胸外科手术后医院肺部感染的特征及预测模型建立  

CHARACTERIZATION AND PREDICTION MODEL OF HOSPITAL LUNG INFECTION AFTER THORACIC SURGERY

作　　者：陈云 谢刚[1] 岳雁鸿[1] 曹波 张勇 CHEN Yun;XIE Gang;YUE Yanhong;CAO Bo;ZHANG Yong(Meishan Hospital of Traditional Chinese Medicine,Meishan 620010,China)

机构地区：[1]四川省眉山市中医医院,620010

出　　处：《中国煤炭工业医学杂志》2023年第2期184-189,共6页Chinese Journal of Coal Industry Medicine

基　　金：四川省医学科研课题立项计划项目(编号:S17083)。

摘　　要：目的 探讨胸外科手术后医院肺部感染的特征及预测模型建立。方法 回顾性选取2020年4月—2022年6月收治的28例胸外科手术后医院肺部感染的患者作为病例组,另选取同期胸外科手术后未发生肺部感染患者100例作为对照组,统计分析两组临床病理特征,采用多因素回归分析胸外科手术后医院肺部感染的影响因素,并构建风险预测模型。结果 28例胸外科手术后医院肺部感染患者中,共分离出病原菌23株,革兰氏阳性菌4株(17.39%),革兰氏阴性菌18株(78.26%),真菌1株(4.35%);采用二元Logistic分析,年龄>60岁(OR=5.162,P=0.010)、吸烟(OR=8.004,P=0.003)、糖尿病(OR=4.505,P=0.014)、慢性阻塞性肺病(OR=10.327,P=0.014)、手术时间>4h(OR=9.065,P=0.007)、术中出血量>500 ml(OR=6.218,P=0.009)、机械通气时间>3d(OR=7.273,P=0.003)、胃管置入(OR=4.366,P=0.017)均为影响胸外科手术后医院肺部感染的独立危险因素;构建胸外科手术后医院肺部感染的风险预测模型:P=1/[1+e(-6.442+1.641)×(年龄)+2.080×(吸烟)+1.505×(糖尿病)+2.335×(慢性阻塞性肺病)+2.204×(手术时间)+1.827×(术中出血量)+1.984×(机械通气时间)+1.474×(胃管置入)],采用Hosmer-Lemeshow检验(χ^(2)=8.367,P=0.399),预测模型的AUC为0.901(95%CI:0.849~0.952),表明该模型有较好的拟合度和预测效能(P<0.05)。结论 年龄>60岁、吸烟、糖尿病、慢性阻塞性肺病、手术时间>4h、术中出血量>500ml、机械通气时间>3d、胃管置入均为影响胸外科手术后医院肺部感染的高危因素,临床可进行早期预防和治疗,减少术后肺部感染率。Objective To explore the characteristics and prediction model establishment of hospital lung infection after thoracic surgery. Methods A total of twenty-eight patients with hospital lung infection after thoracic surgery admitted from April 2020 to June 2022 were retrospectively selected as the case group, and 100 patients without lung infection after thoracic surgery in the same period were selected as the control group.The clinicopathological characteristics of the two groups were statistically analyzed.The susceptibility factors of hospital lung infection after thoracic surgery were analyzed by multivariate regression analysis, and the risk prediction model was constructed. Results Among 28 patients with hospital lung infections after thoracic surgery, 23 pathogenic bacteria were isolated, 4 strains Gram-positive(17.39%),18 strains Gram-negative(78.26%),and 1 strain fungal(4.35%).Binary logistic analysis showed that age60 years(OR=5.162,P=0.010),smoking(OR=8.004,P=0.003),diabetes mellitus(OR=4.505,P=0.014),chronic obstructive pulmonary disease(OR=10.327,P=0.014),operative time4 h(OR=9.065,P=0.007),intraoperative bleeding500 ml(OR=6.218,P=0.009),duration of mechanical ventilation3 d(OR=7.273,P=0.003),and gastric tube placement(OR=4.366,P=0.017) were independent risk factors of hospital lung infection after thoracic surgery.The risk prediction model for hospital lung infection after thoracic surgery was as follow: P=1/[1+e(-6.442+1.641)×(age)+2.080×(smoking)+1.505×( diabetes) + 2.335×(COPD) + 2.204×(duration of surgery) + 1.827×(intraoperative bleeding) + 1.984×(duration of mechanical ventilation) +1.474×(gastric tube placement)],its fit was assessed by Hosmer-Lemeshow test(χ^(2)=8.367,P=0.399),and the AUC prediction model was 0.901( 95%CI:0.849 to 0.952),indicating good fit and predictive efficiency(P<0.05). Conclusion Age60 years, smoking, diabetes, chronic obstructive pulmonary disease, operative time4 h, intraoperative bleeding500 ml, duration of mechanical ventilation3 d, and gastric tube placement

关 键 词：胸外科手术 肺部感染 预测模型 

分 类 号：R655[医药卫生—外科学]