Abstract
Predatory or pseudo journals have recently come into focus due to their massive internet expansion and extensive spam email soliciting. Recent studies explored this urging problem in several biomedical disciplines. In the present study, we identified 69 potential predatory (pseudo) pathology journals that were contrasted to 89 legitimate pathology journals obtained from the major bibliographic databases. All potential predatory journals in pathology shared at least one of the features proposed by previous studies (e.g. a poor web-site integrity, submissions via email, unclear or ambiguous peer-review process, missing names of the editorial board members, missing or pending the journal ISSN). Twenty-one (30%) of the potential predatory pathology journals had misleading titles mimicking those of legitimate journals. Only one of the identified journals was listed in the Directory of Open Access journals whereas none (0%) was indexed in PubMed/MEDLINE or Web of Science, listed in the Committee on Publication Ethics nor have they had a legitimate impact factor in the Journal Citation Reports.
Introduction
Predatory or pseudo journals refer to journals that recruit articles through aggressive marketing and spam emails, promising quick, but not robust review and fast open-access (OA) publication, thus compromising scholarly publishing standards (1-4). Their key motive is a financial benefit via article processing charges (APCs) and other additional fees (1, 3, 4).
The number of OA journals has dramatically risen over the past fifteen years (5), reaching 11,376 journals, indexed in the Directory of Open Access Journals (DOAJ) in 2018 (available at https://doaj.org). This expansion was parallel to the increase in the number of predatory publishers from 18 in 2011 to more than 1100 in 2016 (6, 7). Predatory (pseudo) journals have become more prevalent than ever due to massive internet expansion and extensive spam email soliciting (2, 4, 8). These journals are thought to exert a negative impact on the science including biomedicine.
Since 2011, when Jeffrey Beall, an academic librarian from the University of Colorado Denver, posted his first list of potential predatory, OA publishers and journals (available at: https://beallslist.weebly.com), predatory journals have come into focus (3, 4).
Recent studies have highlighted the significant impact of potentially predatory journals in several biomedical fields including neuroscience/neurology, urology, emergency medicine, physical medicine, orthopedics, anesthesiology and pediatrics (9-11). These studies revealed that by October 2016 >10% of predatory journals in three biomedical fields were indexed in PubMed (12% for rehabilitation, 11.4% for neurosciences, and 20.2% for neurology). By April 2017, these values increased to 23.7% for rehabilitation, 16.1% for neuroscience, and 24.7% for neurology, indicating that these journals evolved rapidly (6, 7). The study on emergency medicine revealed that almost half (25/55) of the OA journals were predatory journals (12). Furthermore, the study on urology found that (7/32) potential predatory journals, which were solicited to an academic urologist, were indexed in reputable databases including: Journal Citation Reports (JCR), Scimago Journal Rankings (SJR), and DOAJ (13). In contrast, a recent study of Kokol et al. exploring predatory journals in pediatrics revealed 26 such journals; however, none of them were indexed in PubMed, Scopus or Web of Science (9).
The role and presence of potential predatory journals in pathology have not been explored so far. In this study, these journals have been critically evaluated and were contrasted to the legitimate pathology journals.
Materials and Methods
1. Journals Identification and Selection
Between January and May 2018, Beall’s list of predatory journals served as an initial database of suspected journals related to pathology, as it was previously used in other studies (2, 7). The term “potential predatory” was only used after assessing each journal separately based on the recommended criteria that are summarized in Table 1 (2, 8). On the other hand, the major bibliographic databases were explored (PubMed/MEDLINE, PubMed Central (PMC), and Web of Science/Science Citation Index (Science Citation Index/Science Citation Index Expanded (SCI/SCIE)) to identify legitimate journals in pathology. The journal titles were retrieved using the following key words: anatomic pathology, cellular pathology, clinical pathology, cytopathology, diagnostic pathology, experimental pathology, histopathology, human pathology, immunopathology, medical pathology, molecular pathology, molecular biomarkers, neuropathology, pathology, and surgical pathology.
2. Data Collection (phase 1)
After collecting the available number of journals related to pathology, each journal was assessed based on, Clemson and colleagues’ criteria; its indexing status, clarity of peer-review process, availability of its archive, legitimacy of editorial board, the status of International Standard Serial Number (ISSN), emphasis on OA, and country of origin (8).
2.1 Status of ISSN and country of origin
To identify the journal’s country of origin and the ISSN legitimacy of the suspected journals, the ISSN Portal service was used (available at https://portal.issn.org), which is an online database that provides information about the journal’s resource and record including the country of origin.
2.2 Amount of APCs
In order to obtain the amount of requested APCs, each journal’s web site was checked for the Creative Commons ‘Attribution’ License Non-Commercial (CC BY-NC) price. Furthermore, APCs of other currencies were converted to the US Dollar (USD) using the currency converter (available at https://xe.com).
