A saved PDF may not display its present status
A file downloaded before the retraction, or copied into another repository, may carry no visible warning. Start with the DOI and official journal record. Look for banners, watermarks, linked notices and updated metadata. Search the full title and DOI in an established subject database as a second route.
Do not rely on one search snippet. A library catalogue, reference manager or inherited reading list can be stale. A news post may alert you to a problem, but the journal's notice and registered metadata are the primary documents for the publication status and stated reason.
Distinguish retraction from other updates
A retraction marks a published article as unreliable in important respects or withdrawn for another documented reason. An expression of concern warns readers of serious unresolved questions while an investigation proceeds. A correction addresses specified errors without necessarily invalidating the whole paper. Publishers may use “withdrawal” for records removed at different stages.
Read the notice rather than inferring meaning from the label. An action may concern a figure, dataset or section, or it may undermine the central conclusion. Status can develop over time. Record the notice date, issuing body, identifier and its explicit relationship to the original.
Let the documented reason shape your response
COPE guidance emphasises clear reasons, linkage and continuing accessibility of the record. Reasons can include honest error, unreliable data, duplicate publication, ethical breach or manipulation. Retraction is not automatically proof of intentional misconduct by every author. State only what the notice supports.
Ask which claim in your work depends on the article. If a core dataset is judged unreliable, the finding cannot remain a normal foundation. If the notice concerns another component, a reasoned reassessment is still required. Search for independent studies, reanalysis, replications and subsequent corrections.
A retracted article can still be a research object
It should not support the withdrawn scientific proposition as though unaffected. It may, however, be evidence in scholarship about publication ethics, error propagation, citation networks or the history of a controversy. In that setting, every use needs an unmistakable status marker. Cite the retraction notice as well and paraphrase its stated reason without speculation.
A transparent sentence might read: “The article was retracted on [date]; the notice states [brief reason]. It is examined here only as part of the earlier reception.” Avoid personal allegations not established by the notice. Source evaluation does not determine individual culpability.
Audit the references that carry your argument
Prioritise sources supporting central conclusions, controversial claims and material saved long ago. Export titles and DOIs for organisation, but inspect authoritative status records individually. Automated alerts can help, yet matching errors and delayed updates occur. The scholarly response requires documented judgement.
When a problem appears, find every occurrence in prose, notes, tables and figures. Deleting the bibliography entry does not repair an argument shaped by it. Reassess claims and, where needed, results. Check early manuscripts separately using the guide to preprint versions and status.
Respond to a retraction discovered after submission
If a central source is retracted later, preserve the original check and obtain the new notice. The available correction, repository note or formal notification depends on the institution or publisher. A dated erratum is more transparent than silently replacing a public file. Explain which passage was affected and whether your conclusion changes.
Build a final literature-status check into future projects. Resolve DOIs, review update indicators and store the date. This does not guarantee that every developing investigation is known, but it creates a reasonable and reproducible step. Recheck immediately when a supervisor, peer reviewer or database alert raises a specific concern.
For a literature review, retain the retracted record in the screening log with its notice, exclusion reason and check date. This preserves the historical search trail and explains why a later synthesis contains fewer studies than an earlier draft.