No, most evidence syntheses using published, non-identifiable data don’t need ethics approval; approval may apply when identifiable data are handled.
Researchers ask this question at the planning stage because journal editors, funders, and institutions look for clean compliance. The short path is to check whether your project involves living people, identifiable records, or any direct contact. If the answer is no and you are only synthesizing published results, formal review by an Institutional Review Board (IRB) or Research Ethics Committee (REC) is usually not required. Local rules and special cases still exist, so this guide walks through the common scenarios, what statements to include, and where to verify the edge cases with official sources.
What Counts As “Ethical Approval” In This Context
The term covers oversight by bodies such as IRBs in the United States and RECs in the United Kingdom. Their mandate is to protect participants when a study gathers new data, accesses identifiable records, or interacts with people. A synthesis that summarizes already published, de-identified findings does not create new risk for participants, which is the core reason many projects in this genre fall outside mandatory review.
Quick Determination Table
This table gives you a fast read on the common setups in review projects and whether formal ethics review is likely.
| Scenario | Ethics Review Likely? | Notes |
|---|---|---|
| Screen and synthesize published, aggregate results only | No | Standard case for evidence syntheses built on public reports. |
| Use public, de-identified datasets with no re-linking | Usually no | Common Rule/OHRP guidance treats this as not human-subjects research. |
| Handle individual participant data (IPD) under a data-use agreement | Often yes | Oversight or a formal exemption is commonly needed; security measures required. |
| Contact study authors or participants for new or missing data | Often yes | Direct interaction can trigger review requirements. |
| Access identifiable health records from a provider or registry | Yes | Expect IRB/REC review and privacy approvals. |
| Quality-assurance literature scan internal to a service team | Varies | May fall under local governance instead of research ethics pathways. |
Ethical Approval For Evidence Syntheses: When It’s Needed
Most projects synthesize published reports and stay in the “no review required” lane. There are clear triggers that move a project into oversight territory.
Trigger 1: Identifiable Data
If the dataset contains direct identifiers or combinations that could re-identify a person, you move into human-subjects territory. In the United States, Office for Human Research Protections (OHRP) materials explain that analysis of de-identified, public data is not human-subjects research, while use of identifiable or re-linkable data can require review or an exemption letter. See the OHRP guidance on coded private information for definitions and boundaries (link: coded private information guidance).
Trigger 2: Direct Interaction
Contacting authors to clarify methods is common and usually benign. Contacting participants or collecting any new data crosses into research conduct, which can require IRB/REC review. If in doubt, request a determination from your local office and file the letter with your project records.
Trigger 3: Records Under A Health Provider Or National System
In the UK, the Health Research Authority provides a decision tool for NHS REC requirements. Routine syntheses of published studies do not need NHS REC review, while work that uses NHS patient data or facilities can. The HRA’s “What approvals and decisions do I need?” page explains these routes in plain terms (link: HRA approvals guidance). You can also use their ethics decision tool to check your case (link: NHS REC decision tool).
When Ethics Review Is Not Typically Required
Large segments of review work involve only published, aggregate data. University IRB manuals echo that review is not required for analysis of public, de-identified datasets and for literature syntheses that do not add new data collection. Examples include several U.S. university guides that state analysis of public, de-identified data does not constitute human-subjects research under 45 CFR 46.
Why Journals Still Ask For An Ethics Statement
Editors need a clear line in your manuscript that either cites the approval or explains why it was not required. Outlets aligned with the International Committee of Medical Journal Editors (ICMJE) expect a statement either way. The BMJ, for instance, requires every research article to include a sentence confirming approval or explaining the absence of a requirement. Include the name of the committee if approval exists, or a short rationale when it does not.
Fit Your Methods To Reporting Standards
Ethics oversight is one axis; transparent reporting is another. For health interventions, the Cochrane Handbook lays out accepted methods for planning, screening, bias assessment, and synthesis. For reporting, the PRISMA 2020 statement is the widely endorsed checklist. These resources do not replace ethics review; they improve clarity and reproducibility. See the Cochrane Handbook overview and the PRISMA site for the current materials.
What To Say In Your Manuscript
Place a short “Ethics” line in your Methods or Declarations. Keep it factual and specific to your project. Here are ready-to-use models you can adapt to your journal’s template:
- No approval required: “This evidence synthesis used only published, aggregate data and did not involve human subjects as defined by 45 CFR 46; institutional review was not required.”
- Exemption/determination: “The project received a Not Human Subjects Research determination from the University IRB (File #2025-123).”
- Approval required: “The project was approved by the REC/IRB at [Institution] (Ref #12345). Data were handled under a data-use agreement with access controls.”
