摘要
Meta分析本质上是一种观察性研究,偏倚对Meta分析结果的真实性和可靠性有很大影响。该文详细介绍了Meta分析中抽样偏倚、选择偏倚和研究内偏倚等三类偏倚的产生原因及减少各种偏倚的常用手段与方法,重点讨论了发表偏倚的识别和控制。
Meta-analysis should be viewed as an observational study of the evidence. Bias often causes conflicting results of the meta-analysis and threatens its internal validity and reliability. There are at least three types of bias involved in meta-analysis: sampling bias, selection bias and within study bias. In this article, the causes and consequences of bias would be introduced briefly, and the methods of detecting and correcting for publication bias would be discussed.
出处
《循证医学》
CSCD
2002年第4期216-220,共5页
The Journal of Evidence-Based Medicine
