Abstract
ARAB JOURNAL OF UROLOGY 2022, VOL. 20, NO. 3, 168–173 https://doi.org/10.1080/2090598X.2022.2056976 ORIGINAL ARTICLE Vaibhav Aggarwal Department of Surgical Disciplines, All India Institute of Medical Sciences (AIIMS), New Delhi, India ABSTRACT ARTICLE HISTORY Received 16 January 2022 Objective: We aimed to determine the diachronous self-citation rate and the various article Accepted 12 February 2022 characteristics which can influence the rate and percentage of diachronous author self- citations using papers published in high-rank urology journals. KEYWORDS Methods: We included all papers (N = 327 articles) published between January 2015 to Author self-citations; April 2015 in the European Urology, The Journal of Urology and the BJU International. We bibliometrics; diachronous determined author self-citations using the Scopus database and used negative binomial self-citations; negative regression to determine which article characteristics affect self-citations. binomial regression Results: 262 articles (80.2%) contained at least one self-citation.The mean number and per- centage of author self-citations were 6.5 and 14.2 respectively. Adjusted analysis showed that the experimental/animal study design and the number of authors were significantly associated with both the number (IRR = 2.12, P = 0.011; IRR = 1.03, P = 0.002) and percentage of author self-citations (IRR = 2.95, P = 0.003; IRR = 1.03, P = 0.012). The number of citations in the Scopus and publication in European Urology were significantly associated with only the number of author self-citations. Conclusion: Diachronous author self-citation rate in urology is higher compared to general medical literature but similar to other surgical subspecialties. It may depend on the study design and the number of authors in the paper. For a more comprehensive evaluation, future studies should look at the context in which self-citations were made. Introduction that among surgical subspecialties, urologists and neu- The quest to publish in academia has grown signifi - rosurgeons had a considerably higher number of cantly over the years and academic surgery is no h-index [11]. The reasons for such differences are not exception. Nonetheless, This is much needed to fill in clear. Moreover, article characteristics on which self- the evidence gaps to guide decision making, deliver citations depend have not been adequately evaluated the best possible quality of care and reduce variations in the surgical literature. in surgical practice [1]. Academically productive sur- This study aimed to determine the diachronous self- geon scientists are often preferred for hiring and pro- citation rate and the various article characteristics motions to major academic positions. Scholarly which can influence the rate and percentage of dia- productivity is usually assessed by the number of pub- chronous self-citations using papers published in lications, grant funding or citation based metrics such a 4 month period in the 3 high ranking urology sub- as the number of citations and h-index [2]. However, specialty journals. citation-based metrics are prone to manipulation through self-citations [3]. This practice could be mis- Methodology leading and can potentially distort the scientific litera- ture [4]. Study design and data collection Lawani has defined 2 genera of self-citations: syn- As the study did not involve human participants, chronous (the ones which are contained in the refer- research ethics committee approval was not applic- ence list of the indexed article) and diachronous (the able. This is a retrospective study that included articles ones which are contained in the citations the indexed published between January 2015 to April 2015 in the article receives) [5]. Previous studies have tried to European Urology (EU), The Journal of Urology (JU) and quantitatively measure the impact of author self- BJU International (BJUI) through hand searching. These citations on citation metrics in general medicine [6], were chosen based on their importance in urology plastic surgery [7], otolaryngology [8], orthodontics [9] according to the SCImago Journal Rank indicator. and radiology [10]. However, Little is known about These journals were accessed through institutional diachronous author self-citations rates in surgical sub- access. specialties such as urology. In addition, it was found CONTACT Vaibhav Aggarwal aggarwalvaibhav1995@gmail.com Department of Surgical Disciplines, All India Institute of Medical Sciences (AIIMS), New Delhi, India © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. ARAB JOURNAL OF UROLOGY 169 The articles were subdivided by study design such signifying overdispersion. In addition, these variables as systematic reviews/ meta-analysis/ reviews, inter- did not have an excessive number of zero values. ventional studies, prospective observational studies, Therefore, we rejected the Poisson regression and the retrospective studies/ case series/ case reports and zero-inflated models. We selected the negative bino- experimental/ animal studies. Articles types such as mial regression model to determine the effect of the abstracts, editorials, and meeting reviews were article characteristics on self-citations (Figure 1). To excluded. transform the percentage of author self-citations into In April 2021, after 6 hours of training in the use of count data, we rounded off the percentage to the Scopus at the institute’s library, the author extracted nearest whole number. We included the following certain article characteristics from each of the above- independent variables in the model: 1) The journal in selected articles as described below. The author ran- which the article appeared; 2) The study design; 3) domly selected 40 articles to confirm the accuracy of Study sub-topic; 4) Country of the corresponding data collection. author; 5) Whether the research was funded; 6) For each article we extracted the following Whether the article is an open-access; 7) Number of characteristics: characters in the title; 8) The total number of authors; 9) Number of references; 10) The total number of cita- (1) The journal in which the article appeared (EU, tions in the Scopus database. JU, BJUI) The analysis was performed on the entire study (2) Study sub-topic cohort. There were no missing values in the data set. (3) The study design (experimental/ animal stu- All statistical tests were two-sided, and P values of less dies; guidelines; case reports/ case series/ ret- than 0.05 were considered to indicate statistical rospective studies/ cross sectional studies; significance. prospective observational studies; interven- tional studies and reviews/ systematic review/ Results meta-analysis) (4) The month of publication (January 2015 to Overall, 327 articles were included in the analysis. April 2015) Characteristics of the included articles are shown in (5) Country of the corresponding author Table 1. Of these, EU, JU and BJUI constituted 21.7%, (6) Total number of authors 53.5% and 24.8% of total articles respectively. 37.3% of (7) Number of characters in the title (excluding the articles received funding and 30% of the articles trailing or double spaces) were open access.‘Prostate’ was the most studied topic (8) The total number of citations in the Scopus (35.8%). Retrospective studies and case series were the database (this includes the absolute number most commonly used study designs. The correspond- of citations received by publications until ing author was affiliated with an institution in the April 2021). United States in 45.9% of the articles. The mean num- (9) Whether the research was funded. ber of title character count, reference count and (10) Whether the article is an open access authors were 121.6 (range 36–275), 28.1 (range 5–80) and 9.2 (range 1–42) respectively. The mean number of citations and citations per year Dependent variables received by an article in the Scopus database was 45.9 (1) Number of author self-citations in Scopus (range 1–411) and 7.7 (range 0.16–68.5). 262 articles database (80.2%) contained at least one self-citation. The mean (2) Percentage of author self-citations in Scopus number of author self-citations was 6.5 (range 0–65). database (determined by the proportion of The percentage of author self-citations (calculated author self-citations to the total number of cita- from the proportion of the sum of author self- tions received by the article) citations of all the included articles with the sum of all citations of all the included articles) was 14.2%. Statistical analysis Negative binomial regression analysis We inserted all the data for each included article into The variance inflation factor (VIF) was calculated for each spreadsheet software. This database was then ana- variable included in the model. All VIF values were <3 lysed using IBM SPSS Statistics v26. Descriptive statis- showing that there was no multicollinearity. Therefore, tics were used to quantitatively describe the features all variables were considered in the regression analysis. of the sample. The variance in the number of