多水平模型在肝癌死亡病例与关联因素生态学分析中应用

发布时间：2018-04-23 04:26

本文选题：肝癌 + 关联因素　；参考：《中国公共卫生》2017年07期

【摘要】：目的分析肝癌死亡病例的生态学关联因素,为卫生行政部门了解肝癌死亡病例的关联因素、合理配置卫生资源提供依据。方法利用全国疾病监测系统161个死因监测点的肝癌死亡监测资料、乙肝和丙肝报告发病数据、乙肝血清流行病学调查数据、全国行为危险因素调查数据和有关社会经济指标,采用三水平超Poisson方差模型,以肝癌死亡病例作为第一水平,161个区县作为第二水平,省(自治区、直辖市)作为第三水平,分析肝癌死亡病例的生态学关联因素;应用SAS 9.4和MLwiN 2.30软件,进行单因素和多因素分析,拟合三水平超Poisson方差随机截距模型。结果同省(自治区、直辖市)不同性别、年龄组间的肝癌死亡率不同,且年龄组间肝癌死亡率的差异在不同性别间也不同,性别与年龄之间存在交互作用;多水平模型在三、二水平的随机截距均有统计学意义,即肝癌死亡在不同省(自治区、直辖市)、不同区县水平上差异均有统计学意义;多因素分析结果显示,在岗职工平均工资、乙肝表面抗原阳性率、乙肝发病率、丙肝发病率、性别、年龄、性别与年龄交互项与肝癌死亡病例之间存在独立的生态学关联关系。结论在岗职工平均工资高与肝癌死亡病例呈负关联关系;乙肝表面抗原阳性率、乙肝和丙肝报告发病率与肝癌死亡病例呈正关联关系;男性、高年龄组与肝癌死亡病例呈正关联关系。
[Abstract]:Objective to analyze the ecological correlation factors of liver cancer death cases, and to provide the basis for the health administration department to understand the related factors of liver cancer death cases and to allocate health resources reasonably. Methods the data of mortality of liver cancer were collected from 161 cause of death surveillance sites in the National Disease Surveillance system, the data of hepatitis B and C hepatitis were reported, the data of seroepidemiological investigation of hepatitis B, the survey data of risk factors of national behavior and related social and economic indicators were used. The three level Poisson variance model was used to analyze the ecological correlation factors of liver cancer death cases, taking the death cases of liver cancer as the first level, 161 districts and counties as the second level and provinces (autonomous regions, municipalities directly under the Central Government) as the third level. Using SAS 9.4 and MLwiN 2.30 software, single factor and multivariate analysis were carried out to fit the three-level random intercept model of Poisson variance. Results the mortality rate of liver cancer in different age groups was different from that in different provinces (autonomous regions and municipalities directly under the Central Government), and the difference of mortality rate of liver cancer among age groups was also different between sexes, and there was interaction between sex and age. The random intercept of the two levels was statistically significant, that is, the difference of liver cancer deaths was statistically significant in different provinces (autonomous regions, municipalities directly under the Central Government, different districts and counties), and the results of multi-factor analysis showed that the average wages of employed workers and staff members were significantly different from those in different provinces (autonomous regions, municipalities directly under the Central Government). The positive rate of hepatitis B surface antigen, incidence of hepatitis B, incidence of hepatitis C, sex, age, sex and age interaction have independent ecological correlation with the death of liver cancer. Conclusion there is a negative correlation between the average wage and the death rate of liver cancer, the positive rate of HBsAg, the reported incidence of hepatitis B and hepatitis C and the death of liver cancer. There was a positive correlation between the age group and the death cases of liver cancer.
【作者单位】：中国疾病预防控制中心流行病学办公室;中国疾病预防控制中心研究生院;中国科学院地理科学与资源研究所;北京儿童医院;中国疾病预防控制中心慢性非传染性疾病预防控制中心;
【基金】：淮河流域癌症综合防治项目(1310800003)
【分类号】：R735.7

【相似文献】