基于接尘工人职业健康监测的肺纤维化判别模型研究
|
摘要
目的通过分析接尘工人肺纤维化影响因素,建立接尘工人肺纤维化的Bayes判别模型,为动态监测接尘工人肺纤维化提供参考。方法采用2015-10/2017-05四川省重点职业病职业健康检查监测数据,提取其中接触煤尘有害因素的职业健康体检信息进行分析,采用Logistic回归分析筛选接尘工人肺纤维化的影响因素,构建Bayes判别模型,对接尘工人肺纤维化进行预测。结果根据监测数据,Logistic回归筛选出影响接尘工人肺纤维化有统计学意义的因素有:接触所监测危害因素工龄、血压舒张压、肺功能FEV1/FVC、心电图、企业类型、行业、年龄,综合专业知识、专家意见及判别分析特点,将总工龄、肺功能FVC、肺功能FEV1也纳入Bayes判别模型。模型经回代检验,构建的Bayes判别模型预测准确率为71.1%。ROC曲线下面积为0.801(P<0.05)。结论 Bayes判别模型可以作为动态监测接尘工人肺纤维化的辅助方法,提供划定重点监护人群的参数信息,为早期预测接尘工人肺纤维化提供参考。
Objective To analyze the influencing factors and establish Bayes discriminant model so as to provide reference for dynamic monitoring of lung fibrosis in workers exposed to dust. Methods The monitoring data of occupational health examination for key occupational diseases in Sichuan Province from October 2015 to May 2017 were used to analyze the harmful factors by Logistic regression and construct a Bayesian discriminant model to predict lung fibrosis in workers exposed to dust. Results According to the monitoring data,Logistic regression analysis showed that statistically significant factors affecting lung fibrosis in dust exposed workers included length of service exposed to the hazards( years),diastolic blood pressure,lung function FEV1/FVC,ECG,the enterprise type,profession and age.Considering expertise, expert opinions and the characteristics of discriminative analysis, the total length of service,pulmonary function FVC and FEV1 were also included in the Bayes discriminant model.According to return test,the prediction accuracy was 73. 8 %. The area under the ROC curve was 0. 801( P < 0. 05) and statistically significant. Conclusion The Bayes discriminant model can be used as an auxiliary method for dynamic monitoring of pulmonary fibrosis in dust exposed workers,and provide information about the parameters of the key care population and reference for early prediction of lung fibrosis in dust exposed workers.
Objective To analyze the influencing factors and establish Bayes discriminant model so as to provide reference for dynamic monitoring of lung fibrosis in workers exposed to dust. Methods The monitoring data of occupational health examination for key occupational diseases in Sichuan Province from October 2015 to May 2017 were used to analyze the harmful factors by Logistic regression and construct a Bayesian discriminant model to predict lung fibrosis in workers exposed to dust. Results According to the monitoring data,Logistic regression analysis showed that statistically significant factors affecting lung fibrosis in dust exposed workers included length of service exposed to the hazards( years),diastolic blood pressure,lung function FEV1/FVC,ECG,the enterprise type,profession and age.Considering expertise, expert opinions and the characteristics of discriminative analysis, the total length of service,pulmonary function FVC and FEV1 were also included in the Bayes discriminant model.According to return test,the prediction accuracy was 73. 8 %. The area under the ROC curve was 0. 801( P < 0. 05) and statistically significant. Conclusion The Bayes discriminant model can be used as an auxiliary method for dynamic monitoring of pulmonary fibrosis in dust exposed workers,and provide information about the parameters of the key care population and reference for early prediction of lung fibrosis in dust exposed workers.
引文
[1]黄昭维,金焱.煤工尘肺防治的研究进展[J].职业卫生与病伤,2005,20(2):110-112.
[2]张志平,陈小刚,曾永平,等.丰城矿务局煤工尘肺发病情况分析[J].临床医药实践,2016,25(2):119-120,160.
[3]李润根,焦新强,高鑫.尘肺普查高千伏X线胸片质量的相关影响因素[J].职业与健康,2011,27(20):2307-2308.
[4]高蒙,张旭峰,刘权芳.基于R软件的Bayes判别分析在医学中的应用[J].数学的实践与认识,2014,44(22):120-124.
[5]周明珍,赵焕虎,杨华.Bayes判别分析在尘肺诊断中的应用[J].中国预防医学杂志,2008,9(1):78-79.
[6]王瑞芝,郭俊琴.煤工尘肺的研究进展[J].中国城乡企业卫生,2004,100(2):14-17.
[7]何凤生.贯彻职业病防治法,迎接经济全球化的挑战[J].中华劳动卫生职业病杂志,2002,20(3):161-164.
[8]冯佳洁,刘晨光,高玉洁,等.近5年我国尘肺研究论文的文献计量分析[J].中国工业医学杂志,2010,23(4):299-300.
[9]Knorr AL,Jain R,Srivastava R.Bayesian-based selection of metabolic objective functions[J].Bioinformatics,2007,23(3):351-357.
[10]陈雪莲.贝叶斯判别分析在财务预警中的应用[J].经营管理,2008,11:91-92.
[11]Jensen DB,Hogeveen H,De VA.Bayesian integration of sensor information and a multivariate dynamic linear model for prediction of dairy cow mastitis[J].Journal of Dairy Science,2016,99(9):7344-7361.
[12]金建刚,陈虎,刘丽辉,等.Bayes判别分析全血细胞减少的分类诊断[J].中国组织工程研究与临床康复,2009,13(6):1139-1142.
[2]张志平,陈小刚,曾永平,等.丰城矿务局煤工尘肺发病情况分析[J].临床医药实践,2016,25(2):119-120,160.
[3]李润根,焦新强,高鑫.尘肺普查高千伏X线胸片质量的相关影响因素[J].职业与健康,2011,27(20):2307-2308.
[4]高蒙,张旭峰,刘权芳.基于R软件的Bayes判别分析在医学中的应用[J].数学的实践与认识,2014,44(22):120-124.
[5]周明珍,赵焕虎,杨华.Bayes判别分析在尘肺诊断中的应用[J].中国预防医学杂志,2008,9(1):78-79.
[6]王瑞芝,郭俊琴.煤工尘肺的研究进展[J].中国城乡企业卫生,2004,100(2):14-17.
[7]何凤生.贯彻职业病防治法,迎接经济全球化的挑战[J].中华劳动卫生职业病杂志,2002,20(3):161-164.
[8]冯佳洁,刘晨光,高玉洁,等.近5年我国尘肺研究论文的文献计量分析[J].中国工业医学杂志,2010,23(4):299-300.
[9]Knorr AL,Jain R,Srivastava R.Bayesian-based selection of metabolic objective functions[J].Bioinformatics,2007,23(3):351-357.
[10]陈雪莲.贝叶斯判别分析在财务预警中的应用[J].经营管理,2008,11:91-92.
[11]Jensen DB,Hogeveen H,De VA.Bayesian integration of sensor information and a multivariate dynamic linear model for prediction of dairy cow mastitis[J].Journal of Dairy Science,2016,99(9):7344-7361.
[12]金建刚,陈虎,刘丽辉,等.Bayes判别分析全血细胞减少的分类诊断[J].中国组织工程研究与临床康复,2009,13(6):1139-1142.