ABSTRACT
Selective publishing of study results is a common threat to the validity of meta-analyses and systematic reviews, leading to bias of pooled treatment effects and inaccurate conclusions. This chapter provides an overview of publication and outcome reporting bias and their negative consequences, as well as preventive measures imposed by governing bodies to reduce their frequency and impact. Statistical methods for detecting and correcting for bias are also presented, including graphical diagnostics, regression tests, trim-and-fill, and selection models, with a discussion of their advantages and limitations. A worked example is included to illustrate the use of these methods in practice.
