论文部分内容阅读
目的 探讨广义Poisson回归模型在流行病学队列资料分析中的应用与价值。方法 拟合超额相对危险度模型 ,绝对超额危险度模型 ,非标准模型等分析云锡现场队列研究资料。结果 在云锡肺癌现场 ,每一个观察人年每一工作水平月的氡子体暴露 ,会产生 1 0 9× 10 -5个肺癌超额病例。在总的 336例病例中 ,归因于氡子体暴露的有 12 3例。只有当氡子体累积暴露高于 5 88 37WLM时 ,暴露人群患肺癌危险性才高于基线对照。结论 广义Poisson回归模型的拟合可以得到对于危险因素的更深入描述 ,如计算出每一单位危险因素变化时 ,可得超额相对危险度的变化、以及由危险因素所引起超额病例数、超额绝对危险 (EAR)、归因危险比 (AR)以及超额发病率等
Objective To explore the application and value of generalized Poisson regression model in epidemiological cohort data analysis. Methods The relative risk model, absolute excess risk model, and non-standard model were used to analyze the data of the Yunxi field cohort study. Results At Yunxi lung cancer site, each observing person’s exposure to the radon progeny for each working month of the year would produce an excess of 109 million lung cancer cases. Of the total 336 cases, 123 were attributed to the exposure of the coriander. Only when the cumulative exposure of scorpions was higher than 5,88,37 WLM, the risk of lung cancer among the exposed population was higher than the baseline control. Conclusion The generalized Poisson regression model can be used to obtain a more in-depth description of the risk factors. For example, when calculating the change of each unit risk factor, the excess relative risk, the number of excess cases caused by the risk factors, and the excess absolute Risk (EAR), attributive hazard ratio (AR), excess morbidity, etc.