肿瘤发病人数估计的分层捕获-再捕获方法

Stratified ‘capture-recaptured’ methodology in the estimation on the number of cancer incidence cases

目的 建立一种估计广州市越秀区恶性肿瘤发病人数方法。方法 根据死亡统计和医院病案两种来源收集肿瘤病例 ,建立分层捕获 再捕获模型 ,应用Bayesian方法估计模型的参数。分析程序由SAS编程语言编写。结果 越秀区男性和女性恶性肿瘤人数的估计值分别为 6 10例和 5 2 0例 ,现有调查方式获得的男性和女性肿瘤发病资料的漏报率的估计值分别为 8.5 %和 5 .4%。结论模拟研究的结果说明分层捕获

Objective To develop a new method in estimating the number of cancer incidence in Yuexiu district in Guangzhou. Methods Data of cancer incidence was collected thorugh hospital records and death certificates to develop a stratified capture recaptured model. Related procedures for the parameter estimation were given and implemented on SAS language. Results Estimated numbers of both male and female cancer incidence in Yuexiu district were 610 and 520 respectively. The estimated miss reporting rates on both male and female cancer incidence from hospital records and death certificates were 8.5 % and 5.4 % respectively. Conclusion The result of simulation studies showed that stratified ‘capture recaptured’ model proposed in this paper was reliable for the estimation of the number of cancer incidence cases.