Duplicate publication of articles used in meta-analysis in Korea
© Choi et al.; licensee Springer. 2014
Received: 24 June 2013
Accepted: 2 October 2013
Published: 9 April 2014
With the increasing use of meta-analysis, duplicate publication of original research is particularly problematic. Duplicate publication can result in an inappropriate weighting of the study results. The purpose of our study was to assess the incidence and characteristics of duplicate publications in Korea, and to estimate the impact of duplicate publication on meta-analyses. The meta-analysis literature written by Korean authors was searched using the online search engines PubMed, KMbase, and KoreaMed. Duplication patterns were classified into the following 4 combinations: identical samples and identical outcomes (copy), identical samples and different outcomes (fragmentation), increased samples and identical outcomes (imalas), and decreased samples and identical outcomes (disaggregation). To estimate the multiple publication bias, we performed a meta-analysis with and without duplicated data. We estimated that 6 (6.9%) of the 86 analyzed meta-analyses included duplicate publications, and 6 of the 1,194 articles (0.5%) used in the meta-analyses were duplicate publications. In this study, duplicate publications were usually due to disaggregation and overlapping (imalas) publications. Of 6 duplicated articles, 1 was considered a copy (16.6%); 1, a fragmentation (16.6%); 2, imalas (33.3%); and 2, disaggregations (33.3%). There was an increase in the mean effect size and fail-safe number with duplicated data. Our study found only 6 instances of duplicate publication after analyzing 1,194 articles used in meta-analyses written by Korean authors. However, 6.9% of the meta-analyses included duplicate publications. Our findings suggest that meta-analyses should be interpreted cautiously, taking into account the possibility of duplicated studies.
KeywordsDuplicate publication Meta-analysis Korea Bias
Over the past few decades, there has been an explosive increase in biomedical publications, and the practice of duplicate publication is sometimes problematic in the scientific medical community. According to the International Committee of Medical Journal Editors, duplicate (or redundant) publication can be defined as “the publication of an article that overlaps substantially with one already published in print or electronic media” (2013). Duplicate publication is unethical because it is wasteful of the time, effort, and resources of journals, editors, peer reviewers, readers, libraries, and electronic databases; it also delays the publication time for papers from other researchers (Leopold 2013). This practice calls into question the integrity of science and even distorts the academic reward system. Furthermore, multiple publications of the same work can artificially exaggerate the significance of a particular set of findings or ideas, compromising efforts to conduct effective systematic reviews and meta-analyses.
The incidence of duplicate publications has been measured, and varied incidences of duplicate publication have been demonstrated in previous studies. A total of 7.6% of publications in the orthopedic literature were found to have some degree of redundancy (Gwilym et al. 2004), 1.8–8.5% of articles in the otolaryngology literature (Rosenthal et al. 2003; Bailey 2002), 14% in surgical journals (Schein and Paladugu 2001), < 1% of the plastic surgery literature (Durani 2006), 8.3% of anesthesia and analgesia journals (von Elm et al. 2004), 1.39% of ophthalmology publications, and 2% of hand surgery publications have been identified as duplicate articles (Mojon-Azzi et al. 2004; Chennagiri et al. 2004). In Korea, Kim et al. (Kim et al. 2008) reported that 27 (5.93%) of 455 articles were duplicated.
Meta-analysis is now a commonly used technique for summarizing published data. With the increasing use of meta-analysis, duplicate publication of original research is particularly problematic. Because duplicate publication can result in an inappropriate weighting of the study results, it may result in multiple publication bias (Huston and Moher 1996; Wood 2007; Murphy and Wyllie 2009; Johnson 2006; Tramer et al. 1997). The data from the same patient will be analyzed more than once, leading to biased estimates of treatment efficacy, exaggerated accuracy, and a false impression of drug safety. Thus, both researchers and readers must be alert to the possibility of data overlap due to duplicate publication.
