论文部分内容阅读
医学科学研究的结论,理论上应是某些客观规律,是经得起实践的检验,即具有可重复性的。但是我们经常听到一些医生诉述:某文章中的疗效分析或某病的临床规律无法重复获得。从医学统计学的角度分析,出现以上矛盾的原因有:(1)总体条件不同,如疾病诊断依据不同,观察疗效的方法不同,疗效标准不统一等。(2)样本具有偏性,如样本含量太少,抽样未遵循随机原则等。(3)分析对比的资料缺乏可比性,把某些假象当作客观规律。本文着重说明有关可比性问题。
The conclusion of medical scientific research should theoretically be some objective law, and it can stand the test of practice, that is, it is reproducible. But we often hear from some doctors that the efficacy analysis in an article or the clinical law of a disease cannot be repeated. From the perspective of medical statistics, the reasons for the above contradictions are: (1) The overall conditions are different, such as different diagnosis basis, different methods for observing curative effect, and inconsistent curative effect standards. (2) The sample is biased, if the sample content is too small, the sampling does not follow the random principle. (3) There is a lack of comparability in the analysis of comparative data, and some illusions are regarded as objective laws. This article focuses on the issue of comparability.