Antibiotics: A Bibliometric Analysis of Top 100 Classics

Citation frequencies represent the most significant contributions in any respective field. This bibliometric analysis aimed to identify and analyze the 100 most-cited publications in the field of antibiotics and to highlight the trends of research in this field. “All databases” of Clarivate Analytics’ Web of Science was used to identify and analyze the 100 publications. The articles were then cross-matched with Scopus and Google Scholar. The frequency of citation ranged from 940 to 11,051 for the Web of Science, 1053 to 10,740 for Scopus, and 1162 to 20,041 for Google Scholar. A total of 513 authors made contributions to the ranked list, and Robert E.W. Hancock contributed in six articles, which made it to the ranked list. Sixty-six scientific contributions originated from the United States of America. Five publications were linked to the University of Manitoba, Canada, that was identified as the educational organization, made the most contributions (n = 5). According to the methodological design, 26 of the most cited works were review-type closely followed by 23 expert opinions/perspectives. Eight articles were published in Nature journal, making it the journal with the most scientific contribution in this field. Correlation analysis between the publication age and citation frequency was found statistically significant (p = 0.012).


Introduction
The bibliometric analysis provides a quantitative review of literature in any field of research based on the citation frequency of the conducted research. This type of analysis identifies the countries,

Institution and Country of Origin
The author's affiliation and origin country of publication were retrieved from the ES database as complete information for the marked list was not available from the WoS database. The retrieved information was then hand-searched and compared with the original text for each manuscript. Although corresponding addresses are considered a reliable source to identify the country of origin of publication; however, upon searching manually, it was seldom recorded. Each institution contributing to the publication was recorded as a single entry.

Data Analysis
The "Visualization of Similarities (VOS) viewer software" is widely used to graphically illustrate the bibliometric parameters in mapping networks, which allow easy visualization of critical elements [2,[19][20][21]. The current study used VOS to represent a graphical mapping of keywords as identified bibliometric analysis to identify the focus of research in recent decades.

Statistical Analysis
The descriptive data and associations of citation frequency, citation density, publication age, and CCI were analyzed using IBM SPSS Statistics ® , version 22, using the Spearman rank test. The normality of data was checked using the Shapiro-Wilk test. To explore the difference between two or more independent groups, the Kruskal-Wallis test was performed. Post-hoc testing was performed to confirm the difference between variables. Mann-Kendall trend test was performed to determine increasing and decreasing time trends. A p-value of < 0.05 was considered statistically significant.

Bibliometric Parameters
The marked list of top 100 classics received a sum of 167,320 citations based on WoS, 165,947 citations based on ES, and 262,727 based on the GS database. The frequency of citations ranged from 940 to 11,051 (WoS), 1053 to 10,740 (ES), and 1162 to 20,041 (GS). Citation density is defined as the average number of citations/annum; it was calculated as 2742 (WoS), 2720 (ES), and 4307 (GS) for the 100 classics. "Antibiotic susceptibility testing by a standardized single disk method" was identified as the most cited "classic" with 11,051, 10,740, and 20,041 citations according to WoS, ES, and GS databases, respectively, with a citation density of 205 [22]. "Antimicrobial peptides of multicellular organisms" was ranked second with 5685, 5668, 7994 citations according to WoS, ES, and GS databases, respectively, with a citation density of 316 [23]. "Transformation of mammalian cells to antibiotic resistance with a bacterial gene under the control of the SV40 early region promoter" was ranked third with 3891, 2319, 3875 citations according to WoS, ES, and GS databases, respectively, with a citation density of 102 [24]. The marked list of top 100 classics along with their citation frequency from WoS, ES, and GS databases, publication age, citation density, and CCI 2019 is presented in Table 1. Shapiro-Wilk test revealed non-normal data on the citation frequency, citation density, and age of publication (years). Figure 1a shows a statistically significant upward trend of citation frequency was noted with the increase in publication age (R 2 = 0.044, p = −0.012). Figure 1b shows a downward trend of citation density was noted with an increase in the age of publication (R 2 = 0.304, p = −0.551), which was not statistically significant. The Supplementary Figure S1 illustrates the distribution of citation frequency over the last six decades.

Year of Publication
Chronologically, the oldest classic with 60 years of publication age was published in 1959 [60], and three articles with four years of publication age were published in 2015 [92,109,119]

Methodological Design and Evidence Level (EL)
The distribution of the list based on methodological design is illustrated in Figure 2. Based on the level of evidence, 71 publications were graded as level-V, two were graded as level-IV, one belonged to level-III, four publications were graded as level-II, and 17 were graded as level-I. The evidence level and methodological design of five publications [24,52,60,77,111] were not identified as full-text of the articles were not accessible through different electronic sources.

Year of Publication
Chronologically, the oldest classic with 60 years of publication age was published in 1959 [60], and three articles with four years of publication age were published in 2015 [92,109,119]

Methodological Design and Evidence Level (EL)
The distribution of the list based on methodological design is illustrated in Figure 2. Based on the level of evidence, 71 publications were graded as level-V, two were graded as level-IV, one belonged to level-III, four publications were graded as level-II, and 17 were graded as level-I. The evidence level and methodological design of five publications [24,52,60,77,111] were not identified as full-text of the articles were not accessible through different electronic sources.

