What changes when a study is retracted

The National Library of Medicine links the original article in PubMed to the notice published by the journal. That relationship distinguishes the retracted publication from the document reporting its retraction.

The bibliographic record does not replace reading the notice. NLM does not automatically separate retractions for honest error from those involving misconduct. The label therefore does not justify alleging fraud: the reason must come from documentation stating it.

PubMed records a 2025 notice retracting the 2014 article on angiotensin IV derivatives and the HGF/c-Met system. The link identifies exactly which publication was affected.

For a product page citing that work, the immediate consequence is reviewing claims that depended on it. Retaining the mechanism and merely appending a notice link is insufficient. Nor does it follow that every study of the molecule has been retracted: each reference retains its own status.

A citation analysis published in 2022 examined retracted systematic reviews. It found that they continued receiving citations, many from articles published after the notice. Only a small proportion identified the retraction in the bibliography.

This observation explains a specific risk: a recent review may repeat an old claim without updating its foundation. A recent date on the citing document does not guarantee that the status of every source was checked.

A useful way to correct content is to build a short chain: claim, direct source and additional independent evidence. If a review refers back to the retracted article, it does not count as new confirmation of the same result.

Then distinguish two situations. Other research may support a more limited version of the claim, or sufficient support may not have been found. In the first case, rewrite within that source's scope; in the second, remove the conclusion or present it only as historical background.

The editorial approach is to record which statement changed, which notice prompted the review and which sources remain as support. This prevents an earlier version from reintroducing the claim when text is reused.

The original article remaining accessible in PubMed does not contradict its retraction: the system preserves the documentary relationship. It can be consulted to study the history of the finding, provided its status is explicit. Presence in a database does not restore validity to affected conclusions.