应用Poisson模型分析2001-2011年北京市丰台区居民死亡特征及预测  

The analysis and prediction of death among residents in Fengtai district of Beijing in 2001-2011

作　　者：信振江[1] 谢俊卿[1] 白俊梅[1] 杜忠仁[1] 李洁[1] 

机构地区：[1]北京市丰台区疾病预防控制中心,北京100071

出　　处：《中国预防医学杂志》2017年第8期583-586,共4页Chinese Preventive Medicine

基　　金：北京市科技计划课题(Z111107056811042);丰台区卫生系统科学研究项目(33081)

摘　　要：目的了解北京市丰台区2001-2011年居民死亡率与死因构成的历年水平与变化趋势,为更好地开展疾病预防工作提供科学依据与参考。方法对2001-2011年北京市丰台区居民死因数据进行统计分析,计算历年死亡率、死因构成和死因顺位,利用Poisson回归分析对死亡趋势进行预测。结果北京市丰台区2001-2011年居民死亡率为553.77/10万(2002年)~677.92/10万(2010年),男性居民死亡率为685.89/10万,女性居民死亡率为548.06/10万,男性死亡率高于女性。不同性别、年龄别死因顺位不同。Poisson模型对丰台区2011年死亡率的预测平均相对误差在2.97%,预测精度较高。结论北京市丰台区2001-2011年居民死亡率呈下降趋势,男性死亡率高于女性。慢性非传染性疾病成为影响丰台区居民健康的主要原因,需进一步加强慢性病的预防与控制,开展两癌筛查和脑卒中的早诊早治、加强健康教育和健康促进、倡导健康生活方式、倡导健康饮食、控制慢性病危险因素。Objective To understand the dynamic change of mortality and the cause of death among residents in Fengtai district of Beijing from 2001 to 2011,and to provide the scientific evidence for disease prevention and control. Methods Data of death of local residents recorded from 2001 to 2011 were statistically analyzed using SAS 9.2 software.Annual mortality and cause of death were calculated and sorted.Poisson regression model was used for death prediction. Results The overall mortality of residents in Fengtai District ranged from 553.77/100 000（2002）to 677.92/100 000（2010）,the death rate of male residents was higher than that of females（685.89/100 000 vs 548.06/100 000）.The cause of death was age and gender dependent.The average relative error of death predicted using Poisson model was only 2.97% with high precision. ConclusionsThe mortality of residents in Fengtai District shows a descending trend from the year 2001 to 2011,which is higher among male resident.The chronic non-communicable diseases become major influential factors for the health of local residents.Therefore,health education,cancer screening,healthy life style and diet need to be strengthened and promoted in order to control chronic disease risk factors.

关 键 词：死亡率 死因构成 死亡特征 Poisson回归模型 死亡预测 

分 类 号：R195[医药卫生—卫生统计学]