Publication Bias

As we have seen in Chap. 1 , systematic reviews and meta-analyses of randomised controlled trials generate the highest level of evidence. However, pooled estimates generated from systematic reviews sand meta-analyses can be biased due to methodological weakness of the individual trials such as inadequate randomisation, lack of allocation concealment, or incomplete outcome data. Furthermore, meta-analytical pooled estimates can be severely affected if not all existing evidence is included in the analysis. We call this publication bias and is the subject of this Chapter.