# How to Report a Meta-Analysis ## Quick Answer - Report a meta-analysis by following the PRISMA guideline structure, covering search strategy, study selection, risk of bias, effect estimates, and heterogeneity. - Select the correct reporting guideline from the [EQUATOR Network](https://www.equator-network.org/) before writing, since journals increasingly require guideline adherence. - A meta-analysis report is incomplete without effect sizes, confidence intervals, and a forest plot, and it must disclose data-sharing plans under policies like the [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy). ## At a Glance | Report Section | Required Elements | Common Omission | | --- | --- | --- | | Methods: Search Strategy | Databases, dates, keywords, filters, registration number | Reporting only one database or no date limits | | Methods: Study Selection | Inclusion and exclusion criteria, screening process, PRISMA flow diagram | Describing criteria vaguely without operational definitions | | Results: Effect Estimates | Pooled effect size, confidence interval, weights, heterogeneity statistics | Reporting only p-values without effect magnitudes | | Results: Forest Plot | Point estimates, confidence intervals, study weights, pooled diamond | Omitting individual study confidence intervals | | Discussion: Limitations | Risk of bias, publication bias, heterogeneity sources, generalizability | Acknowledging limitations without explaining their impact | ## Why Reporting Standards Matter for Meta-Analyses A meta-analysis combines results from multiple studies to produce a single estimate of effect. The value of that estimate depends entirely on how completely and transparently the methods are reported. Readers cannot judge whether the pooled result is trustworthy if they cannot see how studies were found, selected, and evaluated. The [EQUATOR Network](https://www.equator-network.org/) maintains a comprehensive library of reporting guidelines for health research. For systematic reviews and meta-analyses, the PRISMA statement is the primary guideline. Journals increasingly require authors to submit a completed PRISMA checklist alongside their manuscript. Editors and peer reviewers use the checklist to verify that critical methodological details are present. Reporting standards serve several practical functions. They allow readers to assess the validity of the methods. They enable replication by other research teams. They help clinicians and policymakers determine whether the findings apply to their specific context. They also protect against selective reporting, where authors present only favorable results. The [Committee on Publication Ethics](https://publicationethics.org/core-practices) core practices emphasize the importance of transparent reporting as part of responsible research conduct. Authors have an obligation to describe their work accurately and completely. Incomplete reporting can constitute a form of research misconduct when it misleads readers about what was actually done. ## Core Principles of Transparent Meta-Analysis Reporting ### Completeness Over Brevity Meta-analysis reports fail most often because authors omit methodological details in the interest of brevity. Journal space constraints are real, but supplementary materials and online appendices provide space for full descriptions. A report that cannot be replicated from its methods section is not complete. The [National Library of Medicine](https://www.ncbi.nlm.nih.gov/books) hosts authoritative biomedical books and research-method references that describe the rationale for detailed reporting. The core principle is that every decision made during the review process should be documented, including decisions that seem minor at the time. ### Reproducibility as the Standard A meta-analysis report should contain enough detail for another research team to repeat the search and selection process. This means reporting the exact search strings used for each database, the date each search was run, and the number of records retrieved. It also means describing how screening decisions were made and how disagreements between reviewers were resolved. Reproducibility extends to the statistical analysis. Authors should report the software and version used, the specific commands or functions employed, and the parameters chosen for the analysis. This level of detail allows readers to verify the calculations and to re-run the analysis with alternative assumptions if they wish. ### Distinguishing Protocol from Report A meta-analysis protocol describes what the authors plan to do before they begin. The report describes what was actually done. When these differ, the report must explain why. For example, if the authors planned to search five databases but could only access four, the report should state this deviation and its potential impact. Prospective registration of systematic review protocols is increasingly expected. Registration provides a public record of the planned methods, which helps prevent selective outcome reporting. The [NIH Grants and Funding](https://grants.nih.gov/) pages describe how funding agencies evaluate the rigor of proposed research methods, including the expectation that systematic reviews follow established standards. ## The PRISMA Framework for Reporting ### What PRISMA Requires The PRISMA statement provides a checklist of items that should be reported in a systematic review or meta-analysis. The checklist is organized into sections that correspond to the structure of a typical manuscript: title, abstract, introduction, methods, results, discussion, and funding. The methods section of a PRISMA-compliant