|Year : 2020 | Volume
| Issue : 4 | Page : 258-260
Corrections in biomedical literature: Errata, expressions of concern, and retractions
Keerthi Talari1, Vinod Ravindran2
1 Department of Rheumatology, Yashoda Hospitals, Secunderabad, Telangana, India
2 Centre for Rheumatology, Kozhikode, Kerala, India
|Date of Submission||11-Nov-2020|
|Date of Decision||13-Nov-2020|
|Date of Acceptance||16-Nov-2020|
|Date of Web Publication||18-Dec-2020|
Centre for Rheumatology, Kozhikode - 673 009, Kerala
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Talari K, Ravindran V. Corrections in biomedical literature: Errata, expressions of concern, and retractions. Indian J Rheumatol 2020;15:258-60
The process of the biomedical research publishing is not perfect. Guidelines and policies exist to avoid and correct any deviations or errors (including ethical issues) that slink in during research or publication. However, in cases where they have been overlooked, recourses are available on how they ought to be reported. This editorial covers these avenues (i.e., errata, expressions of concern, and retractions) for corrections in biomedical research publications.
A biomedical research publication is the outcome of the applied scientific rigor contributing to the evidence-based medicine, so notifying an error in it is as important as the research itself. Notification of errors in research and their publication are errata and expression of concern (EoC) and in its extreme form, retraction. Relevant bodies, namely, the International Committee of Medical Journal Editors (ICMJE) and Committee on Publication Ethics (COPE) have published comprehensive recommendations or guidelines on these. Individual journals/publishers who are not members of these bodies also either follow their policies or have their own set of similar policies.
COPE suggests erratum be published “If only a small part of an article reports flawed data, and especially if this is the result of a genuine error.” COPE also distinguishes erratum (a publishing error) from corrigendum (an author error). However, the ICMJE makes no such distinction and the National Library of Medicine publishes all errors as errata regardless of the origin or nature of error. Errata could include typographic errors, honest errors resulting in scientific data misinterpretation, authors' name inclusions or deletions, authors' affiliations, and rarely author name corrections (only when it makes a significant impact). Exchange of first name with last name and minor typographic errors which do not result in misinterpretation of findings or do not affect the reader's understanding of the article may not be published as errata.
EoC as originally introduced by the ICMJE is a publication notice generally made by an editor to draw attention to a possible scientific misconduct in a biomedical research publication. Alongside an EoC, editor is required to start appropriate proceedings to assess the integrity of work. The outcome of such an investigation should also be subsequently published. However, if evidence of misconduct is subsequently forthcoming, then an EoC would lead to a retraction.
Retraction is a formal withdrawal of an article from scientific literature which is published in the journal and also incorporated in the indexing databases. Retractions are usually effected when there are errors that misinform readers, when the data and results of a research are not reproducible, when scientific misconduct is obvious or when the publication is duplicate, or plagiarism is obvious. Retractions can at certain times be partial when an honest error results in misinterpretation of a part of the article. Nevertheless, a partial retraction is more significant action than an erratum. However, when a portion of the article is retracted, the overall results of the study are usually not affected.
Although strict enforcement of these guidelines on corrections is essential, as each journal is required to implement them individually, irregular application does ensue. Ideally errors of any kind should be notified as soon as possible. Details of the resultant changes made should be clearly included and all prior versions of the article should be accessible either directly or on request. Corrections should be published in a citable form, i.e., they should be on a numbered print page for correct indexing. The title of the original article should be included in the heading of the correction published. It is best that corrections are not published as letters to editor and rather be labeled specifically as “errata” or “expression of concern” or “retraction.”
For retractions, COPE states that a retraction should involve the following.
- It should be linked to the retracted article wherever possible (i.e., in all electronic versions)
- It should clearly identify the retracted article (e.g., by including the title and authors in the retraction heading)
- It should be clearly identified as a retraction (i.e., distinct from other types of correction or comment)
- It should be published promptly to minimize harmful effects from misleading publications
- It should be freely available to all readers (i.e., not behind access barriers or available only to subscribers)
- It should state who is retracting the article, and
- It must state the reason(s) for retraction (to distinguish misconduct from honest error.
For the last statement though, it may not always be possible to distinguish misconduct from an honest error and the editor is usually the final authority to decide on this.
Corrections may be reported by the authors, editors, publishers, readers, or even the institution where the research was performed. The decision to publish a suitable form of correction is ultimately of the editor. Retractions are usually published by a mutual consent between the author and the editor. But if the author is not forthcoming and scientific misconduct is proven after an appropriate evaluation, the editor can take a decision to retract the article.
The role of “whistle blowers” in the process of biomedical research publishing is not perfect. They raise concerns regarding the possibility of a potential misconduct pertaining to a particular piece of research. They could be the readers, reviewers, or members from the authors' institution. Editors thoroughly investigate all such concerns as a journal's decision may impact adversely on the author's careers. A whistle blower is usually kept anonymous to avoid harsh consequences to them. Lay readers usually take results at their face value and usually may not raise concerns; however, this may not always be the case. Post publishing peer review process is another tool where the editor invites reviews for an article which has already been published where scientific misconduct is a concern. And needless to say, it would be more stringent.
