预测宫颈癌合并糖尿病患者术后切口感染风险的Nomogram模型建立

Construction of Nomogram model for predicting the risk of postoperative incisional infection in patients with cervical cancer combined with diabetes mellitus

目的旨在构建预测宫颈癌合并糖尿病患者根治性子宫切除术术后切口感染风险的列线图模型,并进行模型评估。方法选取在本院接受根治性子宫切除术治疗的宫颈癌合并糖尿病患者267例作为研究对象,其中发生术后切口感染(感染组)97例,未发生术后切口感染(未感染组)170例。采用单因素和多因素Logistic回归筛选与术后切口感染相关的独立影响因素。采用R软件和相关程序包绘制列线图模型。结果与未感染组相比,感染组年龄≥55岁、体重指数(body mass index,BMI)≥25、引流管留置时间≥7 d、住院时间≥14 d、血清白蛋白<30 g/L的患者比例明显增加(P<0.05)。与未感染组相比,感染组血清白细胞介素6(interleukin-6,IL-6)、肿瘤坏死因子α(tumor necrosis factor-α,TNF-α)、C反应蛋白(C-reactive protein,CRP)和降钙素原(procalcitonin,PCT)水平显著升高(P<0.05)。多变量Logistic回归分析表明,年龄≥55岁、BMI≥25、引流管留置时间≥7 d、住院时间≥14 d,以及血清IL-6、TNF-α、CRP和PCT水平升高均是术后切口感染发生的独立危险因素(P<0.05)。用于评估宫颈癌合并糖尿病患者术后切口感染发生风险的列线图具有良好的预测准确性(C指数为0.947)、区分度(受试者工作特征曲线下面积为0.947)、一致性(Hosmer-Lemeshow拟合优度检验平均绝对误差为0.011)和临床效能。结论本研究基于围手术期特征构建了预测宫颈癌合并糖尿病患者术后切口感染的列线图模型,该模型可能有助于加强感染控制意识,为根治性子宫切除术后高危患者的管理提供参考。

Objective To construct a nomogram model to predict the risk of postoperative incisional infection in patients with cervical cancer combined with diabetes mellitus(DM)undergoing radical hysterectomy and to evaluate the model.Methods A total of 267 patients with cervical cancer and DM who received radical hysterectomy in our hospital were selected as the research subjects,including 97 patients with postoperative incision infection(infection group)and 170 patients without infection(uninfection group).Univariate and multivariate logistic regression analyses were used to screen for independent influencing factors related to postoperative incision infection.A nomogram model was plotted using R software and related packages.Results Compared with the uninfection group,the proportion of patients with age≥55 years,body mass index(BMI)≥25,retention time of drainage tube≥7 days,length of hospital stay≥14 days,and serum albumin<30 g/L in the infection group was significantly increased(P<0.05).Compared with the uninfection group,the infection group had significantly increased levels of serum interleukin-6(IL-6),tumor necrosis factor-α(TNF-α),C-reactive protein(CRP)and procalcitonin(PCT)(P<0.05).Multivariate logistic regression analysis showed that age≥55 years,BMI≥25,retention time of drainage tube≥7 days,length of hospital stay≥14 days,as well as elevated levels of serum IL-6,TNF-α,CRP and PCT were independent risk factors for postoperative incision infection(P<0.05).The nomogram used to assess the risk of postoperative incision infection in patients with cervical cancer and DM had good prediction accuracy(C index was 0.947),discrimination[area under the receiver operating characteristic(ROC)curve(AUC)was 0.947],consistency(mean absolute error of Hosmer Lemeshow goodness of fit test was 0.011),and clinical efficacy.Conclusion In this study,a nomogram model for predicting postoperative incisional infection in patients with cervical cancer combined with DM was constructed based on perioperative characteristics.The model may help to enhance awareness of infection control and provide a reference for the management of high-risk patients after radical hysterectomy.