report must describe the eligibility criteria for studies, all information sources searched, the search strategy in full, the process for selecting studies, the data collection process, and the risk of bias assessment. The results section must report the number of studies screened, assessed for eligibility, and included, with reasons for exclusion at each stage. The [EQUATOR Network](https://www.equator-network.org/) provides access to the full PRISMA checklist and related resources. Authors should download the current version of the checklist and use it as a writing guide, not as an afterthought. ### Adapting PRISMA to Your Research Question PRISMA is a general guideline, not a rigid template. The specific items that matter most depend on the research question and the types of studies being synthesized. A meta-analysis of randomized controlled trials will emphasize different reporting elements than a meta-analysis of observational studies. The key is to use PRISMA as a minimum standard and to add discipline-specific details where appropriate. For example, a meta-analysis in animal science might need to report housing conditions, breed, and husbandry practices that could affect the outcomes. These details are not explicitly listed in PRISMA but are essential for interpreting the results. ### Reporting Guidelines Beyond PRISMA The EQUATOR Network hosts a comprehensive, searchable registry of reporting guidelines covering observational studies, clinical trials, systematic reviews, qualitative research, and statistical analysis. Authors should check the target journal's instructions to authors early in the writing process. Some journals specify which guideline they expect authors to follow. Others leave the choice to the authors but require that a completed checklist be submitted with the manuscript. ## Preparing the Methods Section ### Search Strategy Reporting The search strategy is the foundation of a meta-analysis. Without a complete and reproducible search, the results cannot be trusted. The methods section must describe every database searched, the platform used for each database, the exact search strings, and the dates of the searches. A common failure is reporting only the number of records retrieved without providing the search strings. This makes it impossible for readers to assess whether the search was comprehensive or to update the search in the future. Full search strings should be included in the main text or in a supplementary appendix. The search strategy should also describe any limits applied, such as language restrictions, publication date limits, or study design filters. Each limit can introduce bias, so the rationale for each limit should be stated. ### Study Selection and Screening The methods section must describe how records were screened for eligibility. This includes the number of reviewers involved, whether screening was done independently, and how disagreements were resolved. The report should also describe whether full texts were retrieved for all potentially eligible records and how many could not be obtained. The eligibility criteria themselves must be reported in detail. These criteria should be specific enough that another research team could apply them to the same set of records and obtain the same set of included studies. Vague criteria such as "studies on the topic" are not acceptable. ### Data Extraction and Risk of Bias The methods section must describe what data were extracted from each included study and how the extraction was performed. This includes whether extraction was done in duplicate, how discrepancies were resolved, and whether study authors were contacted for missing data. Risk of bias assessment is a critical component of any meta-analysis. The methods must describe which tool was used, who performed the assessment, and how the results of the assessment were incorporated into the analysis. Some meta-analyses exclude studies at high risk of bias, while others include them and perform sensitivity analyses. ## Reporting the Results ### Effect Sizes and Confidence Intervals The results of a meta-analysis should report the pooled effect size with its confidence interval. The effect size metric should be defined clearly, whether it is a mean difference, standardized mean difference, odds ratio, risk ratio, or another measure. The confidence interval should be reported alongside the point estimate, beyond in a figure. Individual study results should also be reported, either in a table or in the forest plot. Readers need to see the individual study estimates to understand the contribution of each study to the pooled result. Reporting only the pooled estimate hides important information about the consistency of findings across studies. ### Heterogeneity Assessment Heterogeneity refers to the variability in effect estimates across studies. The report should describe how heterogeneity was assessed and what the results were. Common measures include the I-squared statistic and the chi-squared test for heterogeneity. The report should also describe how heterogeneity was handled in the analysis. Some meta-analyses use a random-effects model when heterogeneity is present, while others use a fixed-effect model. The choice of model should be justified based on the degree of heterogeneity and the assumptions of each model. ### Forest Plots and Other Visualizations A forest plot is the standard visualization for meta-analysis results. It shows the effect estimate and confidence interval for each individual study, along with the pooled estimate at the bottom. The plot should be clearly labeled with study identifiers, effect measures, and