The rate of corrections in biomedical literature is increasing in numbers. A study reported that 235 articles were retracted over the 30 years between 1966 and 1996. The same group found 1164 retracted articles between 1997 and 2009. Of the later, 54% were issued by authors, 29% by editors, 5% by publishers, and 7% by authors, publishers, and editors together. Majority of these were due to scientific misconduct that was either admitted (44%) or presumed (14%). Other reasons for retraction were irreproducible results, problems with data, methods, analysis, interpretation, and sampling or due to duplicate publication. Conflict of interest is another potential reason which may lead to corrections including retractions. In rheumatology, before 2010, there were 26 retracted articles, whereas post 2010, 134 articles have been retracted mostly due to duplication, plagiarism, copyright issues, concerns over validity of data, and lack of ethical approval or patient consent.
Gone are the days when the publication of errata or retraction notices used to build the impact factor and prove the veracity of a journal. While notification of an error highlights the intent of a journal to provide accurate information, in the present digital world with rapid information dissemination, these errors may result in loss of trust in a journal and would perhaps raise doubts regarding its credibility. One recent example of this is the retractions of the articles on therapy for COVID-19 infection by the two leading journals. To avoid this kind of unfortunate situation, authors should publish only reliable and reproducible data and it ultimately remains the author's responsibility. It is hoped that ideally, a peer review process would pick the problems up before publication. However, it may not always happen and not necessarily reflects poorly on the good intention of editorial processes.
For the authors, despite strict guidelines on how these corrections should be indexed, many corrected/retracted articles continue to be referenced time and again. The reasons for this appear to be both, the way these are indexed and the perfunctory attitude of certain authors. A potential solution for making the corrected versions hard to ignore could be to watermark any such article that has an erratum or EoC, be it the abstract or full text article and also prominently in a retracted article. In this context, it is noteworthy that >2000 post retraction citations exist for the 235 articles retracted between 1966 and 1996 with less than 8% acknowledging the retraction itself. Moreover, for the 1164 articles retracted between 1997 and 2009, 391 citing articles were found, omly 22 (6%) acknowledged the retraction, but 369 (94%) made no mention of it. It is possible that increasing the visibility of such retractions will act as a deterrent aided by increasing awareness among the authors regarding the drawbacks of citing retracted articles. The important contribution of a retraction database retractionwatch.com in this regard has to be acknowledged. Popular reference management software such as EndNote and Zotero automatically flag up retracted articles.
To conclude, though biomedical research publications should ideally be beyond reproach, corrections are at times unavoidable. Errors can set foot in at several stages – protocol preparation, data collection, statistical analysis, manuscript preparation, and publication. While the onus of publication errors lies with the publisher, it is essentially the authors who are responsible for the integrity of the data published by them. The editors try to weed out manuscripts with misconducts and the peer review process acts as a further filter. However, the editor and the peer reviewers can only explore and evaluate what has been submitted to them and hence cannot be held answerable for all types of errors or scientific misconduct that pops up later. Flaws and fallacies may still be discovered which usually are unintentional. But regardless of its nature, owning up to them and correcting them is of paramount importance in maintaining the integrity and trust of biomedical research publications. To quote Jonas Salk– “Life is an error-making and an error-correcting process.”
| References|| |
National Library of Medicine. Fact Sheet: Errata, Retractions, Partial Retractions, Corrected and Republished Articles, Duplicate Publications, Comments (Including Author Replies), Updates, Patient Summaries, and Republished (Reprinted) Articles Policy for MEDLINE. Available from: http://www.nlm.nih.gov/pubs/factsheets/errata.html
. [Last accessed on 2020 Nov 01].
COPE Council. COPE Guidelines: Retraction Guidelines; November, 2019.
Allahbadia GN. Why correcting the literature with errata and retractions is good medical practice? J Obstet Gynaecol India 2014;64:377-80.
Vie KJ. How should researchers cope with the ethical demands of discovering research misconduct? Going beyond reporting and whistleblowing. Life Sci Soc Policy 2020;16:6.
Budd JM, Severt M, Schultz TR, Scoville C. Effects of article retraction on citation and practice in medicine. Bull Med Libr Assoc 1999;87:437-43.
Misra DP, Ravindran V. Conflicts of interest in academic publishing: When in doubt, declare! J R Coll Physicians Edinb 2019;49:179-81.
Misra DP, Agarwal V. Integrity of clinical research conduct, reporting, publishing, and post-publication promotion in rheumatology. Clin Rheumatol 2020;39:1049-60.
Piller C, Travis J. Authors, elite journals under fire after major retractions. Science 2020;368:1167-8.
Misra DP, Agarwal V. Blaming the peer reviewer: Don't shoot the messenger!! Indian J Rheumatol 2020;15:162-4.