Borderline Cases And How To Handle Them
Real projects sometimes land in a gray zone. These patterns come up often:
Individual Participant Data (IPD) Meta-analysis
When sponsors share participant-level datasets under agreement, you step into data-protection territory. Expect to document lawful basis for processing, storage controls, and, in many jurisdictions, IRB/REC oversight or a formal exemption. Lock down who can access the re-identification key, if any.
Registry Or Claims Data
Even without contact, access to identifiable or linkable records typically requires approval. Plan for privacy review, data-use agreements, and security audits.
Quality Improvement Versus Research
Some teams compile literature for internal service improvement. In systems such as the NHS, this may fall under local clinical governance rather than research ethics. Always confirm with the local office listed by your institution or provider.
Do Evidence Syntheses Require IRB Or REC Approval?
Here is an at-a-glance grid you can reference when a co-author or editor asks for the rule. It pairs common review designs with the usual oversight path. Use it during protocol drafting and when crafting your ethics statement for the manuscript.
| Review Type | Typical Oversight Path | Practical Tip |
|---|---|---|
| Intervention review using published aggregate data | No IRB/REC; add a “not required” statement | Cite PRISMA; add search/selection details for transparency. |
| IPD meta-analysis with data-use agreements | IRB/REC or exemption letter | Document privacy controls and access limits to the key. |
| Scoping review of published ethics or policy sources | Usually no oversight | Still include a clear ethics statement in the paper. |
| Review using identifiable registry or EHR extracts | IRB/REC plus privacy approvals | Confirm HIPAA/legal basis, storage, and audit trails. |
| Methodology review with author surveys | IRB/REC required | Surveys count as interaction with living individuals. |
Align With Editorial Policies
Many journals follow ICMJE recommendations. These call for ethical oversight where human participants or animals are involved, and for clear reporting across submissions. When no approval is required, readers still need a sentence that explains why. Link your statement to an institutional determination or to the governing regulation.
Assess The Ethics Of Included Trials
While your project may not need new approval, you still judge the included trials. Screening for ethics approval and consent in the primary trials supports research integrity and is part of a sound risk-of-bias workflow in medical syntheses. Recent work in BMJ Open has shown that missing ethics information in included trials can affect trust in the pooled findings. Cite that assessment step in your Methods.
Documentation To Keep On File
Editors and auditors sometimes ask for proof. Keep the following in your repository or project folder:
- Copy of any “not human subjects research” or “exempt” determination letter from your IRB.
- Data-use agreements and security plans for IPD or registry work.
- A short memo describing how you avoided identifiable information during screening and extraction.
- Links to the reporting guideline you followed (PRISMA checklist) and the methods reference you used (e.g., Cochrane Handbook).
How To Write The Ethics Line Editors Expect
Use straight, plain wording. Two to three sentences usually do the job:
Template A: No Approval Required
“This project summarized published reports and used only aggregate, public data; per institutional policy and 45 CFR 46, oversight was not required.”
Template B: Determination Letter On File
“The IRB at [Institution] issued a Not Human Subjects Research determination (Ref #____); no participant contact or identifiable data were used.”
Template C: Approval Granted
“The REC/IRB at [Institution] approved the protocol (Ref #____). Individual-level data were analyzed under a data-use agreement with access controls.”
Common Misconceptions To Avoid
“PRISMA Is An Ethics Approval”
PRISMA is a reporting checklist, not an approval body. It helps readers understand why you did the review and what you found. Oversight questions sit with your IRB/REC and with data-protection laws.
“All Reviews Are Exempt By Definition”
Not correct. Any project that touches identifiable information or interacts with people can require review. When in doubt, obtain a written determination from your local office. U.S. and UK sources linked in this article show the boundary clearly.
“Journals Don’t Care If No People Were Involved”
Editors still expect a line that states the status. BMJ and other ICMJE-aligned outlets ask for an ethics statement in every research article, even when the answer is “not required.”
Practical Workflow You Can Reuse
- Draft your protocol. State the data sources and whether you will access any individual-level material.
- Check the boundary. If you will only use published aggregate data, prepare a “not required” statement. If you will handle IPD or records, route for IRB/REC review or an exemption.
- Document the call. Save determination letters and data-use agreements in your repository.
- Report transparently. Follow PRISMA for the manuscript and cite your methods foundation, such as the Cochrane Handbook.
- Assess included studies. Record ethics approval and consent status for each trial during data extraction.
Key Takeaways You Can Apply Today
- Published, aggregate data only → standard case with no IRB/REC review.
- Identifiable or re-linkable data → expect review or a formal exemption.
- Any contact with people → treat as human-subjects work and seek oversight.
- Journals still want a one-sentence ethics statement in every research article.
- Use PRISMA and methods references such as the Cochrane Handbook for clean reporting.