基于术前血红蛋白浓度构建列线图模型预测结直肠癌术后手术部位感染的研究

A nomogram based on preoperative hemoglobin levels for the prediction of postoperative surgical site infection in patients with colorectal cancer

目的构建并验证结直肠癌患者手术部位感染(surgical site infection,SSI)发生风险的列线图预测模型。方法回顾性分析西安交通大学第一附属医院2018年6月至2019年8月择期在全麻下行结直肠癌手术的患者,通过计算机随机生成的分配顺序按照7∶3的比例将其分为建模组和验证组。经单因素和多因素Logistic回归筛选独立危险因素,基于这些危险因素建立列线图模型预测术后发生SSI的风险。通过C指数(concordance index,C-index)验证模型的区分度,受试者工作特征(receiver operating characteristic,ROC)曲线验证模型的预测性能,Calibration校正曲线验证模型的一致性,并通过决策曲线分析(decision curve analysis,DCA)以确定模型的临床有效性。结果本研究共纳入413例结直肠癌手术患者[其中40例(9.7%)发生SSI],建模组292例,验证组121例。经过单因素和多因素Logistic回归分析,结果显示男性、BMI≥28.0 kg/m^(2)、糖尿病史、术前新辅助化疗、开腹手术、术前中度贫血为结直肠癌患者术后SSI的独立危险因素。通过整合这6个因素构建了一个预测术后SSI的列线图模型。该列线图在建模组中C-index为0.868(95%CI 0.804~0.932),在验证组中C-index为0.860(95%CI 0.789~0.931)。同时,该模型在建模组和验证组的曲线下面积(area under the curve,AUC)分别为0.862(95%CI 0.800~0.924)和0.873(95%CI 0.806~0.941)。Calibration校正曲线在两组中显示模拟曲线和实际曲线走势基本一致。建模组及验证组DCA显示,当阈值概率分别在1%~74%和1%~80%时,该列线图模型能产生更好的临床效益。结论术前中度贫血是结直肠癌人群术后发生SSI的独立危险因素。通过整合此危险因素和其他临床危险因素构建了一个结直肠癌患者术后发生SSI的列线图预测模型并表现出良好的预测性能。该列线图模型能够给外科医师和麻醉医师提供一个更加准确的风险评估,有助于尽早发现高风险人群,采取个体化预防方案,促进围手术期加速康复外科的实施。

Objective To construct and validate a nomogram for the prediction of postoperative surgical site infection(SSI)in patients with colorectal cancer.Methods Patients who underwent colorectal cancer surgery under general anesthesia in the First Af‑filiated Hospital of Xi'an Jiaotong University from June 2018 to August 2019 were selected and their data were retrospectively ana‑lyzed.The patients were randomly divided into a modelling group and a validation group at a ratio of 7∶3 based on computer‑generated random numbers.Univariate and multivariate logistic regression analyses were used to determine the independent risk factors.Based on these factors,a nomogram model was established to predict the risks of SSI.The discriminative ability of the nomogram was evaluat‑ed by concordance index(C‑index),while the predictive performance of the nomogram was verified by a receiver operating characteris‑tic curve(ROC).The consistency of the model was evaluated by a calibration plot.Decision curve analysis(DCA)was performed to eval‑uate the clinical effectiveness of the current model.Results A total of 413 patients with colorectal cancer were included in the study,where 40 patients(9.7%)were diagnosed,including 292 patients in the modelling group and 121 patients in the validation group.According to univariate and multivariate logistic regression analyses,male,body mass index(BMI)≥28.0 kg/m^(2),diabetes history,preoper‑ative neoadjuvant chemotherapy,open surgery and preoperative moderate anemia were the independent risk factors for postoperative SSI in patients with colorectal cancer.After incorporating the six factors,the nomogram was constructed to predict postoperative SSI.This nomogram achieved good C‑index of 0.868(95%CI 0.804‒0.932)for the modeling group and 0.860(95%CI 0.789‒0.931)for the validation group.Meanwhile,the area under the curve(AUC)of ROC was 0.862(95%CI 0.800‒0.924)for the modeling group and 0.873(95%CI 0.806‒0.941)for the validation group.The calibration curve displayed a general consistency between the modeling and actual curves.According to the DCA of the modeling and actual groups,the nomogram had a better clinical benefit when the probability threshold was set at 1%‒74%and 1%‒80%,respectively.Conclusions Preoperative moderate anemia is an independent risk factor for developing SSI in colorectal cancer patients.Incorporating both preoperative hemoglobin levels and other clinical risk factors,the novel nomogram is established to predict the risks of postoperative SSI for patients with colorectal cancer and good predictive perfor‑mance was showed.This nomogram provides surgeons and anesthesiologists with more accurate risk assessment,identify high‑risk peo‑ple as soon as possible,and adopt individualized prevention programs to promote the implementation of enhanced recovery after surgery.