2.3 Emphasis on OA, legitimacy of editorial board, and clarity of peer-review process
Journals that were promoting OA publishing modality on their websites’ homepage, displaying phrases similar to “Promoting OA, Supports OA, Your Way To OA”, were labeled under “emphasis on open-access”. Editorial boards were reviewed in each journal and a missing editorial board was labeled. Unclear peer-review process was labeled under “ambiguous or unclear peer-review process”.
2.4 Website integrity, number of issues, and misleading titles
For the website integrity, marking was based on presence of major design flaws, compatibility and language mistakes. “Unreal or small number of issues per year” was determined by comparing displayed number with the available number. To identify misleading journal names, a minimum of two identical words was required.
2.5 Indexing status
The indexing status displayed on the journal’s website was appraised via both the National Library of Medicine’s (NLM) catalog (available at https://ncbi.nlm.nih.gov/nlmcatalog/) and the DOAJ.
3. Data Collection (phase 2)
After obtaining preliminary data from phase 1 of the assessment, each journal was further investigated based on Shamseer and colleagues’ criteria; aims and scope, journal name and publisher, home page integrity, contact information, indexing and impact factor, editorial processing and peer review, publication ethics and policies, and publication model and copyright (2). Since DOAJ indexing status, ISSN, APCs and misleading titles were obtained in phase 1, their findings were only reviewed.
3.1 Aims and scope, home page integrity, and contact information
The aims and scope were investigated for inclusion of non-biomedical subjects. Spelling and/or grammar errors, distorted images, author-targeting homepage language, and non-journal affiliated emails, were all collected.
3.2 Editorial processing, and publication copyright, ethics and policies
Presence of editor-in-chief, editorial board, submission system, peer review, retraction, copyright and plagiarism policies, and APC amount, were labeled accordingly.
3.3 Indexing and impact factor
The status of each journal was checked in PubMed/Medline, Web of Science (SCI/SCIE), the World Association of Medical Editors (WAME) (available at http://wame.org), and the Committee on Publication Ethics (COPE) (available at https://publicationethics.org/members). Additionally, the Index Copernicus value was also considered.
4. Statistical Analysis
Shapiro–Wilk test and histogram were used to check for normality. The non-parametric, Mann-Whitney test was used to determine statistical significance of the APC means between OA legitimate and potential predatory journals. All analyses were performed using SPSS Statistics 25.0 for Windows (IBM, Armonk, NY, USA). Statistical significance was accepted at P < 0.05.
Results
Of the total current publishers on Beall’s list (1196), 69 potential predatory journals were identified in the field of pathology (“the black list” journals). The journals’ titles and their respective publishers are shown in Table 2. Only one of the identified potential predatory journals in pathology (Journal of Modern Human Pathology) was indexed in the DOAJ. None of these journals were indexed in PubMed/MEDLINE, Web of Science (SCI/SCIE) nor listed in COPE or had a legitimate impact factor in the JCR (Clarivate Analytics) (“the white list” journals). Of note, 13 potential predatory journals in pathology had at minimum one of their articles archived in the PMC repository (following the policy of PMC as a digital repository archiving free full-text articles that had been published within the biomedical and life science journals). On the other hand, 89 legitimate journals were identified in the field of pathology and used for the comparison (listed in Table 3).
All potential predatory journals in pathology shared at least one common poor-quality characteristics: lack of web-site integrity (n = 21, 31%) vs. (n = 0% in OA legitimate), missing/pending ISSN number (n = 36, 52%) vs. (n = 0% in OA legitimate), unreal or small number of issues per year (n = 22, 32%) vs. (n = 3, 3% in OA legitimate), emphasis on OA policy (n = 40, 58%) vs. (n = 32, 38% in OA legitimate), missing editorial board (n = 20, 29%) vs. (n = 1, 1% in OA legitimate), and ambiguous or unclear peer-review process (n = 38, 55%) vs. (n = 12, 14% in OA legitimate). Moreover, the majority (77%) of potential predatory journals accepted manuscript submissions via email. Absence of retraction, plagiarism, and copyright policies were all characteristics of the suspected journals as shown in Table 4 and Figure 1.
Furthermore, 21 (30%) potential predatory journals had misleading titles, which resemble or appear to be tied to those of legitimate ones (see Table 5). In addition, 31% of the potential predatory journals were indexed in the databases that generate bogus impact factors (e.g. Index Copernicus, Cosmos Impact Factor, and J-Gate). More specifically, 19% of the potential predatory journals presented their Index Copernicus value, whereas only 4% of the legitimate journals presented this impact factor.
Eighty-three percent of the potential predatory journals displayed the required APCs on their web sites. The mean APC was significantly higher among the legitimate OA pathology journals in comparison with the predatory ones (US$ 2837.6 vs. US$ 814.3; range US$ 550-4100 vs. US$ 50-2700; p < 0.001).