weights. The [National Library of Medicine](https://www.ncbi.nlm.nih.gov/books) resources describe best practices for presenting statistical results in biomedical research. Forest plots should be readable in grayscale as well as color, since many readers will print the article or view it on a monochrome device. ### Subgroup and Sensitivity Analyses Many meta-analyses include subgroup analyses to explore sources of heterogeneity. The report should describe which subgroups were planned in advance and which were explored post hoc. Post hoc analyses should be clearly labeled as exploratory, and their results should be interpreted with caution. Sensitivity analyses test whether the results are robust to methodological decisions. Common sensitivity analyses include excluding studies at high risk of bias, using a different effect measure, or changing the statistical model. The report should describe which sensitivity analyses were performed and whether the conclusions changed. ## The PRISMA Flow Diagram ### Constructing the Flow Diagram The PRISMA flow diagram shows the flow of information through the different phases of the review. It documents the number of records identified through database searching, the number of additional records identified through other sources, the number of records after duplicates are removed, and the number screened, excluded, and included. The flow diagram serves a critical function: it shows readers exactly how the final set of included studies was derived from the initial search results. This transparency allows readers to assess whether the selection process was reasonable and whether any steps may have introduced bias. ### Common Errors in Flow Diagrams A frequent error is reporting only the final number of included studies without showing the numbers at each stage. Another error is failing to report the number of records excluded at the full-text stage and the reasons for exclusion. The flow diagram should account for every record identified in the search. The flow diagram should also distinguish between records identified through database searching and records identified through other sources, such as reference lists, citation searching, or contact with experts. This distinction matters because the yield from different sources can vary substantially. ## Risk of Bias and Quality Assessment ### Selecting an Appropriate Tool Many tools are available for assessing risk of bias in individual studies. The choice of tool depends on the study designs included in the review. The methods section must identify the tool used and provide a rationale for the choice. The [EQUATOR Network](https://www.equator-network.org/) provides access to reporting guidelines that often include recommendations for risk of bias assessment. Authors should select a tool that is appropriate for the study designs in their review and that has been validated for that purpose. ### Reporting Risk of Bias Results The results of the risk of bias assessment should be reported for each included study. This is often done in a table or a figure that shows the risk of bias judgments for each domain assessed. The report should also describe how the risk of bias results influenced the analysis and interpretation. A meta-analysis that includes studies at high risk of bias should address this limitation directly. The authors should explain whether the high risk of bias affects the confidence in the pooled estimate and whether sensitivity analyses excluding these studies change the conclusions. ## Publication Bias and Small-Study Effects ### Assessing Publication Bias Publication bias occurs when studies with statistically significant results are more likely to be published than studies with null or negative results. This can lead to an overestimate of the true effect in a meta-analysis. The report should describe how publication bias was assessed. Common methods for assessing publication bias include funnel plots and statistical tests such as Egger's regression test. The report should present the results of these assessments and interpret them in the context of the number of studies included. ### Limitations of Publication Bias Assessment Publication bias assessment has important limitations, particularly when the number of included studies is small. Statistical tests for publication bias have low power when few studies are included. The report should acknowledge these limitations and avoid overinterpreting the results. Funnel plot asymmetry can also result from causes other than publication bias, such as true heterogeneity or poor study quality. The report should consider alternative explanations for asymmetry before concluding that publication bias is present. ## Data Sharing and Reproducibility ### Data Management and Sharing Policies Funding agencies increasingly require data management and sharing plans for research they support. The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) describes expectations for NIH-funded research, including the requirement to submit a data management and sharing plan with grant applications. For meta-analyses, data sharing typically includes the extracted data from included studies, the statistical code used for the analysis, and the search strategies. Sharing these materials allows other researchers to verify the results and to update the review in the future. ### Preparing a Data Sharing Plan A data sharing plan for a meta-analysis should describe what data will be shared, where it will be deposited, and when it will be available. The plan should also address any restrictions on sharing, such as concerns about identifying individual participants from