What is meta-analysis?
Meta-analysis is the statistical combination of results from two or more studies. For example, imagine you had three different studies for intervention X versus placebo. Meta-analysis allows you to effectively pool the data from the three studies, providing the researcher with a weighted, average effect estimate, from the three studies.
The outputs are then usually presented as forest plots. These will include separate plots for each study, showing the effect size for the given outcome, confidence intervals and size of the study. A summary result for the combined analysis is included at the bottom of the plot, presented as a diamond, which represents the point estimate of the averaged studies and the extreme horizontal points of the diamond represent the overall 95% confidence interval.
Why is this helpful? Well meta-analysis is a common element within systematic review and forms an important element of evidence-based healthcare and influence on decision making. The Cochrane collaboration highlight that meta-analysis, as part of systematic review can provide an improvement in precision in effect estimates, they can also provide answer to questions not addressed in individual studies and also help provide answers to conflicting results. Essentially this process of combining available evidence makes best use of the totality of evidence by providing a balanced analysis rather than presentation of selected studies in isolation.
From an evidence perspective I think there are a couple of things to be aware of:
- They can be a very powerful tool in supporting HTA decision making and evidence-based healthcare. In fact, when conducted robustly, a meta-analysis as part of systematic review sits at the top of the evidence pyramid.
- That whilst powerful as a piece of evidence in its own right, one has to be aware that there are areas inherent in the analysis that can produce misleading results, if not properly considered. These areas relate to the types of studies included, bias within the studies and in their reporting and heterogeneity across studies. These aspects need to be evaluated and treated in an appropriate way with the support of a well-defined study protocol. As with many things in health economics there are sources of best practice and guidance that helps researchers to conduct high quality synthesis and to also assess the quality of analyses.
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.4 (updated August 2023). Cochrane, 2023. Available from www.training.cochrane.org/handbook.