Contributing Authors, Institutions, and Countries
Robert E.W. Hancock was identified as the most contributing, authoring six classics, followed by Tomas Ganz, who contributed in four classics. A total of 513 authors contributed to the top 100 classics, among them 26 authors were contributed in two "classics" each. Complete texts for 95

Contributing Authors, Institutions, and Countries
Robert E.W. Hancock was identified as the most contributing, authoring six classics, followed by Tomas Ganz, who contributed in four classics. A total of 513 authors contributed to the top 100 classics, among them 26 authors were contributed in two "classics" each. Complete texts for 95 publications were obtained, and five publications were not accessible through different institutions [24,52,60,77,111]. Based on the institutional address of the corresponding author as retrieved from the ES database, individuals from 26 countries contributed to the "classic" articles. Among these, 69 scientific contributions were from the United States of America. Followed by 18 publications from Canada, 11 from Germany, and four from Sweden. Three publications originated from Belgium, China, and Israel. Two publications originated from Egypt, Denmark, and India. One publication originated from, Argentina, Croatia, Ecuador, France, Kenya, Korea, Netherlands, New Zealand, Pakistan, South Africa, South Korea, Spain, Tanzania, Thailand, United Kingdom, and Australia.
Among 246 international institutions, the greatest contribution to the "classic" articles was made by the University of Manitoba, Canada, in six classics followed by the Stanford University School of Medicine, USA, in five classics. "University of Washington, USA", "University of British Colombia, Canada", "The University of California at Los Angeles, USA", and "Harvard University, USA" contributed in four classics. "University of Kiel, Germany" and "University of California at San Diego, USA" contributed in three classics. "Robert Wood Johnson Medical School, USA", "Weizmann Institute of Science, Israel", "Emory University, Atlanta, Georgia, USA", "Laurentian University, Ontario, Canada", "Rush-Presbyterian-St. Luke's Medical Center, USA", "St. Agnes Medical Center, USA", "the Centers for Disease Control and Prevention, Atlanta, Georgia, USA", and "Veterans Affairs Palo Alto Health Care System, California, USA" contributed to two classics each.

Journal of Publication
The 100 classics were published across 63 different journals. Figure 3 presents the list of journals in which the highest number of classics were published. The list of the remaining journals is available as Supplementary Table S1.

Keywords
The most frequently occurring keywords in the top 100 classics were "anti-bacterial agents" and "antibiotic agent", followed by "antibiotic resistance", "anti-infective agent", and "antimicrobial". Figure 4 is a graphical presentation of keywords arranged in a network of clusters. Colorful nodes represent the linkage of specific keywords to each cluster. Table S2 enlists the total number of index keywords and their frequency of occurrence based on the Elsevier Scopus database.

Keywords
The most frequently occurring keywords in the top 100 classics were "anti-bacterial agents" and "antibiotic agent", followed by "antibiotic resistance", "anti-infective agent", and "antimicrobial". Figure 4 is a graphical presentation of keywords arranged in a network of clusters. Colorful nodes represent the linkage of specific keywords to each cluster. Table S2 enlists the total number of index keywords and their frequency of occurrence based on the Elsevier Scopus database.

Discussion
The current study identified and analyzed the top 100 classics on antibiotics, antimicrobials, or antibacterial agents. Identification of any scientific contribution and inclusion in classics warrants the excellence and acclaimed acknowledgment by the relevant field experts, researchers, and scientists [12]. Theoretically, a higher citation frequency of a publication indicates the quality of the research conducted as identified by the scientific community [122]. Identification is imperative to study whether the classics have elaborated or explored the understanding of a problem and/or provided a comprehensive approach towards its solution, or whether the publication introduced a research trend or provided an expert opinion/summary on a topic of interest. The results of this study present the research perspective in the field of antibiotics, antimicrobials, or antibacterial agents for the last six decades. Additionlly, it illustrates key trends of research as well as clinical practice [2,8].
The definition of "classics" largely depends on the research field/specialty to which the publication belongs. In some fields, 100 or more citations of a publication are considered enough to classify it as a "classic" [6]. In perspective, the article ranked as 100th in the current study received 940 citations in comparison with the article ranked as 1st in the field of physics research in Korea that received 302 citations [123] or with the article ranked as 1st in the dental caries research that received 2003 citations [19]. For the current study, the publications receiving more than 400 citations can be considered classics. However, these publications will not make it to the top 100 due to the immense availability of the highly cited publications.