self- Adjusted analysis showed that experimental/animal citations, and the percentage of author self-citations study design was statistically significantly associated in this study were much greater than their mean with both the number of author self-citations (incidence 170 V. AGGARWAL Figure 1. A. Frequency distribution of number of author self citations. B. Frequency distribution of percentage of author self citations. The distributions are right skewed and follow a negative binomial pattern. rate ratio (IRR) = 2.12; 95% confidence interval (CI) = 1.2 Discussion to 3.77; P = 0.011) and with percentage of author self In this retrospective study of 327 articles, evaluating citations (IRR = 2.95; 95% CI = 1.46 to 5.96; P = 0.003). the impact of various articles characteristics on Similarly, the number of authors in an article was author self-citation counts, we found that the over- strongly correlated with both the number of author self- all percentage of self-citation in urology literature citations (IRR = 1.03; 95% CI = 1.01 to 1.05; P = 0.002) and was 14.2% with a median of 3. Following article with percentage of author self citations (IRR = 1.03; 95% characteristics were associated with a higher num- CI = 1.007 to 1.06; P = 0.012). ber of self-citations: experimental/ animal studies, Publication of an article in European Urology and higher number of authors, publication of an article the total number of citations in the Scopus database in EU and the total number of citations in the was also statistically significantly associated with the Scopus database. Similar results were obtained for number of author self-citations (IRR = 1.67; 95% experimental/animal studies and higher number of CI = 1.14 to 2.45; P = 0.009 and IRR = 1.01; 95% authors when considering the percentage of author CI = 1.01 to 1.02; P = 0.000) but was not associated self-citations as the outcome. However, publication with the percentage of author self-citations. It is impor- in the EU and the total number of citations were tant to note that these isolated significant results on not found to be significantly associated. the number of self citations from regression analysis The rate of self-citations across various disciplines of may just be due to higher citation counts. No statisti- medicine has been variable. Kulkarni et al. report that it cally significant association was found with other vari- amounts to 7% in the general medical literature [6]. ables included in the regression (Table 2). ARAB JOURNAL OF UROLOGY 171 Table 1. Baseline characteristics*. Variables Articles (study group) (N = 327) Journal 71 (21.7) European Urology 175 (53.5) Journal of Urology 81 (24.8) BJU International Month of publication 83 (25.4) January 2015 79 (24.2) February 2015 77 (23.5) March 2015 88 (26.9) April 2015 Study design 31 (9.5) Experimental/ Animal studies 6 (1.8) Guidelines 172 (52.6) Case reports/ case series/ retrospective studies/ cross sectional studies 41 (12.5) Prospective observational studies 40 (12.2) Interventional studies 37 (11.3) Reviews/ Systematic review/ meta-analysis Country of the corresponding author 7 (2.1) Australia 22 (6.7) Canada 30 (9.2) United Kingdom 150 (45.9) United States 52 (15.9) France/Germany/Netherlands/Italy 66 (20.2) others Topic 46 (14.1) Renal 7 (2.1) Ureter 64 (19.6) Bladder 117 (35.8) Prostate 17 (5.2) Urethra/penis 20 (6.1) Sexual/reproductive 56 (17.1) Miscellaneous Funding 122 (37.3) Yes 205 (62.7) No Open Access 98 (30) Yes 229 (70) No Character count of the title 121.6 ± 41.4 Number of authors 9.17 ± 5.8 Reference count 28.12 ± 12.8 Citation count in Scopus database 45.94 ± 54.2 Citations per year in Scopus database 7.7 ± 9 Self citation count in Scopus database 6.54 ± 9.6 Percentage of author self citations 14.2% *Plus-minus values are means ±SD. Number in parenthesis denotes percentages. # Denote these countries: Brazil/ Belgium/ Sweden/ Spain/ South Korea/ Denmark/ China/ Japan/ Switzerland/ India/ Egypt/ Greece/ Ireland/ Turkey/ New Zealand/ Norway/ Portugal/ Singapore/ Taiwan However, the self-citation rate is higher in subspeciality observations [6] [8]. We can anticipate this however literature including diabetes [12], otolaryngology [8], since more authors are available to self cite. This might cardiovascular medicine, infectious diseases [6] and also help increase the interdisciplinary visibility and musculoskeletal radiology [10] similar to the one in our influence of the paper, which may attract more total study. This is true for journal self-citations as well such as