As the Editor-in-Chief of the Journal of Korean Medical Science (JKMS), I have encountered 3 didactic cases of plagiarism since 2012. The first one was copying figures from a JKMS article in a good journal without citation or proper acknowledgment; the second—submission of a slightly modified manuscript, recently published in another journal, to JKMS; and the third—publishing a copied JKMS article in another journal. To investigate these cases, I followed the guidance from the Committee on Publication Ethics (COPE) (2). The authors in all these cases accepted my criticism and apologized for their misconduct.

The first case is rather simple. The original article in JKMS described normal variation of the brachial artery branches in Korean adults and proposed 3 types of variation, which were presented in a schematic drawing in 2008 (3). The author of the JKMS article informed the editorial office that his drawing was copied in an article, which was published in Seminars in Dialysis in 2011. The corresponding author of the plagiarizing article, an American surgeon, confessed that his fellow recycled previously published figures without permission and proper acknowledgment. Such a copying went unnoticed before the manuscript submission and during the review process. The editor of the Seminars in Dialysis and senior staff members of the publisher (Wiley) also apologized for the misappropriation of graphics. But the editor explained that it was a minor misconduct without serious consequences for the literature. The journal neither retracted nor published an official apology for the oversight, leaving a room for similar misconduct in the future. Recently the author of the plagiarized article appealed the editor and publisher again for their neglecting attitude and then the editor replied to consider an official decision for this copyright violation.

The second case was due to the submission of a plagiarized manuscript to JKMS by Chinese authors in 2015. The primary original article was about suppression of microRNA in gastric cancer metastasis, which was published by other Chinese authors in Tumor Biology in 2013 (4). The secondary manuscript was a conspicuous copy with a few minor changes. The case of plagiarism was surfaced during the pre-publication screening by the editorial staff of JKMS. The plagiarizing authors reported that they did not know it was a copy of a previous publication. According to their report, they referred to a Chinese editing company with a request to prepare an original article. And the company prepared the unethical copy. The cover letter for JKMS documented a previous submission to the Asian Pacific Journal of Cancer Prevention. It suggested that the manuscript was submitted to multiple journals. Remarkably, the same article, authored by another group of Chinese authors, was already retracted from the International Journal of Clinical and Experimental Medicine in 2015 (5). The submission to JKMS was rejected, and all the authors were banned from publishing in JKMS for 5 years. The administration of the plagiarizing authors was officially informed about unethical behavior of their employees, but no any response or sanction was followed. This case reveals a new type of misconduct (‘predatory’ editing and brokering services), which is a threat to the evidence accumulation on an unimaginable scale.

Finally, the third case was due to the publication of an article in the International Journal of Clinical and Experimental Medicine in 2015, copying an experimental radiological report published in JKMS a year before (6). The authors from China published almost identical copy with the same text, tables, and figures from the JKMS article. The authors were notified about their misconduct and violation of the JKMS copyrights. They apologized for the misconduct and promised to initiate the retraction process. The editor of the journal and the authors' administration were also notified and presented with PDFs of the primary and secondary publications, but there has been no any response so far. The secondary article is still not retracted.

Surprisingly, instances of text and graphics copying are still revealed despite the plagiarists' awareness of what constitutes unlawful behavior. The pressure to publish may be partly responsible for such a tendency (7). It is also possible that blatant plagiarism remains a big issue because of plagiarists' disrespect toward the scientific community and lack of preventive strategies in their research institutions. Not all research administrators take plagiarism and related retractions seriously, allowing plagiarists to continue working without any sanctions. Impunity creates a vicious circle, compromising research environments.

Journal editors should be aware of predatory editing services, which are offered to non-Anglophone authors. Unethical commercial agencies generate ‘clones’ of primary publications and involve authors in their corrupt businesses. The authors, who refer to the predatory agencies, commit double misconduct by purchasing the clones and submitting them to journals. The authors themselves become ‘predators’ by wasting their money. And the innocent journal editors' duty is to scrutinize submissions that may be products of the corrupt author-agency interactions.

Journal editors should be alerted that even the most powerful text similarity software is incapable of detecting misconduct in some cases. In the above described second case of plagiarism, the initial Crosscheck similarity check, powered by iThenticate^® (Oakland, CA, USA), reported no textual overlap in June 2015. The same software recorded 96% similarity rate when a control check was performed a year later in July 2016. Such a discrepancy is confusing for the software users. But they should know that the similarity check compares accessible parts of published articles within a copyright period of one year. Such a short timeframe allows to check predominantly freely accessible abstracts.

In the third case of plagiarism, the primary source was JKMS, an immediate OA journal, but the similarity check reported 3% overlap only in Table 1. The Crosscheck most probably screened only abstract and/or tables, even though the source is an OA periodical. Strategies aimed at upgrading anti-plagiarism software and providing immediate OA may improve the situation.

It is possible that predatory editing agencies are aware of the deficiencies of the current anti-plagiarism software, and cheat the system. The similarity screening by the Crosscheck needs to be upgraded by securing more access to subscription journals. Editors, who rely on the similarity tests, should interpret the reports carefully. The similarity rates alone are not sufficient for detecting and avoiding plagiarism (8). In some cases, it is advisable to obtain disclaimers from the authors about external editing services and recycling of texts and graphics from other sources. Experts in plagiarism may suggest additional tests, barriers, and educational strategies to avoid misconduct at the pre-publication stage (9).

To sum up, plagiarism still threatens the trustworthiness of the publishing enterprise, and it is likely that journal editors will continue encountering its cumbersome cases in the future. Increased vigilance over the corrupt editing and brokering services in non-Anglophone and poor research environments is required. More efforts should be made to improve the authors' ethical writing skills and facilitate timely retractions of plagiarized items.