直肠癌前切除术后腹腔感染危险因素及列线图预测模型的构建

Risk factors and establishment of nomogram prediction model for intra-abdominal infection after anterior resection of rectal cancer

目的研究旨在识别腹腔感染的危险因素,建立一种能准确预测直肠癌前切除术后腹腔感染的模型。方法回顾性收集某院胃肠外科2017年12月—2020年9月行直肠癌前切除术患者的临床及病理资料,采用单因素分析及多因素logistic回归分析筛选出直肠癌前切除术后独立危险因素。基于独立危险因素建立列线图风险预测模型,并对其性能进行评价。结果共纳入247例直肠癌前切除术患者,32例发生腹腔感染,发病率为13.0%。经单因素和多因素logistic回归分析,筛选出糖尿病(OR=4.430,95%CI:1.177~16.668,P<0.05)、手术时间≥180 min(OR=5.502,95%CI:1.459~20.752,P<0.05),以及术后发生吻合口瘘(OR=18.171,95%CI:4.756~69.416,P<0.05)、肺部感染(OR=7.689,95%CI:2.143~27.593,P<0.05)是直肠癌前切除术后发生腹腔感染的独立危险因素。基于此结果建立列线图预测模型,校准曲线显示该列线图模型的预测发生概率和实际发生概率之间一致性较好,Bootstrap法计算出C-index值为0.945,区分度较好,受试者工作特征曲线下面积为0.945(95%CI:0.871~0.971),决策分析曲线显示价值性高。结论基于直肠癌前切除术后患者发生腹腔感染的独立危险因素建立的列线图预测模型,能直观预测术后发生腹腔感染的概率。

Objective To identify the risk factors for intra-abdominal infection(IAI)and establish a prediction model that can accurately predict IAI after anterior resection of rectal cancer.Methods Clinical and pathological data of patients who underwent anterior resection of rectal cancer in department of gastrointestinal surgery of a hospital from December 2017 to September 2020 were collected retrospectively,independent risk factors after anterior resection of rectal cancer were screened out by univariate analysis and multivariate logistic regression analysis.Based on the independent risk factors,nomogram risk prediction model was established,and performance of the model was evaluated.Results A total of 247 patients undergoing anterior resection of rectal cancer were included in the study,32 patients developed IAI,with an infection rate of 13.0%.Univariate and multivariate logistic regression analysis showed that diabetes mellitus(OR=4.430,95%CI:1.177-16.668,P<0.05),operation time≥180 minutes(OR=5.502,95%CI:1.459-20.752,P<0.05),post-operative anastomotic fistula(OR=18.171,95%CI:4.756-69.416,P<0.05),and post-operative pulmonary infection(OR=7.689,95%CI:2.143-27.593,P<0.05)were independent risk factors for IAI after anterior resection of rectal cancer.Nomogram prediction model was established based on above results,calibration curve showed that predicted occurrence probability of the nomogram model was consistent with the actual occurrence probability,C-index value calculated by Bootstrap method was 0.945,indicating a good degree of differentiation,the area under ROC curve was 0.945(95%CI:0.871-0.971),and decision analysis curve showed high value.Conclusion Nomogram prediction model based on the independent risk factors of IAI after anterior resection of rectal cancer in patients can directly predict the occurrence probability of post-operative IAI.