Abstract
A meta-analysis a statistical method that integrates the quantitative findings from separate but similar studies and provides a numerical estimate of interest. A well conducted systematic review is very important before performing a successful meta-analysis. It is necessary to test heterogeneity between studies included in meta-analysis. Publication bias is a major problem in meta-analysis. So funnel plots should use to detect a possible publication bias. Sensitivity analysis is also an important statistical method to reduce the uncertainty. A forest plot will make easier to interpret the aggregate effect size obtained by combining all the studies included in meta-analysis. In conclusion, careful reviewing of published meta-analyses and a balanced assessment of their deficiencies is likely to become an increasingly important way of resolving uncertainty in healthcare practices.

Key words: systematic review, heterogeneity, publication bias