Readership Awareness Series – Paper 9: Retraction of a Publication

INTRODUCTION

Correcting the literature and ensuring its accuracy and reliability are essential for maintaining transparency and public trust in science. Several modalities are available to carry out correction in the literature, including errata, corrigenda, expression of concern (EoC), and retraction.^1,2 Retraction is the act of removing a fraudulent published paper from the literature to correct the literature and maintain the integrity of scientific literature.^1–7 The purpose of this retraction is to alert the readership of serious flaws in data validity, interpretations, or conclusion of a specific research. Van Noorden called retractions ‘science’s ultimate post-publication punishment’.³ While the number of publications is steadily increasing year on year, so is the number of retractions. A study of biomedical literature has demonstrated that the mean number of publications over the last decade was 828,917 per year and a total number of retractions in the last 10 years were over 4000 and this could only be the tip of the iceberg.⁴ A recent analysis of 18 years found 1182 retracted papers originating from paper mills.⁶ The retraction of PubMed indexed articles increased from 0.002% to 0.02% from the 1980s to 2009.⁷ This increase in retractions can be secondary to several factors including increasing awareness, software to detect plagiarism and image manipulations, and wider reach of papers to the global audience via the internet.

INDICATIONS FOR RETRACTION

In their guidelines, the Committee on Publication Ethics (COPE) has advised that editors can consider issuing a retraction if.²

1. Unreliable data

2. Fabricated or falsified data

3. Redundant publications

4. Plagiarism

5. Peer-review manipulation

6. Copyright infringement

7. Non-disclosure of competing interests or conflicts of interest that influenced validity of the research.

8. Unethical research

9. Usage of material without appropriate authorization

10. Legal issues (privacy/illegal)

Retractions are not advisable if the validity of the findings is beyond doubt or still reliable and can be corrected or if there is inconclusive evidence.² Those conflicts of interest that were reported following publication but did not appear to influence the study outcomes may also be subjected to correction and not retraction.²

The top reasons for retraction in an analysis of cardiovascular literature were duplication (28.1%), plagiarism (19.2%), and falsification (16.8%).⁴ Similar was the analysis of retractions in the surgical literature where duplication (35.3%), Institutional review board violations (18.5%), falsified data (14.7%), and data errors (9.8%) were the most common indications for retraction.⁵

PROCESS OF ISSUING A RETRACTION

Several events can trigger the process of issuing a retraction notice. An investigation of scientific misconduct by a university or the author’s institution or complaints raised by authors or other author groups, or independent investigators are common initiators. The concerns of scientific misconduct are reported to the editor or the publisher officially with evidence to support the claims. There are no foolproof standardized protocols, but generally, the editor or the publisher then assesses the merit of the concern based on the evidence supplied. However, if there is any evidence of misconduct, the editor may seek clarification from the authors or the author’s institute. The author’s institute or university may be notified by the editor, and they may initiate an investigation for significant concerns. Pending the investigation outcome, the editor may publish an EoC to alert the readers on the possible and questionable validity of the results of the specific research work. If the findings are conclusive of a scientific misconduct that warrants a retraction, then the editor or the publisher or both can issue a retraction notice.

COMPONENTS OF A RETRACTION NOTICE

Although the language and proforma may vary, a retraction notice ideally begins with a clear and boldly written ‘RETRACTION’ or ‘RETRACTION NOTICE’, followed by the title and authorship of the research paper in question.^1,2 This is followed by how the editors or publishers were alerted, the concerns raised and who is retracting the article. The reasons for retraction should be factual and objective without any derogatory statements. Clear mention of the investigation that confirmed scientific misconduct and the basis of retraction is critical. The ICMJE has clearly expressed its recommendation as ‘Expressions of concern and retractions should not simply be a letter to the editor. Rather, they should be prominently labelled, appear on an electronic or numbered print page that is included in an electronic or a print Table of Contents to ensure proper indexing, and include in their heading the title of the original article’.¹ The report should be promptly published with a link to the retracted article and should be freely available to all.²

LIMITATIONS OF THE RETRACTION PROCESS

There are inconsistencies in policies and practices across journals. There needs to be more clarity among editors on their capacity, legal position, and implications while issuing a retraction. The lack of desirable author engagement and the prolonged investigation period (for example, a mean of 1.4 years in cardiovascular literature and 3.6 years in surgical literature)^4,5 can lead to anxiety and may harm the reputation of the researchers. Opaque retraction notices (where the reason was not mentioned) may lead to continuous citation of the retracted papers and harm the work and reputation of several other researchers. 8.7% of retractions in surgical literature were opaque notices.⁵ Retraction notices garner less coverage, and this compounded by their improper indexing and paywall restrictions increases the problem of continuous citations even after the paper has been retracted. The median number of citations of retracted papers in cardiovascular literature was eight.⁴ The stigma associated with retraction would prevent honest researchers from coming forward and admitting the mistakes and correcting the wrongdoing.

FUTURE DIRECTIONS

The screening, peer review, editorial handling, and processing are constantly evolving and need to be strengthened.⁸ For example, ‘CrossMark’ logo is placed by publishers on the PDF and clicking it would give the readers (with internet connection) any post-publication updates like retraction or corrections for that specific paper. There is a need to tackle the menace of paper mills which of late has emerged to the forefront of scientific misconduct.⁹ Awareness about retraction databases (e.g., retractionwatch) and how to use them would help prospective authors. To overcome the prolonged and frustrating wait till the investigation is complete, there have been proposals for splitting the process into two.¹⁰ The first phase only determines whether the data is reliable, or the raised concern is valid or not without investigating who is to blame. The retraction could then be published. Subsequently, the second phase can investigate the underlying wrongdoing. While the retraction guidelines will continue to evolve as our understanding evolves, there is a need to minimize the inconsistencies in policies and practices across the journals. Promoting the awareness from the medical school level will play a pivotal role.

DISCLOSURE STATEMENT

Mohammad Javed Ali receives royalties from Springer for the textbook “Principles and Practice of Lacrimal Surgery’ and ‘Atlas of Lacrimal Drainage Disorders’, and the ‘Video Atlas of Lacrimal Surgery’

Additional information

Funding

Hyderabad Eye Research Foundation