8.8.1  Introduction

Statistical considerations often involve a trade-off between bias and precision. A meta-analysis that includes all eligible studies may produce a result with high precision (narrow confidence interval), but be seriously biased because of flaws in the conduct of some of the studies. On the other hand, including only the studies at low risk of bias in all domains assessed may produce a result that is unbiased but imprecise (if there are only a few high-quality studies).

 

When performing and presenting meta-analyses, review authors must address risk of bias in the results of included studies. It is not appropriate to present analyses and interpretations based on all studies, ignoring flaws identified during the assessment of risk of bias. The higher the proportion of studies assessed to be at high risk of bias, the more cautious should be the analysis and interpretation of their results, and the lower will be the grading of the quality of the evidence.