Fake data, paper mills, and their authors: The <i>International Journal of Cancer</i> reacts to this threat to scientific integrity

Abstract

The International Journal of Cancer (IJC) has long been a pioneer in the field of data integrity, and continues to prioritize strict pre-publication quality control measures in order to ensure that only reliable and reproducible data are published. One such measure is our cell line authentication policy. Introduced in 2010, the IJC requires the proper authentication of all human cell lines used in a manuscript, and these documents are extensively verified before the peer-review process even begins.1 This strict procedure has confirmed that many of the papers submitted to the IJC include data based on false cell lines. From our experience within the last years, about 4% of the manuscripts that were selected for peer review and included human cell line data were rejected for severe cell line related issues. While some other journals have since also introduced cell line policies, these remain much less stringent and many still allow authors the option to declare that their cell lines were not authenticated. Furthermore, since 2017, we thoroughly check all papers for image integrity issues before acceptance. In 2019, we started checking the figures already at the first revision stage, and additionally began requiring authors to provide raw data for their images (for Western Blots, but also others, if necessary). While the majority of identified issues could be corrected, about 3% to 5% of all papers with relevant images had to be rejected or were withdrawn due to serious image integrity issues within the last years. Nowadays, the scientific community is facing a new threat. While there had already been reports several years ago about the fabrication of manuscripts,2 at the beginning of 2020 several investigations showed the existence of so-called “paper mills”.3, 4 These paper mills seem to act as commercially oriented “shadow companies” that produce and sell highly similar papers, mainly for medical researchers who are under high pressure to publish. They operate by recycling text and image templates in a high-throughput setting with falsified or fabricated data and “stock images,” resulting in manuscripts that share several common features. In contrast to individual cheating within one paper, this represents large-scale systematic fabrication of manuscripts that is sometimes difficult to spot, as it only becomes evident when comparing papers across different publications and from other journals. Also here at the IJC, we have realized that image manipulation and scientific misconduct are taking on an ever-greater dimension than we could have imagined. Over the last few years, we too have been receiving increasing numbers of such manuscripts. These manuscripts report very common or irrational, and unsophisticated study approaches that lack a clear hypothesis and all present their data in a similar way. Exemplarily, during the first 3 months of 2021, these suspicious and poor quality papers constituted an alarming 5% to 10% of the total submissions to our journal. These manuscripts follow the established patterns (see Box 1) and we are able to match some of them to paper mills originally described by other image detectives. Each of these features* on its own might not be alarming, but appearing together could point to a suspicious pattern. * Modified and extended from references 3-5 The declared aim of the IJC is to publish reliable and trustable data, to help the scientific community to advance in knowledge, and not to build upon invalid results. Therefore, to protect our journal against the publication of fake data, we have implemented and reinforced the following measures that are handled very strictly for manuscripts that carry any suspicion of originating from paper mills. In addition to scrutinizing each paper at submission for the hallmarks of a paper mill production and checking for plagiarism, we require institutional e-mail addresses for all corresponding authors, original source data, valid cell line authentication documents, and good English quality. Equally important for data integrity are a solid data sharing policy, and a quality control assessment for next-generation sequencing data, two important policies that we have recently established and enforced.6 We are aware that initiatives taken at the level of a single journal will not completely stop such papers from eventually being published. As journal policies evolve, these mills could also evolve and operate in other ways and target different fields of research. However, at our journal, all manuscripts first undergo a thorough scrutiny by scientifically reputed editors, and subsequently by expert reviewers to ensure that publications attain the highest scientific standards and reliability. This has already led to the editorial identification and rejection of multiple manuscripts with paper mill features, which are generally below the scientific level of the IJC. Indeed, such manuscripts have low scientific value, poor rationale, unjustified working hypothesis, a poor level of controls (see, e.g., reference 7), and exploit unsophisticated or clearly inappropriate experimental procedures. Therefore, we wish to emphasize to potential customers and the vendors of such paper mill products the very high risk of losing money and reputation if they continue to submit manuscripts to our journal. There is a general agreement that paper mills and the authors purchasing these products pose a great challenge for the integrity of scientific work in all areas of research, and that journals must help to combat this threat by applying stricter control procedures.5, 8-10 Falsified data not only lead to the pursuit of wrong hypotheses and subsequent waste of scientific resources but also eventually will misguide target identification and therapy development, resulting in a significant economic damage and threat to health. Moreover, these unethical practices may lead to a general loss of confidence in scientific research and also discredit and stigmatize specific countries or communities. It is in our common interest that science adheres to principles of integrity. This integrity must be maintained by vigorous and concerted efforts from scientific journals, publishers, research institutions, governments, and should involve the Committee of Publication Ethics (COPE), as a leader in supporting journals and publishers on publication ethics. The International Journal of Cancer will continue to play an active role in this fight. We will take all necessary actions to protect our journal from current and future paper mill activities, as well as to weed out previous paper mill publications in order to maintain the highest scientific standards; we owe this to our authors and to the international scientific community. We are grateful to the IJC editors for their helpful suggestions, and their continuous involvement in keeping data integrity at the journal.

Ähnliche Arbeiten