应用捕获-标记-再捕获法评估济南市户籍居民死亡登记漏报水平

摘要

目的 应用捕获-标记-再捕获法(capture-mark-recapture,CMR)评估2013-2014年济南市居民死亡登记漏报水平.方法 采用分层整群随机抽样方法,按照城乡分层抽取济南市2个区、3个县,从抽中的每个县(区)随机抽取3个乡镇(街道),每个乡镇(街道)随机抽取8个村(居委会).通过民政、公安、妇幼、村医、村干部、社区居委会搜集辖区内2013年1月1日至2014年12月31日的户籍人口死亡名单,与死因监测数据进行比对,计算漏报率.结果 本次调查共收集济南市2013-2014年死亡病例2 903例,死因监测登记同期死亡病例3 113例,CMR估计死亡病例3 772例(95％CI:3 741 ～3 802),死因监测漏报率为17.46％.不同类别的漏报率比较,2013、2014年漏报率分别为19.29％、15.57％(x2=8.92,P＜0.01).城区、农村漏报率分别为20.91％、11.93％(x2=47.35,P＜0.01).男性、女性漏报率分别为17.87％、16.98％(x2=0.48,P＞0.05),＜5岁组、≥5岁组漏报率分别为21.57％、17.45％(x2=0.03,P＞0.05).结论 济南市死因监测存在一定程度的漏报,应用CMR可以评估漏报水平,建议定期开展漏报调查,校正常规监测数据,并针对发现的漏报重点环节进行改进.%Objective To evaluate the under-registration rate related to mortality surveillance program in Jinan city,from 2013 to 2014,using the capture-mark-recapture method.Methods Under the stratified cluster random sampling method,samples were stratified,according to urban and rural areas.Two districts and three counties were chosen,with three towns/streets in each county/district and eight villages/communities in each township/street,randomly selected.With the participation of departments as civil affairs,public security,maternal/child institutions and community committees as well as individuals as village doctors and community leaders,a list of deaths from January 1,2013 to December 31,2014 was collected and compared to the National Mortality Surveillance System on the rates related to under-registration.Results A total of 2 903 records on deaths were collected from 2013 to 2014,while 3 113 deaths were reported in the surveillance system of the same period.3 772 (95％ CI:3 741-3 802) deaths were estimated by the capture-mark-recapture method.The two-year total under-registration rate was 17.46％.The under-registration rates in 2013 and 2014 appeared as 19.29％ and 15.57％ (x2 =8.92,P＜0.01),respectively,with the rates in urban and rural areas as 20.91％ and 11.93％ (x2 =47.35,P＜0.01).The rates on male and female were 17.87％ and 16.98％ (x2 =0.48,P＞0.05).The rates of ＜5 years old group and ≥5 years old group were 21.57％ and 17.45％ (x2 =0.03,P＞0.05) respectively.Conclusions Under-registration in the mortality surveillance was noticed in Jinan.The capture-mark-recapture method seemed useful for evaluating the under-registration rate in the mortality surveillance program.It is suggested that the survey on under-registration rate should be carried out on the regular base.The under-registration rate seemed useful in adjusting the mortality rate.