Brookmeyer R, Day N E, Moss S
Stat Med. 1986 Mar-Apr;5(2):127-38. doi: 10.1002/sim.4780050204.
The duration of the detectable preclinical phase of disease, which we call the sojourn time, is an important quantity to consider when recommending frequency of screens and investigating the natural history of the disease. In this paper, we propose a matched case-control methodology for screening data for joint estimation of the sojourn time distribution and the false negative rate of the screening test. The approach is based on a conditional likelihood for matched sets that involves a comparison of the screening histories of cases and controls. We illustrate the methodology with the data from the Aberdeen screening programme for cervical cancer by the Pap smear. We consider a number of different parametric forms for the sojourn time distribution.
疾病可检测临床前期的持续时间，我们称之为停留时间，在推荐筛查频率和研究疾病自然史时是一个需要考虑的重要指标。在本文中，我们提出了一种匹配病例对照方法，用于筛查数据，以联合估计停留时间分布和筛查试验的假阴性率。该方法基于匹配集的条件似然性，涉及病例和对照筛查历史的比较。我们用来自阿伯丁宫颈癌巴氏涂片筛查项目的数据说明了该方法。我们考虑了停留时间分布的多种不同参数形式。
