Withdrawing articles published in the CDSR


Providing guidance on the Cochrane Library editorial policy on withdrawal of published articles. 





Cochrane reviews

Cochrane reviews should be withdrawn only under exceptional circumstances. Such circumstances include concerns about the conduct of the author team or about the reporting in the Cochrane review.

Serious error

Errors originating from how the review was prepared, or as a result of the retraction from publication of one or more of the review’s included studies. See Serious errors in published articles.

Scientific misconduct

This could refer to fraud or other scientific misconduct in the preparation of the review.

Serious breach of Cochrane's conflict of interest policy

A ruling by Cochrane’s Conflict of Interest Panel that a Cochrane review has seriously breached Cochrane’s conflict of interest policy.

These criteria are in line with COPE's Retraction guidelines, which are formal COPE policy.

The following do not consitute exceptional circumstances: questions related to the usage and currency of the review; the availability of new studies or information, and how this information could impact the review; and whether there are new methods that could make important changes to the review.

The Editor in Chief must approve each withdrawal of a Cochrane review.

Only the most recent version of a Cochrane review can be withdrawn. Withdrawing a Cochrane review generates a new citation version, which results in a new entry in PubMed. The PubMed entry displays as the article title preceded by “WITHDRAWN:”, followed by the abstract of the withdrawn version.



Protocols for Cochrane reviews

As of March 2022, protocols must not be withdrawn for any reason other than serious error, scientific misconduct, or serious breach of CoI policy (see definitions under 'Cochrane reviews', above).

Prior to March 2022, protocols could be withdrawn for the following reasons. These circumstances are now covered by Editorial notes:

  • The protocol was out of date and did not meet Cochrane’s current methodological standards.
  • Authors had made no progress with the protocol in XX months/years, and new authors were being sought to take over the protocol.
  • The protocol had been split into or merged with other protocols.

The Editor in Chief must approve each withdrawal of a protocol for a Cochrane review.

As with the withdrawal of a review, withdrawing a protocol for a Cochrane review generates a new citation version and new entry in PubMed.



Republication after withdrawal

The editorial policy on withdrawal of published articles provides for the republication of a withdrawn article.

Also see Select the correct What's new event in RevMan Knowledge Base for cases where a withdrawn article is republished in order to change text in the withdrawn notice.



Steps to withdraw a published article

1. Notify authors

The Managing Editor must email the contact author, with copies to all co-authors, to

    • notify of the intention to withdraw the review or protocol;
    • give the opportunity for discussion; and
    • ask the authors if they agree with the proposed withdrawal. The authors' agreement or disagreement should be included in the withdrawn statement.

2. Draft the withdrawn statement

Each withdrawn statement must include key information and must be kept neutral and factual to avoid potential defamation.

Statements for Cochrane reviews must include:

    • A first sentence beginning: "This Cochrane review has been withdrawn from publication"
    • Who is withdrawing the review, and whether the authors agree
    • The reason(s) for withdrawal
    • Which section(s) of the review are affected
    • Whether the intention is to revise and republish the review in the future

Statements for protocols for Cochrane reviews must include:

    • A first sentence beginning: "This protocol for a Cochrane review has been withdrawn from publication"
    • Who is withdrawing the protocol, and whether the authors agree
    • The reason(s) for withdrawal
    • Whether the intention is to revise and republish the protocol in the future

3. Seek approval from Editor in Chief

The Managing Editor must email the Editor in Chief to seek advance approval for the withdrawal. Provide the

    • Title of the review or protocol
    • DOI of the version being withdrawn
    • Reason for withdrawal (preferably including the drafted withdrawn statement)

4. Notify Evidence Production & Methods Directorate

After receiving approval from the Editor in Chief, forward the approval email to the Evidence Production & Methods Directorate (emd@cochrane.org), including all information from step 3 (title, DOI, and reason for withdrawal).

5. Withdraw the Cochrane Review

Complete steps 1-4 before processing a withdrawal. Follow the steps available in the Archie Knowledge Base; see also Withdrawing published content in the Editorial Manager Knowledge Base.



Display of published withdrawn articles

Cochrane reviews and protocols that have been withdrawn are not removed from the CDSR. Instead, they are retained in the CDSR and the relevant version of the article is flagged as 'withdrawn'.

  • Published versions that precede a withdrawn review are available via the 'History' section of the Cochrane review.
  • Up to 16 December 2014, withdrawn protocols were removed from the CDSR after one issue of publication.

The following information should be included in the withdrawn version:

  • Reason for withdrawing the protocol/review (as inserted in the 'Published Notes' section in the RevMan file), preceded by "Reason for withdrawal from publication".
  • The following text inserted automatically after "Reason for withdrawal from publication": "The editorial group responsible for this previously published document has withdrawn it from publication."
  • Standard publication information: Title, authors (as listed on previous citation version), citation, DOI, publication date, name of editorial group, sources of support (as listed on previous citation version), and publishing events.
  • Citations of the withdrawn version, as available via Crossref data.



Back to Research integrity  •  EGR homepage