老年女性腹部切口愈合不良的多因素分析及风险预测模型

Multifactorial analysis and risk prediction model of poor abdominal incision healing in elderly women

目的探究老年患者行妇科开腹手术后发生手术切口愈合不良的影响因素,构建风险预测模型。方法回顾性队列研究,收集2021年9月至2023年9月于北京医院行妇科开腹手术的255例老年患者(手术时年龄≥60岁)的围术期资料,根据患者腹部切口愈合情况分为愈合良好组204例和愈合不良组51例,将所有纳人样本按7:3的比例随机拆分为训练集179例和验证集76例。采用单因素分析和多因素Logistic逐步回归分析筛选切口愈合不良的危险因素,并构建列线图模型。采用受试者工作特征(ROC)曲线检验模型预测价值,采用Bootstrap法进行模型的内部验证,绘制校准曲线评估模型的临床应用价值,计算净重新分类指数(NRI)和综合判别改善指数(IDI)比较模型预测性能。结果单因素分析结果显示,年龄、体质指数(BMI)、术前腹壁厚度、术中出血、术前白蛋白、术后最低白蛋白、腹水、疾病性质、围术期补充白蛋白与腹部切口愈合相关(均P<0.05。多因素Logistic逐步回归分析结果显示,围术期补充白蛋白(OR=0.07,95%CI:0.01~0.44,P=0.008)、年龄(OR=1.28,95%CI:1.12~1.52,P<0.001)、术前腹壁厚度(OR=5.75,95%CI:2.81~15.74,P<0.001)、术后最低白蛋白(OR=0.82,95%CI:0.69~0.97,P=0.017)是腹部切口愈合情况的影响因素。ROC曲线分析结果显示,训练集和内部验证集的曲线下面积(AUC)分别为0.982(95%CI:0.967~0.997)和0.961(95%CI:0.906~1.000)。采用Bootstrap自抽样方法对模型进行验证,校准曲线表明该模型预测效果好,NRI和IDI指数表明4项联合指标对切口愈合不良的预测价值高于单一指标。丝结论本研究建立的预测模型(包含年龄、术前腹壁厚度、围术期补充白蛋白、术后最低白蛋白4项指标)经过统计学检验和内部验证,可在围术期预测老年女性发生腹部切口愈合不良的风险,便于临床及时调整高风险患者的治疗。

Objective To investigate the factors influencing the occurrence of poor surgical incision healing after gynecologic open surgery in elderly patients,and to construct a risk prediction model.Methods A retrospective cohort study was conducted to collect perioperative data from 255 elderly patients(age≥60 years at the time of surgery)who underwent gynecologic open abdominal surgery in Beijing Hospital from September 2021 to September 2023,and the patients were divided into 204 cases of the well-healed group and 51 cases of the poorly healed group according to the healing of their abdominal incision conditions,and all the included samples were randomly split into the training set of 179 cases and the validation set of 76 cases according to a ratio of 7:3.Risk factors for poor incision healing were screened using univariate analysis and multifactorial Logistic regression analysis,and a column-line diagram model was constructed.The predictive value of the model was examined using the receiver operating characteristic(ROC)curve,the internal validation of the model was performed using the Bootstrap method,the clinical application value of the model was evaluated by plotting the calibration curve,and the predictive performance of the model was compared by calculating the net reclassification index(NRI)and the integrated discriminant improvement index(IDI).Results The results of univariate analysis showed that age,body mass index(BMI),preoperative abdominal wall thickness,intraoperative bleeding,preoperative albumin,postoperative minimum albumin,ascites,nature of the disease,and perioperative albumin supplementation were correlated with abdominal incision healing(all P<0.05).Multifactorial Logistic regression analysis showed that perioperative albumin supplementation(OR=0.07,95%CI:0.01-0.44,P=0.008),age(OR=1.28,95%CI:1.12-1.52,P<0.001),preoperative abdominal wall thickness(OR=5.75,95%CI:2.81-15.74,P<0.001),and postoperative minimum albumin(OR=0.82,95%CI:0.69-0.97,P=0.017)were the influencing factors of abdominal incision healing.Plotting the ROC curves of the column-line graphical model showed that the area under the curve(AUC)for the training set and the internal validation set were 0.982(95%CI:0.967-0.997)and 0.961(95%CI:0.906-1.000),respectively.The model was validated using the Bootstrap selfsampling method,and the calibration curve showed that the model predicted well with good calibration,and NRI and IDI indicated that the predictive value of the four combined was higher than that of a single index for poor incision healing.Conclusions The prediction model established in this study(including 4 indicators of age,preoperative abdominal wall thickness,perioperative albumin supplementation,and postoperative minimum albumin)has been statistically tested and internally validated to predict the risk of poorly healed abdominal incisions in elderly women in the perioperative period,facilitating timely clinical adjustment of treatment for high-risk patients.