Do systematic reviews need ethics approval? Usually not, because they use published data. The exceptions that do require review, and the ethical duties that apply regardless.
This question comes up early, usually from a student who has been told that all research needs ethics approval and cannot work out how that applies to reading papers. The answer is reassuring in most cases and has enough exceptions to be worth understanding properly.
The general answer is no. But "no ethics committee" is not the same as "no ethical obligations," and the exceptions catch people out.
A standard systematic review of published, aggregate data does not usually require ethics approval, because it does not involve human participants directly and uses data that is already in the public domain. Most institutions and ethics committees treat it as exempt or outside their remit.
Exceptions exist. Reviews involving individual participant data, unpublished or non-public data, or primary data collection (such as surveying or interviewing authors) may require approval. And some institutions require you to submit for a formal exemption decision rather than deciding for yourself.
Always check your own institution's policy. Requirements vary, and the safe move is to ask rather than assume.
Research ethics review exists primarily to protect human participants: their safety, their consent, their privacy. A standard systematic review interacts with none of these. You are not recruiting anyone, intervening on anyone, or handling anyone's identifiable information. You are reading and analyzing results that researchers have already published, and whose participants already consented to the original studies under their own ethical approvals.
There is no new risk to participants, because there are no new participants. That is the logic, and it is why systematic reviews of published aggregate data are routinely treated as exempt from review.
Individual participant data (IPD) reviews. If you obtain the raw, participant-level datasets from the original studies rather than their published summaries, you are now handling individual-level data. Depending on the data, its identifiability, and the agreements governing it, this can require ethics approval and will certainly involve data-sharing agreements and data protection obligations. This is the most common genuine exception.
Unpublished or non-public data. Reviews that draw on data not in the public domain, obtained directly from researchers, companies, or institutions, may fall under different rules, particularly around confidentiality and data governance.
Primary data collection within the review. Some reviews include a component that is genuinely primary research. Surveying study authors, interviewing stakeholders, or running a consultation exercise, as some scoping reviews do, means you are collecting data from people. That component may need approval even though the review itself would not.
Sensitive or identifiable information. If your review would involve handling identifiable information about individuals, ethics and data protection obligations apply regardless of the review label.
Institution-specific requirements. Some institutions require every research project to be submitted, if only to receive a formal exemption. Some journals ask you to state your ethics status. The absence of a participant risk does not always mean the absence of a paperwork requirement.
| Situation | Ethics approval? |
|---|---|
| Standard review of published aggregate data | Usually not required |
| Review using individual participant data | Often required; check governance rules |
| Review using unpublished or non-public data | May be required |
| Review with author surveys or stakeholder interviews | Likely required for that component |
| Review handling identifiable information | Required |
| Any review, at an institution requiring formal exemption | Submission required, even if exempt |
No ethics committee does not mean no ethics. Several duties bind you regardless of whether a review board ever sees your project.
Honest reporting. Report what you found, including findings that contradict your expectations or your prior work. Selectively presenting results is a research integrity failure regardless of participant involvement.
Transparency about methods. Disclose your protocol deviations, your limitations, and anything you streamlined. A review that hides its shortcuts misleads readers, which is an ethical problem, not merely a methodological one.
Avoiding research waste. Duplicating an existing review that already answers your question, or producing a review too poorly conducted to be useful, wastes resources and clutters the literature. Checking for existing reviews and registering your protocol are ethical acts as well as methodological ones.
Declaring conflicts of interest. Funding sources and competing interests must be declared. Readers cannot weigh your conclusions without knowing who paid for them.
Proper attribution and authorship. Credit the studies you draw on, and assign authorship honestly among your team. Ghost authorship and gift authorship are integrity failures.
Respecting the underlying participants. Even at one remove, the data you synthesize came from people who consented to a study. Handling it carelessly, or misrepresenting what it shows, is a failure of respect toward them.
Assuming exemption without checking. Institutional policies differ. Assuming you are exempt and discovering otherwise at submission is an avoidable problem. Ask your ethics office early.
Forgetting the IPD distinction. Teams sometimes drift from aggregate data toward individual participant data without recognizing that the ethical picture has changed. It has.
Overlooking an embedded primary component. A consultation exercise or an author survey is primary data collection, even inside a review. Treat it as such.
Treating exemption as the end of ethics. Honest reporting, transparency, and conflict declaration apply to every review. Exemption from committee review is not exemption from research integrity.
Do systematic reviews need ethics approval? Usually not, when they use published, aggregate data and involve no human participants. Check your institution's policy, as some require a formal exemption decision.
Why are systematic reviews usually exempt? Because they do not involve human participants directly. They analyze already-published results whose participants consented under the original studies' approvals, so no new participant risk is created.
When does a review need ethics approval? When it uses individual participant data, unpublished or non-public data, involves primary data collection such as author surveys or stakeholder interviews, or handles identifiable information.
Do I need approval for a meta-analysis? A meta-analysis of published aggregate data generally does not. A meta-analysis of individual participant data may, and will certainly involve data governance requirements.
Do journals ask about ethics for reviews? Many ask you to state your ethics status. Being able to say clearly that approval was not required, and why, is the answer they need.
A standard systematic review of published aggregate data does not usually require ethics approval, because it creates no new risk to participants. The exceptions worth knowing are individual participant data, non-public data, and any embedded primary data collection, and some institutions require a formal exemption regardless.
Check your own institution's rules rather than assuming, and remember that the absence of a committee does not remove your obligations. Honest reporting, transparent methods, declared conflicts, and avoiding research waste apply to every review that has ever been written.
Running a review and keeping the process transparent? Verflux documents your screening and appraisal decisions as you go, so your methods are reportable rather than reconstructed.
All analytical features included in the free trial. No credit card, no installation, no R or Python.
Create free account