Research

Meta-analysis – A Methodology for Basic Research

A meta-analysis is a statistical approach that summarize data from different studies. Meta-analysis can be used to identify this common impact when the treatment effects (or effect sizes) are constant across studies. 

What is Meta-analysis?

A meta-analysis is a statistical approach that summarize data from different studies. Meta-analysis can be used to identify this common impact when the treatment effects (or effect sizes) are constant across studies. Meta-analysis can be performed to determine the cause of variance in the effect when it differs between studies.

Objectives of Meta-analysis

The following objectives are achieved by meta-analysis:

  • Establishing statistical significance in studies with conflicting results.
  • Create a more accurate estimation of effect size.
  • Provides more thorough study of the risks, safety information, and advantages.
  • In order to explore subgroups with non-significant individual numbers.

Where does meta-analysis fit into the scientific process?

1. Publications

This method replaces the classic narrative review. Many journals encourage authors to submit systematic reviews and meta-analyses that compile the available data on specific topics. In other articles, meta-analyses also support the main ideas. For example, an article describing the result of a new primary study may include a meta-analysis in the introduction to summarize the previous study and given an overview of the study in perspective.

2. Designing new studies

Meta-analyses are very useful in designing new studies. The meta-analysis results can be used to determine which issue have already been addressed and which require further research. It can also be used to determine which population or outcome measures are most likely to produce meaningful result and which variations of the intended intervention are likely to have the greatest impact.

3. Grant proposals

Grant proposal often use meta-analyses to support the need for additional research. A meta-analysis helps to contextualize the available data and indicates the potential value a planned investigation. Graphical components of meta-analysis, such as forest plot, provides a way to present data clearly and capture the attention of reviewers. Some funding organizations require studies meta-analysis of previous research as a part of funding application to fund future research.

How can I do a meta-analysis?

  1. Create a query (based on theory).
  2. Perform a search (using PubMed/Medline, Google Scholar, or another source).
  3. Check the title and abstract of each paper.
  4. Create a summary of the final selected articles.
  5. Evaluate the content of these articles to determine their creditability. In addition to the GRADE criteria, this is done by assessing their internal validity.
  6. How diverse are these articles, and how many are there?
  7. Create a forest plot to estimate the total effect size as an odds ratio while also using fixed and random effects models.
  8. Run a funnel plot to determine the level of publication bias for these articles.
  9. Perform subset analysis and meta-regression to determine if there is a subset of studies that captures the aggregate effects.

Advantages of meta-analysis

  • Increased statistical power
  • Confirmation-based data analysis
  • Improved extrapolation to the whole affected population
  • Being considered a reliable source

Limitations in meta-analysis

  • Finding relevant studies is difficult and time-consuming.
  • Not all studies provided sufficient data for analysis and inclusion.
  • A sophisticated statistical approach is required
  • Study population heterogeneity