This is an archived version of the Handbook. For the current version, please go to training.cochrane.org/handbook/current.

When review authors find evidence of small-study effects, they should consider sensitivity analyses examining how the results of the meta-analysis change under different assumptions relating to the reasons for these effects. We stress the exploratory nature of such analysis, due to the inherent difficulty in adjusting for publication bias and lack of research into the performance of such methods applied conditionally based on the results of tests for publication bias considered in Section 10.4.3. This area is relatively underdeveloped; the following approaches have been suggested.