> For the complete documentation index, see [llms.txt](https://docs.nested-knowledge.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nested-knowledge.com/best-practices/best-practices-by-review-type/meta-analyses.md).
# Meta-Analyses
Learn about the best practices for conducting meta-analyses.
Here the best practices in conducting a meta-analysis are covered.
For step-by-step instructions in **Nested Knowledge**, see the [Meta-Analysis](/guides/guidance-by-review-type/meta-analysis.md) guide. The guidance below applies regardless of the software you use.
### Confirm The Studies Are Eligible To Pool
Before you commit to a meta-analysis, check that your included studies share enough in common to combine meaningfully. Studies should generally share a similar design and patient populations should be reasonably comparable across studies (age, disease severity, comorbidities). The interventions being compared need similar dosing, duration, and administration, and the outcome itself needs to be measured the same way, at a comparable timepoint, across every study.
### Weigh Whether You Have Enough Studies To Pool
There's no universally agreed minimum number of studies for a meta-analysis. Too few limits your ability to assess heterogeneity or detect a genuine effect. Let your research question and the available evidence, not an arbitrary target count, determine when you have enough.
### Extract Exact Values, Not Approximations
Pull the specific means, standard deviations, event counts, or effect estimates and confidence intervals a study reports. Reading a value off a plot or accepting a rounded summary introduces error that compounds once it's pooled with other studies' data.
### Test And Report Heterogeneity, And Let It Guide Your Model Choice
Statistically assess between-study heterogeneity (for example, with I² or Cochran's Q) and choose a fixed-effect or random-effects model based on what that test shows, rather than defaulting to one model regardless of the data. Report the heterogeneity statistics alongside your pooled estimate, not just the estimate itself.
### Assess Study Quality As Part Of Feasibility, Not Just Reporting
Appraise included studies before finalizing your decision to pool, and consider a sensitivity analysis that reports the pooled estimate with and without lower-quality studies included.
### Check For Publication Bias
Where you have enough included studies, assess whether smaller or null-result studies are underrepresented (for example, with a funnel plot). A pooled estimate built on a literature base skewed toward positive results overstates the true effect.
### Report Indirect Comparisons As Indirect
If your comparators weren'ts studied head-to-head and you're using network meta-analysis to compare them anyway, say so explicitly and report the method's own assumptions (for example, transitivity) rather than presenting the result as equivalent to a direct comparison.
### If You're Pooling Outside Nested Knowledge
Several general-purpose statistical tools can run a meta-analysis's underlying calculations directly, which is useful when your analysis needs a specific model or sensitivity test a platform doesn't automate.
| Tool | Notes |
| ---------- | --------------------------------------------------------------------------------------------------- |
| **Stata** | Its `metan` command runs a standard meta-analysis from the command line. |
| **R** | The `meta` package is a free, open-source option for running and customizing a meta-analysis. |
| **RevMan** | Built by the Cochrane Collaboration specifically for systematic review and meta-analysis workflows. |
| **Excel** | The MetaEasy add-in automates common meta-analysis calculations directly in a spreadsheet. |
### Next Steps
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.
Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:
```
GET https://docs.nested-knowledge.com/best-practices/best-practices-by-review-type/meta-analyses.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.