Out of the suspected journals with valid ISSN (33/69), the vast majority (n = 23, 70%) of the targeted journals were originated from the United States, followed by India (n = 4, 12%). The remaining journals were distributed among the United Kingdom (n = 3, 10%), Nigeria (n = 2, 6%), and the United Arab Emirates (n = 1, 3%).
Discussion
The Internet has dramatically transformed academic publishing, most notably, due to the introduction of OA publishing (2-5). Recently, there has been a rapid rise of online journals described as ‘predatory’ (3, 4). Such journals actively solicit manuscripts and charge publication fees without providing a robust peer review and proper editorial services (e.g. copyediting and proofreading) (2-4, 8, 9). Yet, it is important to note that OA is not correlated with the legitimacy of the journal.
In the present study, the impact of potential predatory journals in pathology was explored. Similar to other biomedical fields, our data indicate a substantial burden of such journals in academic pathology. One of the important findings of our study is that none of the potential predatory journals in pathology has been indexed in the major bibliographic databases such as PubMed/MEDLINE and Web of Science nor listed in COPE. These results are in line with a recent study on such journals in pediatrics (9). However, the studies in other biomedical fields (neurology/neurosurgery, physical medicine and emergency medicine) revealed the substantial contamination (up to 25%) with such journals in the major bibliographic databases (7, 12, 13).
Despite having none of the potential predatory journals indexed in PubMed, it is not a predatory-free database (14, 15). PubMed indexing policies are less strict compared with MEDLINE (14). In fact, 98% of the legitimate journals in this study were indexed in PubMed, whereas 81% were in MEDLINE. This might explain how some predatory journals managed to leak into the PubMed database (14). Although MEDLINE and PubMed are distinctively different databases, using the PubMed search engine queries both databases simultaneously (as well as PubMed Central).
Our study indicates that a substantial proportion (30%) of potential predatory journals in pathology may have similar names to the legitimate and renowned pathology journals (Table 5). In addition, predatory publishers usually send spam emails through which they invite authors to contribute to their journals and conferences promising a fast-track review and publishing (8). Summarized in Table 4 are the key characteristics of predatory journals/publishers (8), which may also guide the pathology researchers/scientists before deciding to submit an article to not well-known/unknown pathology journals (16).
Another strikingly distinct feature of potential predatory journals is APC (17). In line with previous studies in other medical disciplines, this study confirms that the mean of APC of potential predatory journals in pathology was significantly lower than that of legitimate OA journals in pathology (∼US$2837 vs. ∼US$814) (17). This marked difference along with “easy to publish” spam announcements might be a reason to attract some inexperienced and/or young academic pathologists to submit a paper to such journals.
Although the US and India were the countries of origin for the majority of the potential predatory journals, the methodology used in this paper provides better precision regarding the journal’s country of origin. In previous reports, Google Maps and 3D Street View were used predominantly to determine the journals and/or publishers’ country of origin based on the addresses displayed in their websites (7, 17). Such methodology is not credible to assess the country of origin, simply, because any random address can be presented in their web sites.
Notwithstanding its implications, two limitations to this study are important to highlight. First, the use of Beall’s list has been controversial as it was discontinued in 2017 and is consequently considered outdated (18, 19). Nevertheless, that list served as a start point given that each journal in our study was objectively apprised following the recommended criteria proposed by other researchers (8). Second, the Beall’s list was criticized for being subjective as it was established and updated by a single person; however, several well-conducted studies have relied on the list as an accessible reference for predatory journals (2, 6, 7, 17, 18). Finally, it is worth mentioning that the removal of Beall’s list led to the rise of new alternatives such as Cabell’s Blacklist (20).
In conclusion, this study highlights a substantial burden of potential predatory journals in pathology. A significant proportion of such journals (30%) have name of resemblance to the legitimate pathology journals, which may pose another significant challenge and threat to the academic community within this medical discipline. This study may aid pathology researchers in their decision-making process when submitting manuscripts for publication. Based on the obtained data, authors should check the journal’s status on PubMed/MEDLINE, Web of Science and DOAJ, as well as the previously proposed criteria, confirmed in this study, before submitting a manuscript to a pathology journal.
Conflict of interest
Semir Vranic serves as an editor-in-chief of the Bosnian Journal of Basic Medical Sciences and a consulting editor for Breast Cancer: Targets and Therapy. Saghir Akhtar is an editor-in-chief of the Journal of Drug Targeting. Faruk Skenderi is a managing editor of the Bosnian Journal of Basic Medical Sciences. Other authors declare no competing interests.
Acknowledgment
This study was supported by the student grant number (#QUST-1-CMED-2018-10) provided by the College of Medicine, Qatar University. The preliminary data from the study were presented at the XXXII Congress of the International Academy of Pathology (IAP), October 2018, Amman, Jordan.