To date, there has been no study of multiple publication bias in meta-analyses. Therefore, the purpose of our study was to assess the incidence and characteristics of duplicate publications in Korea, and to estimate the impact of duplicate publication on meta-analyses.
Systematic search and study selection
The meta-analysis literature written by Korean authors was searched using the online search engines PubMed, KMbase, and KoreaMed. PubMed (http://www.ncbi.nlm.nih.gov) was screened using the key word “Korea” and limited with “meta-analysis” in advanced search for “type of article”. To find meta-analyses not indexed in PubMed, the KMbase (http://kmbase.medric.or.kr) and KoreaMed (http://www.koreamed.org) were searched using “meta-analysis” as a search term. The entire search was performed by 1 librarian, and analysis of selected meta-analyses and included studies was performed independently by 4 investigators.
All investigators screened the titles and abstracts of articles included in the 98 meta-analyses. If potential redundancies, such as the same authors, sample, methodology, results, or conclusions, were identified after screening titles and abstracts, the full article of the suspected duplicate was analyzed, and the contents, methods, subjects, and results were compared separately by 4 reviewers. If any discrepancy in duplication resulted, a consensus was reached by discussion.
Criteria for a duplicate publication and duplication patterns
In this study, the criteria for duplicate publication of the Editorial Policy Committee of the Council of Science Editors (Editorial Policy Committee of the Council of Science Editors 1996) and a joint statement on duplicate publication established by editors of cardiothoracic journals (Cho et al. 2000) were used to define a duplicate publication. According to these criteria, duplicate publications have a similar hypothesis, similar numbers or sample sizes, identical or nearly identical methodology, similar results, authors in common, and no or little new information (Cho et al. 2000). We regarded the oldest or the largest article of a cluster as the main article. A cross-reference was considered clear if the corresponding article was acknowledged and referenced.
Duplication patterns were described by the classification of von Elm et al. (2004). They suggested the following 4 combinations of duplication patterns: identical samples and identical outcomes (copy), identical samples and different outcomes (fragmentation, salami slicing), increased samples and identical outcomes (imalas), and decreased samples and identical outcomes (disaggregation). To estimate the multiple publication bias, we performed a meta-analysis with and without duplicated data.
Identification of duplicates
Characteristics of duplicates and of main articles
Pattern of duplicate publications
No. of subjects
Patterns of duplicate publication
Year of publication
Source of funding
Identical study sample and reported partially identical outcomes (copy)
Documented parts of a large trial and reported identical outcomes (disaggregation)
100 (case 43)
202 (case 69)
New data were added to a main article and reported identical outcomes (Imalas)
Duplicates originated from 1 study sample but reported on different outcomes (fragmentation)
64 (case 11)
99 (case 26)
New data were added to a main article and reported almost identical outcomes (Imalas)
Documented parts of a large trial and reported identical outcomes(disaggregation)
Impact of duplication
Among the 6 meta-analyses including duplicate publications, we were unable to collect a full dataset of 4 meta-analyses. To evaluate the impact of duplicate publication on the results of meta-analyses, we thus reviewed 2 cases of meta-analyses including duplicate publications (meta-analyses 2 and 3 in Table 1). When a meta-analysis was performed without duplicated data, the mean effect size was 2.0054 (95% confidence interval [CI]: 1.8553, 2.1554) in meta-analysis 3. However, the mean effect size was increased to 2.1394 (95% CI: 1.6248, 2.6570) with duplicated data. The fail-safe number was also increased to 209.1 with duplicated data compared to 203.6 without duplicated data. In the case of meta-analysis 2, there was no difference in the mean effect size without duplicated data compared to the original article. However, the fail-safe number was increased to 17.5 with duplicated data compared to 14.7 without duplicated data.
To the best of our knowledge, the incidence and characteristics of duplicate publications in meta-analyses have not been evaluated at a nationwide level. Our study suggests that 6.9% of the meta-analysis articles that were written by Korean authors included duplicate publications. However, we found only 6 instances of duplicate publication after analyzing 1,194 articles used in meta-analyses that were written by Korean authors. The proportion of multiple publications among the 1,194 collected articles, 0.5%, may be an underestimate, since most duplicated studies were excluded by authors during the process of conducting a meta-analysis. Furthermore, in this study, we did not include meta-analyses not indexed in online databases.
