论文部分内容阅读
回顾性调查是病因学研究常用的方法。但由于其本身缺点的限制,一次调查的结果很难得出某一因素和疾病关系肯定或否定的结论。如若进行多次调查,而每次调查的结果该因素的危险性又颇不一致,那么,怎样运用一个统一的指标来反映这数次调查的结果呢?本文介绍两种常用的危险性合并方法。以Doren对吸烟和肺癌关系的回顾性调查为例。可供分析的资料共计10次(见附表)。其危险性合并的方法如下。一、估计相对危险性(0R)的合并设病例组具有危险因素(吸烟)的人数为a,不具有该因素的人数为c;对照组具有危险因素的人数为b,不具有该因素的人数为d。
Retrospective surveys are commonly used methods in etiological studies. However, because of its own shortcomings, it is difficult to draw a conclusion that the relationship between a certain factor and disease is positive or negative because of the results of one investigation. If multiple surveys are conducted and the risk of the factors is quite inconsistent in each survey, then how do we use a unified indicator to reflect the results of the surveys? This article describes two commonly used risk merger methods. Take Doren’s retrospective survey of the relationship between smoking and lung cancer as an example. The total number of data available for analysis is 10 (see attached table). The risk of merger is as follows. I. The number of risk factors (smoking) in the combined risk group (OR) is estimated to be a, and the number of people who do not have this factor is c; the number of risk factors in the control group is b, the number of people who do not have this factor. d.