01 · The Question
What Happens to a Review When One of Its Included Studies Is Retracted?
A systematic review or meta-analysis may have been conducted correctly using the publication record available at the time. Months or years later, however, one of its included studies can be retracted.
That creates a second-order problem. The retraction affects the primary study, but the study's data, effect estimate, or conclusions may already have been incorporated into a review that other researchers now treat as higher-level evidence.
Does the entire review become unreliable? Should the retracted study simply be deleted? Or does the synthesis need to be recalculated?
The answer depends on what the retracted study contributed and how much the review's findings depend on it.
03 · What You Need to Know
A Retraction Can Propagate Into Secondary Evidence
The Review May Have Been Correct When It Was Conducted
A later retraction does not necessarily mean that the review authors originally made an error. The primary study may have had no retraction notice when the review's searches and analyses were completed.
This distinction matters when interpreting an older review. Systematic reviews represent an evidence base assembled at a particular time. The underlying literature can subsequently change through new studies, corrections, expressions of concern, and retractions.
Cochrane therefore advises review authors to examine relevant retraction notices and errata both when studies are first considered and when reviews are updated. Its current guidance specifically notes that previously included studies may later have been corrected or retracted.
First Determine What Was Actually Retracted
A study and a publication are related but not always identical units. One study can generate several journal articles, conference reports, follow-up papers, or secondary analyses. A retraction may apply to one publication associated with a study rather than automatically to every report arising from the same research project.
Start by reading the formal retraction notice. Determine which publication was retracted, why, and what findings or data are affected. COPE recommends that retraction notices identify the affected article and state the reason for retraction.
If the retraction makes the study data used in the review unreliable, those data require direct attention. If the situation is more complicated, avoid assuming either that everything associated with the study is invalid or that unaffected-looking data remain trustworthy without further examination.
Trace the Retracted Study Through the Review
Once you know what has been affected, identify where the study contributed to the review.
Possible contribution
What to check
Potential consequence
Meta-analysis
Which effect estimates came from the retracted study
Pooled estimates, confidence intervals, heterogeneity, or conclusions may change
Narrative synthesis
Which claims rely on the retracted findings
Strength or direction of the narrative conclusion may change
Subgroup analysis
Whether the study dominates a small subgroup
A subgroup finding may weaken or disappear
Rare adverse outcome
Whether the study supplies unusual or unique events
Risk interpretation may change substantially
Methodological discussion
Whether the review relies on unaffected or affected material
Impact depends on the scope and reason for retraction
Certainty assessment
Whether removing the study changes the body of evidence
Confidence in the review's conclusion may need reassessment
Do Not Assume a Small Study Has a Small Effect
The number of participants in the retracted study does not by itself tell you how much its removal will matter. Its influence depends on the synthesis.
A relatively small study could be the only evidence for a particular subgroup, outcome, intervention, or population. An unusually large effect could also influence a pooled estimate or the interpretation of heterogeneity. Conversely, a large retracted study might have little effect on a particular conclusion if numerous independent studies provide consistent evidence.
The relevant question is therefore counterfactual: what does the review show when the affected evidence is no longer allowed to support the conclusion?
For Meta-Analyses, Recalculate When Possible
If the retracted study contributed numerical data to a meta-analysis and you have access to the necessary information, removing its contribution and recalculating the synthesis can show whether the pooled result changes materially.
You might examine changes in the pooled effect estimate, uncertainty, heterogeneity, subgroup findings, and the substantive interpretation of the result. A threshold such as statistical significance should not be the only criterion. A pooled estimate can remain on the same side of a significance threshold while changing enough to matter scientifically or practically.
Cochrane's guidance for updating reviews states that where data appear incorrect or possibly fabricated, they should be removed from the review analysis and that this decision should be reported. In 2026, Cochrane also introduced a framework specifically for determining the impact of studies with associated retractions on published reviews.
Watch Out
Do not call a recalculation a sensitivity analysis if the retracted evidence is no longer eligible to support the primary synthesis under the review's applicable rules. A sensitivity analysis can help show how influential the study was, but it does not necessarily justify retaining unreliable evidence in the main estimate.
Narrative Reviews Need Reassessment Too
Retractions are not only a meta-analysis problem. In a narrative review, an influential retracted study may support a central statement, establish an apparent pattern, supply the only evidence for a particular proposition, or shape the authors' interpretation of conflicting results.
Remove the retracted study mentally from the evidential chain and ask whether the same conclusion still follows from the remaining studies.
This can be surprisingly revealing. A paragraph containing six citations may appear well supported until you discover that several of those papers merely repeat a claim originating in the same retracted study. Tracing continued citations of retracted evidence can therefore be necessary even when the retracted article appears only once in the review's reference list.
The Review Itself May Need Correction or Updating
If you are an author of the affected review, discovering the retraction should prompt an assessment of whether the published review needs an update, correction, editorial notice, or another response under the publisher's policies.
Cochrane's 2026 framework explicitly addresses published reviews containing studies associated with retracted articles. Cochrane notes that publishers should inform readers when included studies are retracted and assess the effect on the review's findings and conclusions. Where confidence in the review's conclusions is undermined, stronger editorial action may be necessary.
The appropriate response depends on impact. A retracted study whose removal leaves every important conclusion essentially unchanged presents a different situation from one whose removal reverses the principal finding.
If You Are Only a Reader, Check for an Updated Version
You may encounter the problem while reading someone else's systematic review. Before attempting to reconstruct the analysis yourself, check whether the review has already been updated, corrected, accompanied by an editorial note, or superseded by a newer synthesis.
If no update exists and the review matters to your work, inspect the contribution of the retracted study as far as the available reporting allows. If you cannot determine its influence, that uncertainty should temper how confidently you rely on the review.
Retraction Does Not Automatically Invalidate Every Other Included Study
The presence of one retracted study does not logically contaminate all independent evidence in the review. Other included studies should be evaluated on their own merits.
However, if the retraction involves fabrication or another integrity problem, additional publications from the same research group may warrant scrutiny when there are concrete reasons to suspect related problems. Cochrane guidance advises checking other eligible studies by the same authors for similar issues when data appear incorrect or possibly fabricated.
That is a reason for targeted verification, not guilt by association.
07 · A Quick Checklist
When You Discover a Retracted Study Inside a Review
Before relying on the review, check:
Verify the retraction through the publisher, formal notice, or authoritative bibliographic record.
Read why the publication was retracted and determine which data or findings are affected.
Identify every analysis, table, outcome, subgroup, or narrative conclusion to which the study contributed.
Check whether the review has already been updated, corrected, or accompanied by an editorial notice.
For quantitative synthesis, examine the result without the affected study where possible.
For narrative synthesis, determine whether the remaining independent studies still support the same conclusion.
Reconsider certainty or confidence in the evidence if removal changes the body of evidence materially.
Document any exclusion, reanalysis, or change in interpretation transparently if you are updating the review.
11 · Cite this Guide
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.
Recommended (Field Guide)
APA
MLA
Chicago
Copy Citation
