踝关节骨折术后手术部位感染的危险因素分析及预测模型构建

Analysis of risk factors and construction of a prediction model for surgical site infection after ankle fracture surgery

目的分析踝关节骨折患者手术部位发生感染的危险因素,并以此建立列线图预测模型。方法回顾性分析2018年1月—2021年11月三亚中心医院骨关节外科行手术治疗的踝关节骨折患者457例,男性241例,女性216例;年龄18~75岁,平均49.6岁。根据患者术后手术部位是否发生感染分为感染组(52例)和未感染组(405例)。收集所有患者年龄、性别、体质量指数(body mass index,BMI)、吸烟、饮酒、合并糖尿病、致伤机制、麻醉方式等临床资料,应用单因素分析和Logistic回归分析手术部位感染的危险因素。采用R语言包绘制预测患者手术部位感染的列线图模型;用校准曲线、受试者工作特征(receiver operating characteristic,ROC)曲线判断预测效果。结果单因素分析结果显示,患者BMI、合并糖尿病、合并慢性阻塞性肺病、应用免疫抑制剂、应用糖皮质激素、手术时间、手术切口类型、美国麻醉医师协会分级、开放性骨折与手术部位感染相关(P<0.05)。Logistic回归分析显示,合并糖尿病(OR=2.992)、手术时间(OR=1.125)、手术切口类型为Ⅱ~Ⅲ类(OR=2.383)、开放性骨折(OR=1.965)、合并慢性阻塞性肺病(OR=1.065)、应用免疫抑制剂(OR=1.198)、应用糖皮质激素(OR=1.145)为踝关节骨折患者手术部位感染的独立危险因素(P<0.05)。列线图校准曲线斜率接近于1,ROC曲线下面积为0.722(95%CI:0.652~0.793),霍斯莫-莱梅肖(Hosmer-Lemeshow)拟合优度检验χ^(2)=7.565,P=0.428。结论合并糖尿病、手术时间、手术切口类型、合并慢性阻塞性肺病、应用免疫抑制剂、应用糖皮质激素及开放性骨折为踝关节骨折患者手术部位感染的独立危险因素,在此基础上建立的列线图预测模型可有效在术前及术后早期识别高危人群。

Objective To analyze the risk factors of postoperative surgical site infection in patients with ankle fractures,and to establish a nomogram model to predict the onset of postoperative surgical site infection.Methods A total of 457 patients with ankle fractures treated at the department of bone and joint surgery of Sanya Central Hospital from Jan.2018 to Nov.2021 were analyzed retrospectively.The age ranged from 18 to 75 years,with an average of 49.6 years.There were 241 males and 216 females.According to whether infection occurred at the surgical site,patients were divided into infection group(n=52)and non-infection group(n=405).Clinical data of all patients including age,gender,body mass index(BMI),smoking,drinking,diabetes mellitus,injury causes and mode of anesthesia were collected.Univariate analysis and Logistic regression were used to analyze the risk factors of surgical site infection.The R language package was used to establish a nomogram model to predict the surgical site infection,and the calibration curve and receiver operating characteristic(ROC)curve were used to judge the prediction effect.Results Univariate analysis showed that BMI,combination of diabetes mellitus or chronic obstructive putmonary disease(COPD),use of immunosuppressive drugs or glucocorticoids,operation time,surgical incision types,American Society of Anesthesiologists grade,and open fracture were associated with postoperative surgical site infection(all P<0.05).Logistic regression analysis further showed that the combination of diabetes mellitus(OR=2.992)or COPD(OR=1.065),operation time(OR=1.125),type Ⅱ/Ⅲ surgical incision(OR=2.383),open fracture(OR=1.965),and use of immunosuppressive agents(OR=1.198)or glucocorticoids(OR=1.145)were independent risk factors for postoperative surgical site infection in ankle fracture patients(all P<0.05).The slope of the nomogram calibration curve was close to 1,the area under the ROC curve was 0.722(95%CI=0.652-0.793),and the Hosmer-Lemeshow goodness of fit test wasχ^(2)=7.565,P=0.428.Conclusion Combination of diabetes mellitus/COPD,operation time,types of surgical incision,use of immunosuppressive agents/glucocorticoids,and open fracture are independent risk factors for postoperative surgical site infection in patients with ankle fractures.The nomogram prediction model based on the above-mentioned risk factors can identify the high-risk patients early and effectively.