Can You Stop Searching Even Though You Know More Relevant Papers Probably Exist?

Must you find every relevant paper before you are allowed to stop?

Suppose you have searched the major databases relevant to your topic, checked reference lists, followed citations, tried alternative terminology, and reviewed hundreds of records. New relevant papers still occasionally appear.

You could probably find more if you kept going. Does that mean you have to?

This is one of the uncomfortable realities of literature searching: stopping rarely means proving that nothing else exists. Even systematic-review searches may miss relevant studies, and methodological guidance recognizes both the importance of comprehensive searching and the practical problem of deciding when further searching has sufficiently little value.

The relevant question is therefore not simply whether another paper exists. It is whether the possibility of additional unidentified papers creates enough risk to the purpose, interpretation, or conclusions of your work that searching should continue.

Yes, but the reason for stopping must be defensible

Completeness is not the same as certainty that nothing remains

There is an important difference between exhaustive and sufficiently comprehensive

An exhaustive search sounds simple in principle: keep searching until every relevant publication has been found. The difficulty is epistemic. To know that you have found every relevant paper, you would somehow need to know the complete population of relevant papers against which your results could be compared.

Usually, that population is precisely what you are trying to discover.

This is why being certain that you found every relevant paper is a much stronger claim than demonstrating that you conducted an appropriately comprehensive search.

Cochrane describes systematic-review searching as systematic and comprehensive and recommends searches that aim to identify as many relevant studies as possible. At the same time, its guidance acknowledges that searches may miss relevant studies and that determining objectively when search development is complete is difficult.

The fact that another paper exists does not tell you how much that paper matters

Imagine two unidentified studies.

The first is another small study using a familiar method in a population already well represented in your review. Its findings closely resemble those of many included studies.

The second examines a population absent from your review and reports results that challenge your emerging conclusion.

Both are relevant. Their potential consequences are very different.

This distinction is central to a sensible stopping decision. The existence of additional relevant literature matters, but so does the likely informational value of what remains. Campbell guidance explicitly recognizes diminishing returns in searching: after some point, additional effort tends to yield progressively fewer relevant references. It also notes that the appropriate investment depends on the review question and available resources.

The acceptable risk of missing papers depends on what you are doing

Qualitative evidence-synthesis guidance makes this variation particularly clear. Where authors do not intend a comprehensive search, they should justify their stopping decision. Where comprehensiveness is intended, the concept of bibliographic sufficiency may be relevant when extending terms or sources ceases to identify additional relevant studies.

Diminishing returns can justify stopping, but inconvenience cannot

Searching has costs. Every additional database, citation trail, search term, and screened record requires time that could otherwise be spent appraising studies, analyzing evidence, writing, or improving other parts of the research.

That creates a genuine trade-off. Campbell guidance explicitly recognizes diminishing returns in literature searching rather than assuming that every additional hour has equal value.

But diminishing returns should not become a respectable-sounding synonym for impatience.

If an additional search route is still identifying important studies, an unexplored database is particularly relevant to the topic, or your search consistently fails to retrieve known key studies, stopping is harder to defend. Cochrane recommends checking whether strategies retrieve known relevant publications and using citation searching and reference checking as additional indicators of search performance.

Stopping is ultimately a decision under uncertainty

Researchers have proposed statistical methods for estimating how much relevant literature may remain unidentified. Capture-mark-recapture, for example, uses overlap among searches to estimate the possible population of studies. Research has shown that such approaches can potentially inform stopping decisions, but suitable universal cut-offs have not been established.

In ordinary research practice, you will therefore rarely receive a mathematical certificate saying, "Search complete."

You instead accumulate evidence that the search is performing adequately: major sources have been covered, alternative terminology has been explored, known relevant studies are retrieved, supplementary searches mostly converge with the main search, and additional effort produces increasingly little consequential evidence.

This is closely related to recognizing when further searching stops improving the review. The stopping point is not where uncertainty reaches zero. It is where the remaining uncertainty becomes acceptable for the purpose of the work.

What if you are still finding occasional relevant studies?

Now change one detail. Suppose citation searching suddenly identifies a cluster of 18 relevant studies from a research tradition absent from the database results. That would be evidence that the apparent saturation was misleading. The search should probably be reconsidered because the newly discovered cluster reveals a systematic gap rather than a handful of residual papers.

Why knowing that more papers exist does not settle the decision

Ask what you risk by stopping, not whether another paper exists

When deciding whether to stop, imagine that another relevant paper will eventually be found. Then ask what kind of paper would make you regret stopping.

One particularly useful next question is whether additional searching is likely to change the conclusion. That does not replace methodological requirements, but it forces you to think about the consequence of what might still be missing rather than merely the existence of more literature.

Whatever decision you make, preserve enough information to explain it later. A stopping decision becomes much easier to defend when it follows an observable pattern of search convergence rather than an undocumented feeling that you have probably read enough.

Before stopping while more papers may still exist

Questions about stopping before every paper is found

You can stop without claiming that nothing remains

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.