结直肠癌术后切口感染危险因素及Logistic回归模型

Risk factors for postoperative incision infection in colorectal cancer and logistic regression model

目的探讨结直肠癌术后切口感染危险因素及模型评价。方法回顾性收集2019年10月-2022年10月南京市江宁医院普外科和江苏省肿瘤医院结直肠外科收治的104例结直肠癌术后患者的临床资料,根据切口感染情况分为感染组(26例)和未感染组(78例);采用多因素Logistic回归分析归纳结直肠癌术后切口感染的危险因素,并构建预测模型,采用受试者工作特征(ROC)曲线分析预测模型对结直肠癌术后切口感染的预测价值。结果104例结直肠癌术后患者发生切口感染26例,切口感染发生率为25.00%;多因素Logistic回归分析结果显示,合并糖尿病、引流管放置时间是结直肠癌术后切口感染的危险因素(P<0.05);构建预测模型:Logit(P)=-10.721+合并糖尿病×1.416+引流管放置时间×1.523;ROC曲线分析结果显示,预测模型预测结直肠癌术后切口感染的曲线下面积(AUC)值为0.928,95%CI为0.860~0.969,敏感度为92.31%,特异度为80.77%。结论结直肠癌患者术后感染危险因素包括合并糖尿病、引流管放置时间,据此构建预测模型对结直肠癌术后切口感染具有较好的预测价值。

OBJECTIVE To investigate the risk factors of postoperative incision infection in colorectal cancer and model evaluation.METHODS The clinical data of 104postoperative colorectal cancer patients admitted to the Department of General Surgery Nanjing Jiangning Hospital and Department of Colorectal Surgery Jiangsu Cancer Hospital from Oct.2019to Oct.2022were retrospectively collected,and the patients were divided into infection group(26cases)and non-infection group(78cases)according to the status of incision infection.Multivariate logistic regression analyses were used to summarize the risk factors of postoperative incision infection of colorectal cancer,and the predictive model was constructed.The predictive value of the prediction model for postoperative incision infection of colorectal cancer was analyzed by receiver operating characteristic(ROC)curve.RESULTS Incisional infections occurred in 26cases of 104patients after colorectal cancer surgery,and the incidence of incisional infection was 25.00%.Multivariate unconditional Logistic regression analysis showed that the combination of diabetes mellitus and drainage tube placement time were the risk factors for postoperative incision infection in colorectal cancer(P<0.05).The predictive model was established as follows:Logit(P)=-10.721+diabetes mellitus×1.416+drainage tube placement time×1.523;ROC curve analysis showed that the area under the curve(AUC)of the predictive model for predicting postoperative incision infection in colorectal cancer was 0.928,with a 95%CI of 0.860-0.969,a sensitivity of 92.31%,and a specificity of 80.77%.CONCLUSION Risk factors for postoperative infection in colorectal cancer patients included combined diabetes mellitus and drainage tube placement time,and the prediction model constructed accordingly had good predictive value for postoperative incision infection in colorectal cancer.