occurrence of postoperative lower limb lymphedema in patients with gynecological malignant tumors(P<0.05),but had nothing to do with factors such as body mass index,age,postoperative chemotherapy,living environment,and hypertension(P>0.05);further multivariate Logistic regression analysis was performed,and the results showed that postoperative radiotherapy,lymph node dissection,TNM staging,postoperative drainage time,and diabetes were independent risk factors affecting the occurrence of lower limb lymphedema in patients with gynecological malignant tumors after surgery(P<0.05).The area under the ROC curve of the modeling group was 0.817,and the sensitivity and specificity were 66.04%and 83.33%,respectively;the result of H-L test wasχ^(2)=3.215,P=0.278.The area under the ROC curve of the validation group was 0.773,the sensitivity was 66.35%,and the specificity was 84.62%;the result of H-L test wasχ^(2)=3.112,P=0.254,indicating that the early warning model had good accuracy and discrimination.Conclusion:Postoperative radiotherapy,lymph node dissection,TNM staging,postoperative drainage time,and diabetes are independent risk factors affecting postoperative lower limb lymphedema in patients with gynecological malignancies.The early warning model established by this has good discrimination and accuracy.It can help clinically identify people at higher risk and provide prevention and timely treatment.