Goldberg J D, Wittes J T
Biometrics. 1978 Mar;34(1):77-86.
In a medical screening program for early detection of disease, one or more screening modes are administered to an apparently healthy population. Knowledge of the true disease status for all screened individuals would allow estimation of the false negative and false positive rates for each mode of detection and for the program as a whole. This paper develops capture-recapture methods applicable to programs as a whole. This paper develops capture-recapture methods applicable to programs when follow-up of individuals negative on screening is not performed or is incomplete. The methods require at least two independent modes of detection. Data from a breast cancer screening program illustrate the procedure. The results of four screening examinations at approximately one-year intervals and the long-term follow-up of all screened individuals support the usefulness of these methods in the evaluation of a screening program.
在一项用于疾病早期检测的医学筛查项目中，会对看似健康的人群采用一种或多种筛查方式。知晓所有被筛查个体的真实疾病状况，将有助于估算每种检测方式以及整个项目的假阴性率和假阳性率。本文开发了适用于整个项目的捕获-再捕获方法。本文还开发了在不对筛查呈阴性的个体进行随访或随访不完整的情况下适用于项目的捕获-再捕获方法。这些方法至少需要两种独立的检测方式。来自一项乳腺癌筛查项目的数据说明了该程序。大约每隔一年进行的四次筛查检查结果以及对所有被筛查个体的长期随访，支持了这些方法在评估筛查项目中的实用性。