Varied incidences of duplicate publication have been demonstrated in previous studies. In a larger study of the prevalence of redundant publication in the Association Archives of Otolaryngology-Head and Neck Surgery over 8 years, 1,965 authors of 24,353 articles were screened for duplicate publications (Bailey 2002). They found that 201 authors had published 443 redundant articles (1.8% duplication rate). Rosenthal et al. (2003) found that of 492 otolaryngology articles, 42 (8.5%) represented some form of duplication. Gwilym et al. (2004) reported that of 343 “original” articles published in the Journal of Bone and Joint Surgery in 1999, 26 (7.6%) had some degree of redundancy. Gøtzsche (1989) reported 44 (18%) multiple publications among 244 collected studies involving comparisons of nonsteroid anti-inflammatory drugs; the majority of these articles were published within 1 year of each other, and without subsequent reference to the original publication. More recently, Cheung et al. (2013) reported this relatively high incidence of duplicate publication has not significantly changed over 10 years. On the other hand, other studies have reported much lower incidences of duplicate publication. Durani found that 4 (<1%) of 431 original plastic surgery articles had some degree of redundancy (Durani 2006). Chennagiri et al. (2004) found only 14 duplicated articles (2%) among 600 articles published in the Hand Surgery literature. Arrive et al. (2008) reported only 2 instances of redundant publication among 362 original research articles published in Radiology.
In this study, duplicate publications were usually due to disaggregation and overlapping (imalas) publications. An imalas publication is an expansion of a main article through the addition of more data to produce a high-impact article. Disaggregation is the publication of several fragmented parts of a multicenter trial. Reporting the same trial more than once is unacceptable, except in the case of large, multicentre trials with numerous endpoints and following the acceptance of the editor. Generally, the title was modified, and the authors, study sample, and method underwent a minor change in the duplicate publications. All of the duplicate articles in our study were published within 2 years of the main article, and no clear cross referencing was present in the majority (83.3%). However, in 1 case in our study (meta-analysis 3), there was no common authorship. In this case, the study sample, method, outcome, and the department of the authors of the 2 articles were almost identical; we therefore concluded that the articles were overlapping (imalas) publications. The use of different authors for an identical study may be referred as plagiarism, and obscures duplication. Indeed, it may be extremely difficult for reviewers to determine whether 2 papers represent duplicate publications of 1 trial or 2 separate trials when 2 articles reporting the same trial do not share a single common author.
Meta-analyses in the nursing literature had the highest proportion of duplicate publications (33%), and all of the duplicate publications in this study were published before 2005. These cases of duplication might have happened accidentally, or through negligence, rather than by intention. In Korea, the concept of duplicate publications and publication ethics has been raised since 2005 following Professor Hwang’s case (Saunders and Savulescu 2008). Before 2005, Korean researchers might have had little knowledge of publication ethics, which might explain the relatively high rate of duplicate publications.
Detection and exclusion of duplicate data is an important step in conducting a meta-analysis. Our study found an increase in the mean effect size and fail-safe number with duplicated data. This finding means that the effect and reliability was overestimated in a meta-analysis. Multiple publication bias may be an important threat to the validity of meta-analyses. Most importantly, studies with significant results are more likely to lead to multiple publications and presentations (Easterbrook et al. 1991), which makes it more likely that they will be located and included in a meta-analysis. For instance, Tramer et al. (1997) concluded that the inclusion of redundant data in a meta-analysis led to a 23% overestimation of the treatment effectiveness of an antiemetic agent. Murphy and Wyllie (2009) reported that removing the overlapping case would decrease the sample size by 20% in a meta-analysis that quantified the association between initial inappropriate antimicrobial therapy and increased mortality in patients with ventilator-associated pneumonia.
