11 December 2022
On November 1-2, 2022, PubMed listed over 190,000 entries for the search term “Alzheimer’s disease” (AD), with over 13,600 entries for 2022 alone [1], suggesting that this topic is highly researched by academics within the biomedical research community. In this letter, the topic of ‘tortured phrases’ [2], as applies to the term “Alzheimer’s disease”, is explored. Ultimately, the question is asked, if other terms are used to describe AD, would it still be recognized as such?
The term ‘tortured phrases’ refers to a new phenomenon in which the lack of linguistic specificity results in the creation of novel terms that replace standardized jargon. Such terms may arise from the need by authors to reduce the level of textual similarity when submitting their papers to a journal, for example by using online thesauruses or reverse translation techniques, to avoid claims of plagiarism [2], clumsy replacement of terms with synonyms, such as “illness” rather than “disease”, lack of fluency or familiarity with English, or other as-yet-unknown reasons. While it is likely that an AD specialist would appreciate that terms such as “Alzheimer’s malady” (AM), “Alzheimer’s ailment” (AA), “Alzheimer’s infection” (AI), or “Alzheimer’s sickness” (AS) would all likely be referring to AD, the existence of this range of unspecific terms to describe an established term, “Alzheimer’s disease”, may sow confusion among junior members of the medical community, such as students or interns. The greater risk is that they themselves might employ the alternatives—i.e., the ‘tortured phrases’—rather than the established term AD itself.
Ultimately, the existence of such non-existent terms may dilute the specificity of medical jargon, and thus potentially affect the integrity of the medical literature [3]. By virtue of inter-paper citation among ranked and unranked journals, cross-contamination of such ‘tortured phrases’ may occur, even entering the literature of highly ranked journals that supposedly carry out peer review and with strict editorial screening that would surely have detected such false positives of AD.
To better appreciate the existence of the above five listed ‘tortured phrases’ for AD, the Tortured Phrases Detector of the Problematic Paper Screener (PPS) [4] was consulted on November 1-2, 2022. At PPS, a total of 22, 37, 43, and 56 entries were discovered for AM, AA, AI, and AS, respectively. These were categorized along four dimensions: 1) open access (OA) (DOIs highlighted in green); 2) document category; 3) publisher; and 4) number of citations (see the Supplementary Material (Excel file) for raw data).
A more detailed assessment of those findings revealed the following, noting that a single document might contain one or more of the four ‘tortured phrases’): 1) The relative percentage of documents that were OA (DOIs highlighted in green in the Supplementary Material) for AM, AA, AI, and AS were 41, 22, 30, and 27%, respectively, allowing for the free access and confirmation by any member of the public; 2) the number of these ‘tortured phrases’ that were found in the following four document categories (articles; chapters; preprints; proceedings) were: AM (11; 9; 1; 1), AA (13; 8; 1; 5), AI (20; 14; 1, 8), and AS (24; 22; 2; 8), indicating that they were most common in articles and book chapters; 3) the top three publishers based on counts of the four ‘tortured phrases’ in any document were: AM (7 Springer Nature, 6 Elsevier, 2 Taylor & Francis), AA (14 Springer Nature, 6 Elsevier, 5 IEEE), AI (12 Springer Nature, 8 Elsevier, 8 IEEE), and AS (17 Elsevier, 15 Springer Nature, 9 IEEE); and 4) the four ‘tortured phrases’ in these documents were cited a total of 165, 135, 119, and 413 times for AM, AA, AI, and AS, respectively. The paper that accrued the most citations among all four ‘tortured phrases’ (DOI: 10.1109/ACCESS.2020.3007939) was retracted, and was one of three retracted documents, one of which (DOI: 10.4172/2167-7182.1000247) was “silently” retracted [5].
Absent a public explanation by the authors, it is impossible to know the origin of such ‘tortured phrases’. In journals that claim to use ethical guidelines, such as those established by the ICMJE or COPE, editors need to be cautious when resolving cases associated with ‘tortured phrases’ because their existence might not be an ethical issue, merely the lack of attention to detail (by authors), or the lack of attention to quality control (by peer reviewers, editors, and copy editors). There is an ethical component, however, if the authors have used a third-party service (such as a paper mill or translation service), which may have introduced such terms into the paper, but failed to declare their reliance on such a service in their acknowledgement [6].
There are currently still only about 150 documents, mainly academic papers, that include one or more of these four ‘tortured phrases’, 13 of which are indexed in PubMed. These instances are not merely a linguistic anomaly or curiosity. There may be, as the literature that is discovered to have such non-standard terms expands, far-reaching consequences on the integrity, sensu lato, of the biomedical literature, so additional debate and solutions are needed. One suggestion is to introduce a double DOI-based “publication history” for papers that would allow authors (and editors) to constantly or real-time upgrade or correct papers, whenever oddities or errors are detected [7].
Limitations
This assessment did not—because not all full texts were available (not OA) or accessible—make judgements based on an examination of original texts and relied exclusively on the accuracy of PPS. Only AD-related ‘tortured phrases’ were examined although several of the papers carried multiple non-AD-related ‘tortured phrases’. Even though several of the documents had more than one AD-related ‘tortured phrase’, citations for each (i.e., AM, AA, AI, and AS) were considered separately. One way to refine academics’ searches for AD in PubMed to allow for a more precise output would be to use the MeSH terms function [8].
Jaime A. Teixeira da Silva
Independent researcher, Ikenobe 3011-2, Kagawa-ken, 761-0799, Japan. E-mail: jaimetex[at]yahoo.com
REFERENCES
[1] PubMed (2022) Alzheimer’s disease. https://pubmed.ncbi.nlm.nih.gov/?term=Alzheimer’s+disease
[2] Cabanac G, Labbé C, Magazinov A (2021) Tortured phrases: A dubious writing style emerging in science. Evidence of critical issues affecting established journals. arXiv (preprint, not peer reviewed) 27 pages. https://doi.org/arXiv:2107.06751v1
[3] Teixeira da Silva JA (2022) Tortured phrases dilute the specificity of medical jargon. J Health Social Sci 7, 137-140.
[4] Problematic Paper Screener (2022) Tortured Phrases Detector. https://dbrech.irit.fr/pls/apex/f?p=9999:24:::NO:::
[5] Teixeira da Silva JA (2016) Silent or stealth retractions, the dangerous voices of the unknown, deleted literature. Publishing Res Q 32, 44-53.
[6] Teixeira da Silva JA (2021) Outsourced English revision, editing, publication consultation and integrity services should be acknowledged in an academic paper. J Nanoparticle Res 23, 81.
[7] Teixeira da Silva JA, Nazarovets S (2022) Publication history: A double DOI-based method to store and/or monitor information about published and corrected academic literature. J Scholarly Publishing 53, 85-108.
[8] PubMed (2022) Alzheimer’s disease[MeSH Terms]. https://pubmed.ncbi.nlm.nih.gov/?term=alzheimer+disease%5BMeSH+Terms%5D&...