the original studies. The [NIH Grants and Funding](https://grants.nih.gov/) pages provide guidance on preparing data sharing plans that meet funding agency expectations. Authors should review the specific requirements of their funding agency and ensure that their plan complies. ## Author Identity and Attribution ### Registering an ORCID An ORCID identifier provides a persistent digital identifier that distinguishes researchers from one another. The [ORCID for Researchers](https://info.orcid.org/researchers) pages describe how researchers can register for an ORCID and maintain their record. Many journals now require authors to provide their ORCID identifiers at submission. An ORCID record can link a researcher to their publications, grants, and peer review activities, which helps ensure proper attribution of research outputs. ### Managing Author Contributions The [Committee on Publication Ethics](https://publicationethics.org/core-practices) core practices address authorship and contributor ship. Authors should agree on the order of authorship and the contributions of each author before submission. The report should include a statement of author contributions that describes who did what. Disputes about authorship can delay publication and damage professional relationships. Clear communication about authorship expectations at the start of a project can prevent these problems. ## Common Failure Patterns in Meta-Analysis Reports ### Incomplete Search Reporting The most common failure in meta-analysis reports is incomplete reporting of the search strategy. Authors often report the databases searched but omit the exact search strings, the dates of the searches, or the number of records retrieved. This makes the search impossible to replicate or update. ### Selective Outcome Reporting Some meta-analyses report only the outcomes that showed statistically significant results. This practice biases the literature and misleads readers. The report should describe all outcomes that were assessed, including those with null results. ### Ignoring Heterogeneity Some meta-analyses report a pooled estimate without adequately addressing heterogeneity. When studies show widely varying results, a single pooled estimate may be misleading. The report should describe the heterogeneity and discuss its implications. ### Overstating Conclusions Meta-analyses sometimes overstate the strength of their conclusions. The discussion should acknowledge the limitations of the evidence and avoid causal language when the underlying studies are observational. The confidence in the pooled estimate should be calibrated to the quality and consistency of the included studies. ## Practical Steps for Writing the Report ### Step 1: Select the Reporting Guideline Before writing, identify the appropriate reporting guideline for your study type. The [EQUATOR Network](https://www.equator-network.org/) provides a searchable database of guidelines. Download the checklist and use it to structure your writing. ### Step 2: Review the Target Journal Requirements Check the target journal's instructions to authors for specific reporting requirements. Some journals require a completed checklist at submission. Others have specific formatting requirements for forest plots or flow diagrams. ### Step 3: Write the Methods Section First The methods section is the foundation of the report. Write it in enough detail that another research team could replicate your work. Include the search strategy, eligibility criteria, screening process, data extraction methods, and risk of bias assessment. ### Step 4: Report Results Transparently Present the results of the search and selection process in the flow diagram. Report the characteristics of included studies in a table. Present the pooled effect estimates with confidence intervals and the heterogeneity statistics. ### Step 5: Interpret With Appropriate Caution The discussion should interpret the results in the context of the limitations. Address the risk of bias in the included studies, the heterogeneity, and the potential for publication bias. Avoid overstating the conclusions. ### Step 6: Prepare Supplementary Materials Prepare the supplementary materials that support the report, including the full search strategies, the data extraction forms, and the statistical code. These materials should be submitted with the manuscript or made available in a public repository. ## Records and Measurements to Maintain | Record Type | Content | Purpose | | --- | --- | --- | | Search Log | Database, platform, date, search string, results count | Reproducibility and updating | | Screening Log | Records screened, excluded, and reasons | PRISMA flow diagram | | Data Extraction Form | Variables extracted from each study | Verification and data sharing | | Risk of Bias Table | Judgments for each domain and study | Transparency and sensitivity analysis | | Analysis Code | Software, version, commands, parameters | Reproducibility and verification | ## Professional Escalation Criteria ### When to Consult a Statistician A statistician should be consulted when the analysis involves complex methods that the authors do not fully understand. This includes network meta-analysis, individual patient data meta-analysis, or meta-regression with multiple covariates. A statistician can also help with the interpretation of heterogeneity and publication bias statistics. ### When to Consult a Librarian A librarian or information specialist should be consulted when designing the search strategy. Librarians have expertise in database-specific search syntax and can help ensure that the search is comprehensive. They can also assist