citations and self-citations. Studies conducted on ‘cita- dermatology [13] and orthopaedics [14]. Moreover, we tion classics’ in surgery and its subspecialties consis- found that experimental/ animal studies are associated tently show that western and European countries with higher self-citations similar to observations made produce the most number of top-cited papers [15] by authors of otolaryngology literature [8]. This is [16]. But this did not translate to a high self-citation expected since expertise in subspecialized fields may rate in our study and other similar studies. This may be be limited. In addition, science built on basic hypotheses because literature considered ‘important’ in the field is more likely to inspire further research by the same does not require self promotion to be well recognized group. When such expertise is coupled with novel by peers irrespective of the topic of study. This argu- research questions and a limited number of journals to ment is further strengthened by the finding that the publish in, self-citations become inevitable. percentage of author self-citations was slightly nega- The number of authors was independently signifi - tively correlated with the number of non-self citation cantly associated with higher self-citations. Previous counts in Scopus that reached statistical significance studies on self-citations have yielded similar (r = −0.13, P = 0.018). In addition, some data indicate 172 V. AGGARWAL Table 2. Negative binomial regression IRR for the outcomes. Variable Number of author self citations Percentage of author self citations Journal 1.17 (1.14–2.45) 1.48 (0.9–2.41) European Urology 1.27 (0.93–1.7) 1.25 (0.87–1.8) Journal of Urology 1 1 BJU International Study design 2.12 (1.19–3.77) 2.95 (1.46–5.96) Experimental/ Animal studies 1.27 (0.55–2.94) 1.49 (0.51–4.34) Guidelines 1.49 (0.93–2.37) 1.67 (0.94–2.94) Case reports/ case series/ retrospective studies/ cross sectional studies 1.56 (0.92–2.65) 1.68 (0.88–3.21) Prospective observational studies 1.23 (0.7–2.16) 1.38 (0.71–2.69) Interventional studies 1 1 Reviews/ Systematic review/ meta-analysis Country of the corresponding author 1.4 (0.64–3.05) 1 (0.37–2.68) Australia 0.91 (0.56–1.5) 0.74 (0.4–1.36) Canada 0.95 (0.6–1.5) 0.87 (0.49–1.54) United Kingdom 1.04 (0.76–1.42) 0.89 (0.6–1.31) US 1.22 (0.83–1.8) 0.99 (0.6–1.62) France/Germany/Netherlands/Italy 1 1 others Topic 0.82 (0.54–1.24) 0.74 (0.45–1.21) Renal 0.8 (0.36–1.84) 0.7 (0.26–1.9) Ureter 1.07 (0.73–1.56) 1.1 (0.7–1.73) Bladder 1.06 (0.75–1.49) 1.06 (0.7–1.6) Prostate 1.03 (0.59–1.79) 1.06 (0.54–2.06) Urethra/penis 0.86 (0.49–1.49) 0.91 (0.48–1.76) Sexual/reproductive 1 1 Miscellaneous Funding 0.98 (0.76–1.27) 0.93 (0.68–1.28) Yes 1 1 No Open Access 1.16 (0.88–1.53) 1.08 (0.78–1.5) Yes 1 1 No Character count of the title 1 (0.99–1.0) 0.999 (0.996–1.003) Number of authors 1.03 (1.01–1.05) 1.03 (1.007–1.06) Reference count 1 (0.99–1.01) 1.0 (0.99–1.02) Citation count in Scopus database 1.014 (1.011–1.017) 0.998 (0.995–1.001) *IRR stands for incidence rate ratios. Values in the table are expressed as IRR (95% confidence interval) # Denote these countries: Brazil/ Belgium/ Sweden/ Spain/ South Korea/ Denmark/ China/ Japan/ Switzerland/ India/ Egypt/ Greece/ Ireland/ Turkey/ New Zealand/ Norway/ Portugal/ Singapore/ Taiwan. that the self-citation rate may be higher in papers that study, however. We used a retrospective study design are poorly cited [17]. Thus the rate and percentage of which has an inherent potential for unknown confound- self-citations remained similar and statistically non- ing. Furthermore, this study included articles from the EU, significant across nations and the topic of the study. JU and BJUI which are the high ranking urology subspeci- Self-citations could be viewed as a double edged alty journals. Subspeciality journals often publish papers sword. On one hand, they are healthy for scientific having a narrow clinical readership. Thus, these findings research and are needed (and sometimes even una- may not be generalizable to general medical and surgical voidable) to network with previously established con- journals which have a broad readership and a broad citing cepts or observations and aids in understanding and community. Similarly, these results may not be general- expanding on research hypotheses. This is how Mishra izable to small journals as they often have a small reader- et al [18]. justifies the existence of self-citations, ‘lack of ship and less often publish highly cited papers. These ‘less (self-)citations likely reflects dead-ends and orphans prestigious’ journals may have higher self-citation rates not even nurtured by the scholars with a vested inter- [19]. Lastly, though Scopus has a fair accuracy in deter- est.’ However, when manipulative, self-citations may mining author self-citations, results might vary if other propagate a specific thought process or methodology databases are used. which may be biased and unscientific. It may also be able to create a false perception of the importance or Conclusion authority of an individual in the field particularly in a subspecialty where there are a limited number of Diachronous author self-citation rate in urology is higher well-cited authors. compared to general medical literature but similar to We used robust statistical methods in this paper. We other surgical subspecialties. It may depend on the determined which article characteristics influence self- study design and the number of authors in the paper. citation in subspecialty urosurgery literature using Though generally viewed negatively in the scientific com- a sufficiently long publication window and by adjusting munity, self-citations may be perfectly healthy. It is the for potential confounders. There are limitations to this intention behind self-citations that holds more ARAB JOURNAL OF UROLOGY 173 importance but may be difficult to measure. For a more [8] Tolisano AM, Song SA, Cable BB, et al. “Author self-citation in the otolaryngology literature,” Head comprehensive evaluation, future studies should look at Neck Surg., p. 5. the context in which self-citations were made. [9] Livas C, Delli K, Pandis N. Author self-citation in ortho- dontics is associated with author origin and gender. Prog Orthod. 2021 Jan;22(1):DOI:10.1186/s40510-020- Disclosure statement 00348-y [10] Dessouky R, Zhang L, Wadhwa V, et al. Self-citations in No potential conflict of interest was reported by the musculoskeletal radiology: frequency and pattern author(s). analysis. Acta Radiol. 2019 Nov;60(11):1490–1495. [11] Svider PF, Pashkova AA, Choudhry Z, et al. Comparison of scholarly impact among surgical specialties: an ORCID examination of 2429 academic surgeons. Laryngoscope. 2013;123(4):884–889. Vaibhav Aggarwal http://orcid.org/0000-0001-6820-3304 [12] Gami AS, Montori VM, Wilczynski NL, et al. Author self-citation in the diabetes literature. 3. [13] Reiter O, Mimouni M, Mimouni D. Analysis of References self-citation and impact factor in dermatology journals. Int J Dermatol. 2016 Sep;55(9):995–999. [1] Lancet T. The Lancet. Variation in surgery and surgical [14] Sundaram K, Warren J, Anis HK, et al. Publication research. Lancet. 2013 Sep;382(9898):1071. integrity in orthopaedic journals: the self-citation in [2] Kelly CD, Jennions MD. The h index and career assess- orthopaedic research (SCOR) threshold. 2020 ment by numbers. 2006 Apr;21(4):167–170.Trends Ecol. May;30(4):629–635.Eur. J. Orthop. Surg. Traumatol. Evol. [15] Long X, Huang J-Z, Ho Y-S. A historical review of classic [3] Bartneck C, Kokkelmans S. Detecting h-index manipula- articles in surgery field. Am J Surg. 2014 Nov;208 tion through self-citation analysis. Scientometrics. (5):841–849. 2011;87(1):85–98. [16] Ahmad SS, Ahmed A, Exadaktylos A, et al. Systematic [4] Ioannidis JPA. A generalized view of self-citation: review on citation classics in minimally invasive gas- direct, co-author, collaborative, and coercive induced trointestinal surgery. J. Minimal Access Surg. 2018;14 self-citation. J Psychosom Res. 2015 Jan;78(1):7–11. (4):265. [5] Lawani SM. On the heterogeneity and classification of [17] Aksnes DW. A macro study of self-citation. author self-citations. 2007 Sep;33(5):281–284.J. Am. Scientometrics. 2003 Feb;56(2):235–246. Soc. Inf. Sci. [18] Mishra S, Fegley BD, Diesner J, et al. Self-citation is the [6] Kulkarni AV, Aziz B, Shams I, et al. Author self-citation hallmark of productive authors, of any gender. PLoS in the general medicine literature. PLoS ONE. 2011 Jun;6(6):e20885. ONE. 2018 Sep;13(9):e0195773. [7] Swanson EW, Miller DT, Susarla SM, et al. What effect [19] Fassoulaki A, Papilas K, Paraskeva A, et al. Impact factor does self-citation have on bibliometric measures in aca- bias and proposed adjustments for its determination. demic plastic surgery? Ann. Plast. Surg. 2016;77(3):4. Acta Anaesthesiol Scand. 2002;46(7):902–905.
Journal
Arab Journal of Urology
– Taylor & Francis
Published: Jul 3, 2022
Keywords: Author self-citations; bibliometrics; diachronous self-citations; negative binomial regression
