Newsletter 2024
You are what you read
A non-academic guide to journal safety
The inflation of publications is a sign of the times in medicine like elsewhere. Like in other fields, medicine and sciences are facing a rapid growth of globally shared information, which some call an “infodemic.” This has become a virtual tsunami of data and advice, mixing trustworthy messages with misinformation. The consequences of such an uncontrolled phenomenon are confusion among the public about scientific data and the scientific voice.
Trustworthy scientific journals have respected publication procedures, such as independent peer review of submissions, to preserve the faithfulness of the scientific process. Still, they have come under increasing attack in recent years. There is no doubt that this evolution results from competition among publishers with the sole aim of profit. Also, corrupt conferences are another tool for the diffusion of flawed information and are more and more frequent. For this purpose, the organisers employ aggressive marketing and luxury locations. However, these events offer poor-quality content and may serve special interest groups, promoting their ideas under the guise of publishing scientific papers. Additionally, the system exploits the open-access model in which authors and researchers bear the costs.
A few years ago, very relevant publishers, such as Nature1, aimed to establish a definition for predatory journals using the Delphi method 2. Predatory journals were differentiated from legitimate journals by deceptive information, lack of membership in the Committee on Publication Ethics (COPE, https://publicationethics.org/), and characteristics such as persuasive emails, fake impact factors, low transparency, no retraction policy, inaccessible publisher details, and absence of editors or editors without affiliations.
What are the reasons why some researchers choose low-quality journals? Probably because our criteria for scientific production for academic and hospital careers encourage high publication rates in a publish-or-perish culture.
However, grading scientific journals, especially reliable ones, was not originally intended to mark a researcher’s proficiency and thus determine their academic career.
When the foundation of medical science is a scientific publication, the race to publish may have side effects on the quality of science. Researchers can simply be tricked. However, this often reveals a lack of mentorship for padding the curriculum vitae for career advancement and resisting the pressure to publish for promotion. In some cases, authors experience frustration related to repeated rejections of their manuscripts by standard journals. In a French study, participants were asked to declare their motivation for publishing3. A large portion of them were motivated by their reputation, their career advancement, the pressure from hierarchy, or financial interests on par with scientific or academic motivations.
Which strategies could limit submissions to predatory journals? The most dubious journals could be avoided by improving authors’ knowledge to reduce submissions to predatory journals, developing digital technologies to get information about journal publication practices, and checking if the journal is a member of the Committee on Publication Ethics (COPE) or listed in the Directory of Open Access Journals (DOAJ). Authors can individually check the targeted journal and test it using websites that follow the Think-Check-Submit process, ensuring it is a trusted journal. On the other hand, institutions could base their assessments on scientific content rather than publication metrics. Some have suggested giving negative points for dubious articles or scientific approaches. In France, the Conference of Deans of Medicine and the National Council of Universities for Health Disciplines published a list of preferred scientific journals for publishing research selected by a group of experts.
Academic institutions could quickly evaluate results made public if scientists could save their studies and experimental data in publicly accessible “repositories” before they are sent to a journal.
Another very delicate point is increasing the quality of peer review. Most journals apply a system of anonymous review. Still, this process could be more problematic since the reviewers, who generally work for free and are moved by a passion for science, may be influenced, may have a hidden conflict of interest, or maybe too busy to perform a good evaluation. Some journals name their reviewers, and the reviews themselves are available for all readers.
From an ethical point of view, the wide dissemination of fake information and fraudulent scientific papers risks eroding public trust in science. It may also impact therapeutic strategies, which in turn significantly affect patient outcomes.
More ethical concerns are arising from using ChatGPT or similar platforms in scientific publishing.
In November 2022, OpenAI released ChatGPT, and immediately after, people started using it to write their scientific papers. Rahimi et al. 4 report an incredible rate of 90% of manuscripts being somehow influenced by ChatGPT, and this is understandable since ChatGPT can generate acceptable written responses and, nonetheless, abstracts and introductions of papers. Yet, artificial intelligence (AI) generated text may include flawed information, while plagiarism will increase exponentially along with its large adoption.
AI tools based on large language models do not qualify as authors for the COPE, which requires that authors take full accountability for the integrity and accuracy of the published content. Any AI implementation cannot satisfy these obligations.
Similarly, AI has recently been applied for peer review instead of voluntary, unpaid human reviewers 5. Even if, with the right balance between automation and individual expertise, AI-enhanced peer review could become an acceptable procedure, review experts are obliged to maintain secrecy about their evaluation. Consequently, employing AI to substitute human peer review violates the elementary requirement for confidentiality.
Completing a scientific publication is serious, hard work: it takes serious and hard work, requiring creativity and originality, fact-checking, and testing.
Addressing ethics is crucial in the rapidly advancing invasion of the scientific publishing field by AI applications. Anaesthesiologists should be aware of AI biases and held accountable for the honourability of their publications.
Authors
- Ornella Piazza – Anesthesia and Intensive Care, Department of Medicine, Surgery and Dentistry, University of Salerno, Baronissi,(SA), Italy
- Corresponding author: Anne-Claire Lukaszewicz – Anesthesia and Critical Care Medicine Department, Edouard Herriot Hospital; Hospices Civils de Lyon; Lyon, 69437; France
- EA 7426 « Pathophysiology of Injury-Induced Immunosuppression », Joint Research Unit HCL-bioMérieux; Université Claude Bernard Lyon 1, Hospices Civils de Lyon, bioMérieux; Lyon, 69003; France
References
- Grudniewicz A, Moher D, Cobey KD, et al. Predatory journals: no definition, no defence. Nature 2019; 576:210–212.
- Cukier S, Helal L, Rice DB, et al. Checklists to detect potential predatory biomedical journals: a systematic review. BMC Medicine 2020; 18:104.
- Duracinsky M, Lalanne C, Rous L, et al. Barriers to publishing in biomedical journals perceived by a sample of French researchers: results of the DIAzePAM study. BMC medical research methodology 2017; 17:96.
- Rahimi F, Talebi Bezmin Abadi A. ChatGPT and Publication Ethics. Archives of Medical Research 2023; 54:272–274.
- Donker T. The dangers of using large language models for peer review. The Lancet Infectious Diseases; 2023; 23:781