with documenting the search for reporting purposes. ### When to Seek Ethical Guidance The [Committee on Publication Ethics](https://publicationethics.org/core-practices) provides guidance on ethical issues in research publication. Authors should seek guidance when they encounter questions about authorship, data integrity, duplicate publication, or conflicts of interest. ## Limitations of Meta-Analysis Reporting ### Reporting Cannot Fix Poor Methods A well-written report cannot compensate for poorly conducted research. If the search was incomplete, the eligibility criteria were vague, or the risk of bias assessment was inadequate, transparent reporting will expose these problems instead of solve them. ### Guidelines Are Minimum Standards Reporting guidelines represent minimum standards for transparency. They do not guarantee that the research was conducted well or that the conclusions are correct. Readers must still evaluate the quality of the methods and the strength of the evidence. ### Jurisdiction-Specific Requirements Reporting requirements can vary by jurisdiction and by funding agency. The [NIH Grants and Funding](https://grants.nih.gov/) pages describe requirements for NIH-funded research, but other funders may have different expectations. Authors should review the requirements of their specific funder and journal. ## Frequently Asked Questions ### What is the difference between a systematic review and a meta-analysis? Systematic reviews synthesise the literature qualitatively; a quantitative meta-analysis is added only when included studies share sufficient methodological homogeneity for statistical pooling. ### Do I need to register my meta-analysis protocol before starting? Prospective registration is strongly recommended and increasingly required by journals and funders. Registration provides a public record of your planned methods, which helps prevent selective reporting. Check the requirements of your target journal and funding agency. ### How do I choose between a fixed-effect and a random-effects model? A fixed-effect model assumes that all studies estimate the same underlying effect and that differences between studies are due to sampling error. A random-effects model assumes that the true effect varies across studies and incorporates this variability into the analysis. The choice should be based on the degree of heterogeneity and the assumptions of each model. ### What should I do if my search retrieves thousands of records? A large number of records is common for broad research questions. The screening process should be systematic and documented. Use the PRISMA flow diagram to show how the final set of included studies was derived from the initial search results. ### How do I handle studies with missing data? The methods section should describe how missing data were handled. Options include contacting study authors for missing information, using available case analysis, or performing sensitivity analyses to assess the impact of missing data. The approach should be described in the report. ### What is the role of the PRISMA checklist in manuscript submission? Many journals require authors to submit a completed PRISMA checklist with their manuscript. The checklist documents where each reporting item is addressed in the manuscript. This helps editors and reviewers verify that the report is complete. ### How do I report a meta-analysis with only a few included studies? A small number of included studies is a limitation that should be acknowledged. The report should describe the search and selection process in full, present the individual study results, and interpret the pooled estimate with appropriate caution. Statistical tests for publication bias have low power with few studies. ### Where should I deposit my data and analysis code? Data and code should be deposited in a public repository that provides a persistent identifier, such as a DOI. The [NIH Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy) describes expectations for NIH-funded research. Check the requirements of your funder and journal for specific repository recommendations. ## Related Articles - [Guidelines for Meta-Analysis: Reporting Standards and Best Practices](/blog/guides/guidelines-for-meta-analysis-reporting-standards-and-best-practices) - [Meta-Analysis of Individual Participant Data: Rationale, Conduct, and Reporting](/blog/guides/meta-analysis-of-individual-participant-data-rationale-conduct-and-reporting) - [Meta-Analysis vs Systematic Review: What's the Difference?](/blog/guides/meta-analysis-vs-systematic-review-what-s-the-difference) - [The Importance of Meta-Analysis in Research: A Primer](/blog/guides/the-importance-of-meta-analysis-in-research-a-primer) - [Common Mistakes in Meta-Analysis and How to Avoid Them](/blog/guides/common-mistakes-in-meta-analysis-and-how-to-avoid-them) ## References and Further Reading - [Research Methods Resources](https://www.ncbi.nlm.nih.gov/books). National Library of Medicine. - [EQUATOR Network](https://www.equator-network.org/). EQUATOR Network. - [Core Practices](https://publicationethics.org/core-practices). Committee on Publication Ethics. - [NIH Grants and Funding](https://grants.nih.gov/). National Institutes of Health. - [ORCID for Researchers](https://info.orcid.org/researchers). ORCID. - [Data Management and Sharing Policy](https://sharing.nih.gov/data-management-and-sharing-policy). National Institutes of Health. - [NCBI Literature Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information. - [PubMed](https://pubmed.ncbi.nlm.nih.gov/). National Library of Medicine. > This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.