建立妇科肿瘤患者术前营养状况的列线图评估模型

Establishment of a nomogram model for pre-operative nutritional status of gynecological tumor patients

目的建立妇科肿瘤患者术前营养状况的列线图评估模型。方法选取2016年1月至2019年12月在江南大学附属医院首次进行妇科肿瘤择期手术患者共110例为研究对象,术前3 d采用患者生成的主观整体评估量表(PG-SGA)将患者分为营养良好组44例(0~3分)和营养不良组66例(4~35分)。比较两组患者临床资料和血生化指标。采用多因素Logistic回归分析筛选妇科肿瘤患者营养不良的独立预测因素,在此基础上建立列线图模型。结果单因素比较发现,营养不良组患者年龄增大,进展期肿瘤增多,脂肪百分比、中上臂周长(MUAC)、握力、血红蛋白和白蛋白水平下降;近1个月体重丢失量和百分比、C反应蛋白(CRP)水平升高,差异有统计学意义(P<0.05)。多因素Logistic回归分析显示,进展期肿瘤(OR=3.001,95%CI=2.124~4.125)、近1个月体重丢失(OR=5.201,95%CI=4.526~6.329)和CRP增加(OR=2.012,95%CI=1.320~3.020)、血红蛋白下降(OR=0.562,95%CI=0.232~0.859)和握力下降(OR=0.324,95%CI=0.102~0.667)是妇科肿瘤患者营养不良的独立预测因子(P<0.05)。建立列线图预测模型后采用受试者操作特征曲线(ROC曲线)计算曲线下面积(AUC)为0.889。结论妇科肿瘤患者择期手术前普遍存在营养不良,进展期肿瘤、近1个月体重丢失和CRP增加、血红蛋白和握力下降都是营养不良的独立预测因子;建立可视化效果强、操作简便的定量列线图模型预测营养不良有较高的效能。

Objective To establish a nomogram model for pre-operative nutritional status of patients with gynecological tumors.Method A total of 110 patients underwent elective surgery for gynecological tumors from January 2016 to December 2019 were selected as the research objects,three days before operation,the patients were divided into well-nourished group(0~3 points,n=44)and malnutrition group(4~35 points,n=66)according to the patient generated subjective global assessment scale(PG-SGA).The clinical datas and blood biochemical indexes were compared between the two groups,then multivariate Logistic regression analysis was used to screen the risk factors of malnutrition.Result Univariate comparison found that the patients in malnutrition group were older,more of advanced tumors,less of fat percentage,middle and upper arm circumference(MUAC),handgrip strength,hemoglobin and albumin levels,while higher of weight loss and percentage in recent one month,C-reactive protein(CRP)levels,the difference was statistically significant(P<0.05).Multivariate Logistic regression analysis showed that advanced tumors(OR=3.001,95%CI=2.124-4.125),higher of weight loss in recent 1 month(OR=5.201,95%CI=4.526-6.329)and CRP(OR=2.012,95%CI=1.320-3.020),less of hemoglobin(OR=0.562,95%CI=0.232-0.859)and handgrip strength(OR=0.324,95%CI=0.102-0.667)were the independent predictors to malnutrition in patients with gynecological tumors(P<0.05).The nomogram predictive model was established,and the area under the curve(AUC)of nomogram model for predicting malnutrition by receiver operating curve(ROC curve)was 0.889.Conclusion Malnutrition is common in patients with gynecological tumors before elective surgery,advanced tumors,increase of weight loss in recent one month and CRP,decrease of hemoglobin and handgrip strength are the independent predictors of malnutrition;establishment of a quantitative nomogram model with strong visual effect and simple operation has a high efficiency.