The reasons for duplicate publication may be varied. The authors may have no concept of duplicate publication, or may desire to approach different audience groups. However, in most cases, duplicate publication occurs to boost the author’s career advancement; in other words, the author may feel compelled to publish more for scientific achievement, which is measured by the number of published articles. To prevent duplicate publication, regular education on publication ethics and notifications to members are needed.
There were some limitations in this study. The first limitation is that the judgments and categorizations were subjective in nature. Secondly, the analysis on the effect of duplicate publication on meta-analyses was performed on only 2 meta-analyses including duplicate publications. Furthermore, we did not take into account the confirmation of duplication by the original authors. However, to avoid subjective judgments, 4 reviewers independently reviewed the articles, and if necessary, consensus was reached by discussion. We believe that our findings are helpful in that they will increase the awareness of the problem of duplicate articles and reduce the incidence of duplicate publications in future.
Our study found only 6 instances of duplicate publication after analyzing 1,194 articles used in meta-analyses written by Korean authors. However, 6.9% of the meta-analyses included duplicate publications. This kind of multiple publication bias will result in an accentuated positive effect of interventions. Our findings suggest that meta-analyses should be interpreted cautiously, taking into account the possibility of duplicated studies.
This research was supported by the Korea Research Foundation Grant funded by the Korean Government (KRF-I00030-C00008). We appreciate Kyung-Mi Nam (the Catholic University of Korea) for data collection.
- Arrive L, Lewin M, Dono P, Monnier-Cholley L, Hoeffel C, Tubiana JM: Redundant publication in the journal radiology. Radiology 2008, 247(3):836-840. doi:10.1148/radiol.2473070969 10.1148/radiol.2473070969View ArticleGoogle Scholar
- Bailey BJ: Duplicate publication in the field of otolaryngology-head and neck surgery. Otolaryngol Head Neck Surg 2002, 126(3):211-216. doi:10.1067/mhn.2002.122698 10.1067/mhn.2002.122698View ArticleGoogle Scholar
- Chennagiri RJ, Critchley P, Giele H: Duplicate publication in the Journal of Hand Surgery. J Hand Surg (Br) 2004, 29(6):625-628. doi:10.1016/j.jhsb.2004.04.005 10.1016/j.jhsb.2004.04.005View ArticleGoogle Scholar
- Cheung VW, Lam GO, Wang YF, Chadha NK: Current incidence of duplicate publication in otolaryngology. Laryngoscope 2013. doi:10.1002/lary.24294Google Scholar
- Cho BK, Rosenfeldt F, Turina MI, Karp RB, Ferguson TB, Bodnar E, Waldhausen JA: Joint statement on redundant (duplicate) publication by the editors of the undersigned cardiothoracic journals. Ann Thorac Surg 2000, 69(2):663. doi:10.1016/S0003-4975(00)01088-2 10.1016/S0003-4975(00)01088-2View ArticleGoogle Scholar
- Durani P: Duplicate publications: redundancy in plastic surgery literature. J Plast Reconstr Aesthet Surg 2006, 59(9):975-977. doi:10.1016/j.bjps.2005.11.039 10.1016/j.bjps.2005.11.039View ArticleGoogle Scholar
- Easterbrook PJ, Berlin JA, Gopalan R, Matthews DR: Publication bias in clinical research. Lancet 1991, 337(8746):867-872. doi:10.1016/0140-6736(91)90201-Y 10.1016/0140-6736(91)90201-YView ArticleGoogle Scholar