Discussion
The current study identified and analyzed the top 100 classics on antibiotics, antimicrobials, or antibacterial agents. Identification of any scientific contribution and inclusion in classics warrants the excellence and acclaimed acknowledgment by the relevant field experts, researchers, and scientists [12]. Theoretically, a higher citation frequency of a publication indicates the quality of the research conducted as identified by the scientific community [122]. Identification is imperative to study whether the classics have elaborated or explored the understanding of a problem and/or provided a comprehensive approach towards its solution, or whether the publication introduced a research trend or provided an expert opinion/summary on a topic of interest. The results of this study present the research perspective in the field of antibiotics, antimicrobials, or antibacterial agents for the last six decades. Additionlly, it illustrates key trends of research as well as clinical practice [2,8].
The definition of "classics" largely depends on the research field/specialty to which the publication belongs. In some fields, 100 or more citations of a publication are considered enough to classify it as a "classic" [6]. In perspective, the article ranked as 100th in the current study received 940 citations in comparison with the article ranked as 1st in the field of physics research in Korea that received 302 citations [123] or with the article ranked as 1st in the dental caries research that received 2003 citations [19]. For the current study, the publications receiving more than 400 citations can be considered classics. However, these publications will not make it to the top 100 due to the immense availability of the highly cited publications.
Web of Science was used as a benchmark database because it has citation metrics from 1945 to the present [124]. A significant variance was observed when the citation metrics were cross-matched with other databases. The Elsevier Scopus database reports the citations dated back to 1996, which is a severe flaw while figuring out the most-cited papers. In contrast, the Google Scholar database counts the citations based on published articles, books, conference proceedings, thesis/dissertations, technical reports, and preprints, which explains the higher citation counts reported in the current study [2].
The current study found a statistically significant correlation of the citation frequency with the age of publication, which is similar to the findings of a previous bibliometric analysis report [2]. Although there was an upward trend of citations received by the classics to the age of publication [125], the trend analysis of the influence of age of publication on the citation density revealed that certain topics after reaching maturity show a decrease in citation density. This change in trend can be also be noticed from the current citation index 2019.
It has been reported that the actual impact of a publication can only be assessed at least two decades after it has been published [2,4,17]. Interestingly, this phenomenon has been observed in the current study as the most number of classics were published in 1999. However, it is noteworthy that with the changing trends of how published work is reviewed, the accessibility of literature has increased multifold, and research from around the world can be remotely reviewed without needing access to archives, libraries, and published paper journals. This debate is backed up by the current study, which observed that 63 classics were published during the last two decades. This finding indicates that in the current era of digital technology, classics might require lesser years to reach their maturity stage.
With the evolution of research, several guidelines have been introduced to fulfill the evergrowing need for organized reporting of observational studies [126], laboratory studies [127], clinical studies [128], or reviews [129]. These guidelines allow the scrutinization of scientific information and improve the quality and transparency of reports. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement is used to report systematic review and meta-analysis mainly focusing on evaluating randomized trials to provide the highest level of evidence. Surprisingly, the current study did not identify any systematic review of literature or meta-analysis, which made it to the list. The title of the study report is another key element which is stressed upon in various guidelines. It is entirely possible that some classics were not identified in the current study owing to how their titles were designed. A title should explicitly describe the methodology of study and key elements which identify the study to allow proper indexing of the article.
Keywords play an essential role in the discoverability of any published article [130]. While searching any specific type of literature, scholars tend to methodically utilize search terms which are generally used in a specific field [131]. In this study, prime examples of such terms are antibiotics, antibacterials, or antimicrobials. However, it was noted that keywords only appeared in articles published after 1995 and more so not mandatorily in every publication. It was noted that even though keywords might have been submitted in the journal database during submission of manuscripts, the published articles did not display the keywords [55,63,109]. These incoherencies make the network analysis of keywords somewhat misleading and inconsistent with the actual data if we only rely on hand-searching. Therefore, the ES database was utilized to retrieve the relevant data to allow a presentable and fair network analysis.

Limitations
Firstly, a large amount of "classic" articles had to be excluded from the list as it was not considered possible to perform the bibliometric analysis of 500 or more articles in the current study. Therefore, the top 100 classics which achieved the maximum citations were selected for the present study. Secondly, the most recently published research papers are at a disadvantage irrespective of their content and quality, since they were outside the time window considered. Under this spectrum, it would not be wrong to say that the real impact of a research article cannot be accurately determined for at least five years post-publication.

Conclusions
This bibliometric analysis of the top 100 classics on antibiotics revealed that the increase in the age of publication positively influenced the citation frequency. Unlike times before 1996, the explosion of access to scientific articles in the current era of digital technology means that classics written more recently might require fewer years to reach their mature stage. In spite of substantial developments and advancements in this field/specialty in recent decades, there is a dearth of systematic reviews and meta-analyses among the top 100 publications. Keywords are the cornerstones of the discoverability of any manuscript and therefore, quality journals and publishers should mandate the inclusion of keywords in every publication to ensure maximum visibility of the publication across all databases.
Supplementary Materials: The following are available online at http://www.mdpi.com/2079-6382/9/5/219/s1, Figure S1: Distribution of citations frequency over last six decades, Table S1: List of journals which published top 100 classics, Table S2: List of keywords identified from the Elsevier Scopus database