|
| 晚期肺癌患者肺部感染病原菌特征及预测模型构建和验证 |
| Characteristics of lung infection pathogens and construction of prediction model for lung infection in advanced lung cancer |
| 投稿时间：2022-10-09 |
| DOI：10.3969/j.issn.1000-0399.2023.04.006 |
| 中文关键词: 晚期肺癌 肺部感染 病原菌 危险因素 预测模型 |
| 英文关键词: Advanced lung cancer Lung infection Pathogen Risk factors Prediction model |
| 基金项目:安徽省科学技术厅基金项目（编号：201904a07020016） |
|
| 摘要点击次数: 878 |
| 全文下载次数: 306 |
| 中文摘要: |
| 目的 探讨晚期肺癌患者肺部感染的病原菌特征以及预测模型的构建。方法 统计分析安庆市第一人民医院2019年3月至2021年12月收治的208例晚期肺癌患者肺部感染发生率、感染病原菌分布特点，K-B扩散纸片法分析病原菌对抗菌药物的耐药性，采用多因素logistic逐步回归分析患者肺部感染的危险因素，并根据其回归系数构建晚期肺癌肺部感染预测模型，使用受试者工作特征（ROC）曲线、Hosmer-Lemeshow检验对预测模型进行效能评估。结果 208例晚期肺癌患者中，并发肺部感染49例（23.56%），革兰阴性菌37株（60.66%）、革兰阳性菌20株（32.79%）、真菌4株（6.56%）；铜绿假单胞菌、鲍曼不动杆菌、肺炎克雷伯菌对亚胺培南的耐药性低，表皮葡萄球菌、金黄色葡萄球菌对万古霉素、氨苄西林的耐药性低。logistic逐步回归分析结果显示，糖尿病、低蛋白血症、KPS评分<80分、放疗及化疗是晚期肺癌患者肺部感染的危险因素（P< 0.05）；患者肺部感染预测模型为P=1/1+exp（-2.686+1.566χ糖尿病+1.838χ低蛋白血症+1.336χKPS评分+1.236χ放疗+0.874χ化疗），预测模型的ROC下面积（AUC）为0.817，且Hosmer-Lemeshow检验P=0.529。结论 晚期肺癌患者肺部感染以革兰阴性菌为主，合并糖尿病、低蛋白血症、Karnofsky评分<80分、放化疗是患者肺部感染的危险因素，基于上述因素建立的预测模型有较高的预测效能。 |
| 英文摘要: |
| Objective To investigate the characteristics of lung infection pathogens and construction of prediction model in patients with advanced lung cancer. Methods The incidence of lung infection and distribution characteristics of pathogens were statistically analyzed in the 208 patients with advanced lung cancer admitted to Anqing First People’s Hospital between March 2019 and December 2021. The drug resistance of pathogens to antibiotics was analyzed by Kirby-Bauer (K-B) method. The risk factors of lung infection were analyzed by multivariate logistic stepwise regression analysis, and the prediction model for lung infection was constructed based on their regression coefficients. The predictive efficiency of the model was evaluated by receiver operating characteristic (ROC) curves and Hosmer-Lemeshow test. Results In the 208 patients with advanced lung cancer, there were 49 cases (23.56%) with lung infection. There were 37 strains of Gram-negative bacteria (60.66%), 20 strains of Gram-positive bacteria (32.79%) and 4 strains of fungi (6.56%). The drug resistance of Pseudomonas aeruginosa, Acinetobacter baumannii and Klebsiella pneumoniae was low to imipenem, and drug resistance of Staphylococcus epidermidis and Staphylococcus aureus was low to vancomycin and ampicillin. The results of logistic stepwise regression analysis showed that diabetes, hypoproteinemia, KPS score <80 points, radiotherapy and chemotherapy were independent risk factors of lung infection in patients with advanced lung cancer (P< 0.05). The prediction model for lung infection was as follows: P=1/1+exp (-2.686+1.566 χdiabetes+1.838 χhypoproteinemia +1.336χKPS score +1.236 χradiotherapy +0.874 χchemotherapy). The area under the ROC curve (AUC) of the prediction model was 0.817 (P=0.529 in Hosmer-Lemeshow test). Conclusions The main pathogen of lung infection is Gram-negative bacteria in patients with advanced lung cancer. Diabetes, hypoproteinemia, KPS score <80 points, radiotherapy and chemotherapy are independent risk factors of lung infection. The prediction model for lung infection of advanced lung cancer constructed based on the above factors has high predictive efficiency. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
