Letter from the editor: Meta-analysis, forest plots, and assessment of heterogeneity

To the Editor: Prior letters from the editor addressed the problem of well-performed meta-analyses that produce conflicting results, steps to improve systematic reviews, and the common models and assumptions used in meta-analysis. This fourth letter in the series discusses forest plots, how they can help us understand conflicting data, and situations in which they can augment or substitute for meta-analysis, as well as methods to assess the heterogeneity between studies.