How Do You Know Whether Existing Reviews Already Answer the Question?

Finding a Review on Your Topic Is Only the Beginning

You search for literature related to your proposed review and find several systematic reviews with titles that look uncomfortably close to yours. At this point, it is tempting to conclude that somebody has already answered your question.

But similarity of topic is not the same as equivalence of question.

An existing review may examine a different population, intervention or phenomenon, comparison, outcome, setting, study design, or period of evidence. It may also appear to match your question but contain methodological limitations that affect how confidently its conclusions can be used.

Before deciding that your planned review is redundant, you need to compare what you want to know with what the existing reviews actually established.

Compare Questions and Evidence, Not Just Titles

How to Test Whether an Existing Review Really Answers Your Question

Begin by writing your question independently

Before comparing your proposed review with existing reviews, state your intended question clearly. Otherwise, there is a subtle risk of reshaping your question after seeing what others have already reviewed.

The relevant elements depend on the type of review. For intervention questions, these often include the population, intervention, comparator, and outcomes. Other review questions may instead be structured around exposures, phenomena, concepts, contexts, diagnostic questions, experiences, or other dimensions.

Cochrane emphasizes that a well-formulated review question guides eligibility criteria, searching, data collection, synthesis, and presentation. It also recommends identifying related or overlapping reviews before beginning a new review so that unnecessary duplication can be reduced.

Once your own question is explicit, you have something concrete against which existing reviews can be compared.

Separate a shared topic from a shared question

Suppose your proposed question concerns whether generative AI-supported feedback improves undergraduate students' academic writing. You find a systematic review titled “Artificial Intelligence in Higher Education.” The topical overlap is obvious, but the title tells you almost nothing about whether your specific question has been answered.

The review might concentrate on attitudes toward AI. It might include every form of artificial intelligence but provide little evidence about feedback. It might combine school, university, and professional education. It might examine adoption rather than learning outcomes.

The distinction is fundamental. A crowded topic can still contain unanswered questions, while a narrowly defined question may already have been comprehensively synthesized.

Compare the eligibility criteria element by element

The inclusion and exclusion criteria usually reveal the operational meaning of the review question more precisely than its title.

Construct a comparison between your proposed criteria and those of the closest existing reviews. Depending on the question, examine population, intervention or exposure, comparator, outcome, context, study design, publication period, language restrictions, and other substantive boundaries.

Not every mismatch creates a review gap. The question is whether the difference changes what can reasonably be concluded.

Check what evidence the review could actually have found

A review may ask almost exactly your question yet still fail to contain the evidence needed to answer it adequately.

Inspect the databases and other information sources searched, the search strategy where available, restrictions applied, and the date of the final search. Also examine whether relevant studies you already know about appear among the included studies.

A recent publication date does not guarantee a recent or adequate search. Conversely, an older review may remain informative if little consequential evidence has appeared since its search.

If a seemingly relevant review searched inadequately, its recency alone should not persuade you that the question is settled. A recent review can still leave an evidence gap when its search was poor.

Determine whether the methods support the answer

A review can address the right question and still provide an unreliable answer.

Methodological appraisal therefore matters. Depending on the type of review, consider whether study selection and data extraction were appropriately conducted, risk of bias was assessed, synthesis methods suited the data, heterogeneity was handled appropriately, and limitations of the evidence influenced the interpretation.

Tools can assist with this appraisal when they fit the review type. AMSTAR 2, for example, was developed to critically appraise systematic reviews of healthcare interventions that include randomized or non-randomized studies. ROBIS instead focuses specifically on risk of bias in systematic reviews and has broader applicability across several types of review questions. These tools overlap, but they do not assess exactly the same construct.

Read the results, not merely the authors' conclusion

Authors may conclude that an intervention “appears effective,” that evidence is “promising,” or that further research is required. Those statements are interpretations. You need to understand what evidence produced them.

Look at the number and characteristics of included studies, consistency of findings, precision where relevant, risk of bias, heterogeneity, and any limitations affecting certainty or applicability.

A review can technically address your question while still leaving substantial uncertainty. In that situation, the question has been reviewed, but it may not have been conclusively answered.

Ask whether the answer applies to your intended context

Sometimes the evidence is reasonably complete for one population or context but uncertain for another. That does not mean every demographic, institution, profession, or country deserves its own systematic review.

Instead, identify a credible reason why the distinction could affect the evidence or its interpretation. If your proposed review differs mainly by population, examine whether the population difference could meaningfully change the answer.

The same caution applies geographically. The fact that no review has been published specifically for your country does not itself demonstrate that evidence synthesized elsewhere is inapplicable.

Determine whether the evidence base has changed

Even a rigorous review can become incomplete when consequential new evidence appears.

Do not merely count publications appearing after the previous search. Examine whether they are eligible, how much evidence they contribute, whether they address previous weaknesses, and whether their findings could alter interpretation.

A particularly influential new study may sometimes warrant reconsideration of an existing synthesis. The relevant question is whether the new study could materially change what the review currently tells us, not simply whether it is large or highly visible.

Distinguish “answered” from “reviewed”

This is perhaps the most useful distinction.

A question can therefore be reviewed repeatedly while remaining uncertain. Equally, a single rigorous and current review may already provide a sufficient synthesis.

PRISMA 2020 reflects this logic. When similar systematic reviews already exist, authors are expected to explain why another review was considered necessary, with examples including outdated reviews, discordant results, methodological flaws, or newer review methods capable of addressing the question differently.

Three Reviews on the Topic, but Do They Answer the Question?

The researcher should not claim a gap simply because Reviews A and B differ from the proposed review. Once Review C is identified, the burden shifts to explaining what another synthesis would add.

Why Researchers Misjudge Whether a Question Has Been Answered

Make an Answered-Question Assessment Before Starting the Review

For each existing review that appears close to your proposed question, document the comparison rather than relying on an impression of similarity.

What you are looking for is not an excuse to conduct your preferred review. You are testing whether an unresolved problem actually remains.

Check Whether an Existing Review Already Answers Your Question

Questions About Existing Reviews and Research Gaps

Judge the Evidence the Review Actually Synthesized

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.