应用捕获-再捕获方法评价2009-2011年湖北省死因监测数据完整性

Evaluating the completeness of death-cause data from diseases surveillance points of Hubei province during 2009-2011 by capture-mark-recapture method

目的应用捕获-再捕获方法(capture-mark-recapture method,CMR)评价2009-2011年湖北省疾病监测点(diseases surveillance points,DSPs)死因监测数据的完整性。方法在湖北省国家DSPs的基础上,将DSPs由6个扩大至11个,抽取能代表各DSPs人口水平、经济水平、死亡水平的乡镇(街道)1个,其所有的家庭均为调查家庭,开展漏报调查。收集2009年1月1日至2011年12月31日湖北省DSPs抽样地区常规死因网络报告数据和死因漏报调查数据,人口数据以户籍人口为准。采用CMR估计目标人群的死亡人数,计算死亡率。结果 2009-2011年,湖北省DSPs抽样地区常规死因网络报告和死因漏报调查2个独立来源资料显示,全死因死亡人数分别为9 019和9 481人,死亡率分别为6.16‰和6.48‰。用CMR估计的全死因死亡人数为10 200人,死亡率为6.97‰。结论湖北省DSPs抽样地区死因监测数据完整性较好。但是,死因网络报告质量在不同的DSPs之间分布不均衡,个别DSPs存在一定的漏报现象。建议定期开展漏报调查,使用CMR校正低估的死亡率。

Objective To evaluate the completeness of death-cause data from diseases surveillance points （DSPs） of Hubei province during 2009-2011 by capture-recapture method. Methods On the basis of 6 national DSPs, 11 DSPs in Hubei province were expanded. A town or residential district was selected, which could represent the demographical, economical and death level of 11 DSPs. All families of the town served as the subjects for missing report investigation. The data of the routine death-cause reports and the missing death-cause reports from DSPs in Hubei province were collected. Capture-mark-recapture （CMR） method was used to estimate the number of death among the sampled population, and the mortality was calculated. Results The data from the routine death reporting system and under-reporting investigation system in sampling areas during 2009-2011 indicated that the death numbers were 9 019 and 9 481 people, respectively, the mortalities were 6.16 %0 and 6.48 %o, respectively. The death number and mortality estimated by CMR were 10 200 people and 6.97 %v. Conclusion The data from DSPs sampling areas in Hubei province are complete. But the quality of the routine death reporting system at different DSPs was not equal. There are the missing-reports in some DSPs to some extent. It is suggested that the missing-report investigation should be performed and CMR could be used to adjust the underestimated mortality.