应用捕获-再捕获方法估计林州市全死因死亡率和漏报率

ESTIMATION OF THE ALL-CAUSE MORTALITY AND UNDER-REPORTING RATE BY CAPTURE-RECAPTURE METHOD IN LINZHOU CITY

[目的]应用捕获-再捕获方法估计林州市2004～2005年全死因死亡率,评价全死因登记处收集死亡资料的完整性。[方法]收集林州市死因登记处、民政和公安3部门2004～2005年的全死因死亡登记资料。不同来源数据间的匹配使用姓名、年龄、性别、死亡时间、死亡原因和居住地址等变量。使用对数线性泊松模型估计3样本中都没有出现的死亡人数。[结果]2004～2005年期间3来源资料中共报告全死因死亡人数为13007人,死亡率为6.42‰。其中,登记处报告的全死因死亡人数为12261人,死亡率为6.06‰。用捕获-再捕获方法估计的全死因死亡人数为13402人,死亡率为6.62‰。登记处的漏报率为8.5%,合并样本的漏报率为2.3%。[结论]仅仅使用登记处或合并样本资料会低估全死因死亡率,而利用公安、村访谈、登记处等部门现有的资料,使用捕获-再捕获方法可以校正低估的全死因死亡率。

[Objective] To estimate the all-cause mortality in Linzhou City from 2004 to 2005 using the capture-recapture method, and to evaluate the completeness of data in all death causes registry. [ Methods] The data of all-cause death were collected from the registry office, civil administration and police registries. Cases in every source were matched on the name, sex, age, date of death, cause of death and resident address. A 3-source log-liner Poisson model was used to estimate the number of all death cause which did not appear in three sources. [Results] It was reported that death number of all-cause was 13 007 and the mortality was 6.42 per 1 000 from combined three sources. Among that the number and mortality were respectively 12 261 and 6.06 per 1 000 from the registry. The capture-recapture method estimated that the number of that was 13 402 and the mortality was 6.62 per 1 000. Under-reporting rates of registry and aggregate database were 8.5% and 2.9%. [Conclusion] Registry office or aggregated databases underestimate the all-cause mortality. On the basis of data which are collected from police, civil and registry, using the capture-recapture method can adjust the all-cause mortality.