Published articles may need to be corrected or amended for various reasons. The best matching scenario can be chosen from several possibilities surveyed on this Correction Policy Page, all of them consistent with the Crossmark Policy, the COPE recommendations and our own Ethical Guide.
The most important principle to follow is that any changes made in a published article, with a Digital Object Identifier, must be clearly track recorded and unambiguously recognisable. The ways applied to achieve this aim in our journals are summarised in the following chart.
Various types of errors and mistakes may need to be corrected; the way they can be treated depends on their nature. Typographical mistakes and misprints need to be corrected only if they jeopardise readability or indexing. Errors in the scientific content can only be corrected if they do not affect the main results and conclusions of the paper, and no concerns about its integrity and reliability arise (otherwise different steps described below will be necessary).
Whichever correction method is chosen, the key principle is transparency: it should be unambiguously visible what corrections have been made, when and why.
In situ correction
Obvious misprints, especially in the metadata of an article, as well as smaller technical or display problems, can be corrected in the published file. Such corrections must be documented in footnotes displaying the date and the nature of the correction, and the thus corrected file will replace the original that contained the typo. The publisher may decide to correct an error discovered by the authors, the editors, the publisher, or any reader, informing the Editor-in-Chief about the modification.
Should a mistake by the authors be discovered in a published paper, the Editor-in-Chief must be informed. Based on consultation with the authors, the E-i-C decides about the publication of a separate paper called "Corrigendum" in which the precise nature of the error is explained and the corrected statement/data/information is provided. The Corrigendum will be assigned its own Digital Object Identifier and its online version will be mutually interlinked with the original article: a red label on both articles' web page will indicate that "a related paper is available". It is important to stress that the original, erroneous article is not modified in this case, only the link to the Corrigendum refers to the fact that it has been corrected.
An Erratum is similar to a Corrigendum but it serves to correct a mistake made by the publisher or the editorial office. It is again the Editor-in-Chief who may decide about the publication of an Erratum, detailing the precise nature of the error and the corrected form. The Erratum will receive its own Digital Object Identifier and its online version will be mutually interlinked with the original article: a red label on both articles' web page will indicate that "a related paper is available". Again, the original, erroneous article is not modified, only the link to the Erratum shows that it has been corrected.
In certain rare cases when the nature of the error fits the aforementioned requirements of corrections, but the mere publication of a correction note would not result in a readable paper – e.g., if references or bigger parts of an article are affected – the Editor-in-Chief and the publisher may decide to replace the erroneous original paper with a corrected version. Such a change must be documented in the corrected paper and, on top of that, in a simultaneously published correction note (a Corrigendum or an Erratum, depending on the source of the error). The correction note will get a separate Digital Object Identifier and it will be interlinked with the thus corrected original paper.
Published papers may need to be updated for many reasons, such as additional findings, concerns regarding compliance with various policies, and further changes. Any such situation can be handled by a suitable post-publication note.
Authors of fully correct papers may wish to add their newer achievements — not qualifying for a separate publication — to their published papers in the form of an Addendum. Such manuscripts should be submitted to the Editor-in-Chief, who will decide about its acceptance or rejection as in the case of any other submission. If accepted and published, the Addendum will get a separate Digital Object Identifier and it will be interlinked with the corresponding paper that it supplements (leaving the original paper unchanged).
Editor-in-Chief's expression of concern
In the regrettable cases when serious doubts arise about the integrity or reliability of a paper, a deeper investigation is needed. Temporarily while investigations are ongoing, or permanently if investigations turn out to be inconclusive for any reason, the Editor-in-Chief has to inform readers about the concerns. The Expression of Concern will get a separate Digital Object Identifier and it will be interlinked with the related paper (which remains available in an unmodified form).
Any unusual aspect related to or problem with the publication of a paper that may be of relevance to the readers may be added as a note by the Publisher — provided that the Editor-in-Chief approves it. The Publisher's Note will get a separate Digital Object Identifier and it will be interlinked with the paper in question (which remains available in an unmodified form).
One common reason for retraction is a serious flaw in the scientific content; that is, a substantial error that invalidates a paper's results. Depending on whether the main conclusions are affected or only certain parts, a partial or complete retraction will be needed.
In other, more unfortunate cases retraction of a paper may be necessary because of ethical misconduct during the research or the publication phase. In such situations, the decision about the retraction must be based on an in-depth investigation involving all parties, and must be based on clear evidence.
Whenever the authors initiate the retraction of their paper — after discovering or learning about an essential error in their results — they should inform the Editor-in-Chief. After the approval of the E-i-C, a Retraction Note will be published containing the date and the reason of the withdrawal and the precise identification of the retracted part. The Retraction Note will get a separate Digital Object Identifier and it will be interlinked with the original paper in question. Even if the entire paper is retracted, only its full-text html version will be removed. The pdf file will remain available in an unchanged form, but its pages will carry the label ’RETRACTED ARTICLE’. In addition, the Retraction Note will show that the paper has been withdrawn and give the reason for its retraction.
Papers can also be retracted without the consent of the authors at the Editor-in-Chief's discretion (for scientific or ethical reasons). The procedure is analogous to the authors' retraction: the Retraction Note will get a separate Digital Object Identifier and it will be interlinked with the retracted paper whose pdf file remains available in an unmodified form, its pages bearing the label ’RETRACTED ARTICLE’. The full-text html version will be removed.
The publisher may only retract a paper on legal grounds; e.g., if plagiarism or multiple publication has been proven. Then again, the procedure is analogous to the authors' retraction: the Retraction Note will get a separate Digital Object Identifier and it will be interlinked with the retracted paper whose pdf file remains available in an unmodified form, its pages bearing the label ’RETRACTED ARTICLE’. The full-text html version will be removed.
Complete removal of a published paper is only possible in rare exceptional cases, by court order or by a committee decision declaring that its availability poses a real moral or practical threat. A removed paper will be replaced by a Removal Notice with a separate Digital Object Identifier.