- Editorial Policy Committee of the Council of Science Editors: Redundant publication. CBE Views 1996, 19: 76-77.Google Scholar
- Gotzsche PC: Multiple publication of reports of drug trials. Eur J Clin Pharmacol 1989, 36(5):429-432. doi:10.1007/BF00558064 10.1007/BF00558064View ArticleGoogle Scholar
- Gwilym SE, Swan MC, Giele H: One in 13 ‘original’ articles in the Journal of Bone and Joint Surgery are duplicate or fragmented publications. J Bone Joint Surg Br 2004, 86(5):743-745. doi:10.1302/0301-620X.86B5.14725 10.1302/0301-620X.86B5.14725View ArticleGoogle Scholar
- Huston P, Moher D: Redundancy, disaggregation, and the integrity of medical research. Lancet 1996, 347(9007):1024-1026. doi:10.1016/S0140-6736(96)90153-1 10.1016/S0140-6736(96)90153-1View ArticleGoogle Scholar
- International Committee of Medical Journal Editors: Recommendations for the Conduct, Reporting, Editing, and Publication of Scholarly Work in Medical Journals (ICMJE Recommendations). 2013. . Accessed 15 Aug 2013 http://www.icmje.org/urm_main.htmlGoogle Scholar
- Johnson C: Repetitive, duplicate, and redundant publications: a review for authors and readers. J Manipulative Physiol Ther 2006, 29(7):505-509. doi:10.1016/j.jmpt.2006.07.001 10.1016/j.jmpt.2006.07.001View ArticleGoogle Scholar
- Kim SY, Hahm CK, Bae CW, Cho HM: Duplicate Publications in Korean medical journals indexed in KoreaMed. J Korean Med Sci 2008, 23(1):131-133. doi:10.3346/jkms.2008.23.1.131 10.3346/jkms.2008.23.1.131View ArticleGoogle Scholar
- Leopold SS: Editorial: duplicate submission and dual publication: what is so wrong with them? Clin Orthop Relat Res 2013, 471(5):1401-1402. doi:10.1007/s11999-013-2916-8 10.1007/s11999-013-2916-8View ArticleGoogle Scholar
- Mojon-Azzi SM, Jiang X, Wagner U, Mojon DS: Redundant publications in scientific ophthalmologic journals: the tip of the iceberg? Ophthalmology 2004, 111(5):863-866. doi:10.1016/j.ophtha.2003.09.029 10.1016/j.ophtha.2003.09.029View ArticleGoogle Scholar
- Murphy L, Wyllie A: Duplicate patient data in a meta-analysis; a threat to validity. J Crit Care 2009, 24(3):466-467. doi:10.1016/j.jcrc.2008.12.012 10.1016/j.jcrc.2008.12.012View ArticleGoogle Scholar
- Rosenthal EL, Masdon JL, Buckman C, Hawn M: Duplicate publications in the otolaryngology literature. Laryngoscope 2003, 113(5):772-774. doi:10.1097/00005537-200305000-00002 10.1097/00005537-200305000-00002View ArticleGoogle Scholar
- Saunders R, Savulescu J: Research ethics and lessons from Hwanggate: what can we learn from the Korean cloning fraud? J Med Ethics 2008, 34(3):214-221. doi:10.1136/jme.2007.023721 10.1136/jme.2007.023721View ArticleGoogle Scholar
- Schein M, Paladugu R: Redundant surgical publications: tip of the iceberg? Surgery 2001, 129(6):655-661. doi:10.1067/msy.2001.114549 10.1067/msy.2001.114549View ArticleGoogle Scholar
- Tramer MR, Reynolds DJ, Moore RA, McQuay HJ: Impact of covert duplicate publication on meta-analysis: a case study. BMJ 1997, 315(7109):635-640. doi:10.1136/bmj.315.7109.635 10.1136/bmj.315.7109.635View ArticleGoogle Scholar
- von Elm E, Poglia G, Walder B, Tramer MR: Different patterns of duplicate publication: an analysis of articles used in systematic reviews. JAMA 2004, 291(8):974-980. doi:10.1001/jama.291.8.974 10.1001/jama.291.8.974View ArticleGoogle Scholar
- Wood JA: Methodology for dealing with duplicate study effects in a meta-analysis. Organ res methods 2007, 11(1):79-95. doi:10.1177/1094428106296638 10.1177/1094428106296638View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.