基于检验指标的肺尘埃沉着病患者合并肺炎的影响因素分析及预测模型构建  

Analysis of risk factors and construction of prediction model for pneumonia in patients with pneumoconiosis

作　　者：崔小藏 庄会富[3] 郭毅 徐银海 CUI Xiaocang;ZHUANG Huifu;GUO Yi;XU Yinhai(Clinical Laboratory,the Second Affiliated Hospital of Xuzhou Medical University/General Hospital of Xuzhou Mining Group,Xuzhou,Jiangsu 221006,China;School of Medical Technology,Xuzhou Medical University,Xuzhou,Jiangsu 221004,China;Research Institute of Occupational Health,China University of Mining and Technology,Xuzhou,Jiangsu 221116,China)

机构地区：[1]徐州医科大学第二附属医院/徐州矿务集团总医院检验科,江苏徐州221006 [2]徐州医科大学医学技术学院,江苏徐州221004 [3]中国矿业大学职业健康研究院,江苏徐州221116

出　　处：《检验医学与临床》2024年第16期2345-2350,共6页Laboratory Medicine and Clinic

基　　金：江苏省科学技术厅基础研究计划(自然科学基金)青年基金项目(BK20201013)。

摘　　要：目的 在检验指标中挖掘肺尘埃沉着病患者合并肺炎的影响因素,构建预测模型并评估其可行性。方法 选取2021年9月至2023年8月徐州医科大学第二附属医院收治的243例肺尘埃沉着病患者作为建模组,根据是否合并肺炎,将建模组患者分为非肺炎组130例和肺炎组113例,比较两组患者炎症指标[C反应蛋白(CRP)、白细胞介素(IL)-2、IL-4、IL-6、IL-8和IL-10]、肿瘤指标[癌胚抗原(CEA)、恶性肿瘤生长因子(TSGF)、β_(2)-微球蛋白(β_(2)-MG)和鳞癌相关抗原(SCC)]、肾功能指标(肌酐、尿酸、尿素和视黄醇结合蛋白)、血常规指标(中性粒细胞计数、中性粒细胞百分比、淋巴细胞计数、淋巴细胞百分比和白细胞计数)水平,采用多因素Logistic回归分析肺尘埃沉着病患者合并肺炎的影响因素,并构建预测模型。选取2021年9月至2023年8月徐州医科大学第二附属医院收治的228例肺尘埃沉着病患者作为验证组,绘制受试者工作特征(ROC)曲线分析评估该模型对肺尘埃沉着病患者合并肺炎的预测价值。结果 非肺炎组和肺炎组CRP、IL-6、IL-8、IL-10、_(β_(2)-MG)、肌酐、尿素水平及中性粒细胞百分比、淋巴细胞计数、淋巴细胞百分比比较,差异均有统计学意义(P<0.05)。多因素Logistic回归分析结果显示,CRP、IL-6、IL-8、β_(2)-MG水平升高是肺尘埃沉着病患者合并肺炎的独立危险因素(OR=3.883、1.217、1.215、2.838,P<0.05),将上述4个指标构建肺尘埃沉着病患者合并肺炎预测模型:Logit(P)=1.357×X_(CRP)+0.196×X_(IL-6)+0.195×X_(IL-8)+1.043×X_(β_(2)-MG))-9.772;ROC曲线结果显示CRP、IL-6、IL-8和_(β_(2)-MG)联合预测模型在建模组和验证组中的曲线下面积分别为0.909和0.848,灵敏度分别为86.8%和75.6%,特异度分别为77.9%和85.0%,准确率分别为82.1%和79.7%。结论 CRP、IL-6、IL-8、β_(2)-MG)水平升高是肺尘埃沉着病患者合并肺炎的独立危险因素,基于上述4个指标Objective To explore the risk factors of pneumonia based on laboratory test indexes for patients with pneumoconiosis and construct prediction model to evaluate its feasibility.Methods A total of 243 patients with pneumoconiosis enrolled in the Second Affiliated Hospital of Xuzhou Medical University from September 2021 to August 2023 were selected as the modeling group,which was subdivided into the non-pneumonia group(n=130)and the pneumonia group(n=113)according to whether they were combined with pneumonia or not,and the inflammatory indexes[C-reactive protein(CRP),interleukin(IL)-2,IL-4,IL-6,IL-8 and IL-10],tumor indexes[carcinoembryonic antigen(CEA),malignant tumor growth factor(TSGF),β_(2) microglobulin (β_(2)-MG))and squamous carcinoma-associated antigen(SCC)],renal function indexes(creatinine,uric acid,urea and retinol-binding protein)and routine blood indexes(neutrophil count,neutrophil percentage,lymphocyte count,lymphocyte percentage and white blood cell count)levels in the two groups were compared;multivariate Logistic regression was used to analyze the influencing factors of pneumonia in pneumoconiosis,and a prediction model was constructed.A total of 228 patients with pneumoconiosis admitted to the Second Affiliated Hospital of Xuzhou Medical University from September 2021 to August 2023 were selected as the validation group,and the receiver operating characteristic(ROC)curve was plotted to evaluate the predictive value of the prediction model for pneumonia in patients with pneumoconiosis.Results The differences of CRP,IL-6,IL-8,IL-10,β_(2)-MG,creatinine,urea levels and percentage of neutrophils,lymphocyte count and percentage of lymphocytes between the non-pneumonia group and the pneumonia group were statistically significant(P<0.05).The results of multivariate Logistic regression analysis showed that increased levels of CRP,IL-6,IL-8 and β_(2)-MG were independent risk factors for predicting pneumonia in patients with pneumoconiosis(OR=3.883,1.217,1.215,2.838,P<0.05),and a prediction model for pne

关 键 词：肺尘埃沉着病 肺炎 危险因素 C反应蛋白 白细胞介素-6 白细胞介素-8 β_(2)微球蛋白 预测模型 

分 类 号：R135.2[医药卫生—劳动卫生] R563.1[医药卫生—公共卫生与预防医学]