Bibliometrics of Latin American research on COVID-19 in the first year of the pandemic: the main trends

This study aimed to identify and analyze the characteristics of COVID-19 publications in the Science Citation Index Expanded published by Latin Americans in 2020. Documents including searching keywords in their title, abstract, or author keywords were assessed. The analyzed aspects covered characteristics of document types, languages, Web of Science categories, and journals. Publication performances of countries and institutions were evaluated by six publication indicators and two citation indicators. Results showed that a lower percentage of articles and a higher percentage of Spanish language were found. Web of Science category of general and internal medicine published the most articles. The Clinics was the most popular journal. The Cadernos de Saude Publica and Revista da Associacao Medica Brasileira published the most articles and reviews, respectively. Brazil took a leading position in the six publication indicators. The University of São Paulo in Brazil was the most productive institution. Based on the number of citations from the Web of Science Core Collection since publication to the end of 2020, 10 most frequently cited publications were presented. In addition, the analysis of words in publication titles, author keywords, and KeyWords Plus was performed to find the main research focuses.

2019 (Hui et al., 2020;Zhou et al., 2020). COVID-19 is not only a recently emerged disease but also a new research topic. The Web of Science has been used for bibliometric study on COVID-19 in the Annals of Translational Medicine , the Journal of Family and Community Medicine (Farooq et al., 2021), the Sao Paulo Medical Journal (de Oliveira et al., 2021), the Profesional de la Informacion (Herrera-Viedma et al., 2020), the Pharmacological Research (Su et al., 2021), and the high impact factor journals . COVID-19 and the emerging research in fields, for example, ophthalmology (Kalra et al., 2021), business and management (Verma & Gustafsson, 2020), environmental studies (Usman & Ho, 2021), endocrinology (Atlasi et al., 2021), and the nursing (Oh & Kim, 2020) was concerned. Other related research was presented on the topic of cancer research in the era of COVID-19 (Van Hemelrijck et al., 2021), the effect of COVID-19 on the environment (Casado-Aranda et al., 2021), the application of artificial intelligence in the COVID-19 pandemic (Islam et al., 2021), the coronavirus during the first eight months of COVID-19 in 2020 (Furstenau et al., 2021), the human mobility behavior in COVID-19 (Benita, 2021), the machine learning approaches in controlling COVID-19 outbreaks (Chiroma et al., 2020), and the co-occurrence of COVID-19 publications (Al-Zaman, 2021). It was reported that a higher percentage of document types of editorial materials were published when SARS suddenly happened (Chiu et al., 2004). COVID-19 editorial materials in the beginning stage of the pandemic were analyzed (López-López et al., 2021). The patterns of international collaboration on COVID-19 (Duan & Xia, 2021;Grammes et al., 2020) were studies. COVID-19 pandemic-related research in Africa (Fonkou et al., 2021), in Iran (Shamsi et al., 2020), and in the Latin American contributions (Gregorio-Chaviano et al., 2020) were also reported. Gregorio-Chaviano et al. (2020) concluded that the results of regional participation in COVID-19 research are not only related to the number and exponential growth but also related to quality and excellence.
Due to the Web of Science being updated from time to time, the bibliometric study results cannot be repeated. We collected data from the Science Citation Index Expanded (SCI-EXPANDED) for publications in 2020 after the Journal Citation Report (JCR) reported the journal impact factors. This study aimed to use bibliometric analysis to study COVID-19 related publications in the SCI-EXPANDED in 2020. Six publication indicators and two citation indicators were applied to evaluate publication performances of countries and institutes. Word analysis was used to find the research's main focuses.

MATERIALS AND METHODS
Data used in this study were retrieved from the Clarivate Analytics Web of Science, the online version of the Science Citation Index Expanded (SCI-EXPANDED) on 25 July 2021. The 2020 journal Impact Factor  ) was reported in the Journal Citation Report (JCR) on June 30, 2021. According to the definition of journal impact factor, it is best to search documents published in 2020 from SCI-EXPANDED after IF 2020 were presented. It is reported that SCI-EXPANDED is mainly designed for researchers to find literature and is not used for bibliometric research (Ho, 2018;Ho, 2021a). Thus, it is always necessary to perform data processing, but directly obtain data from SCI-EXPANDED for bibliometric research. Recently, in extensive bibliometric research, by using the 'front page' (including the paper title, abstract, and author keywords in the paper) (Fu et al., 2012) as a filter resulted in a big difference in bibliometric studies (Ho, 2019a;Ho, 2020). KeyWords Plus provides additional search terms, which are extracted from paper titles cited by authors in their bibliography and footnotes in the ISI (now Clarivate Analytics) database, and greatly enhance the title terms and author author-keyword indexing (Garfield, 1990). It has been pointed out that documents that can only be searched through KeyWords Plus (KP) have nothing to do with the search subject (Fu & Ho, 2015).
In the SCI-EXPANDED, the corresponding author is labeled as reprint author, but in this study, we used the term corresponding author (Ho, 2012). In a single-institute publication where the corresponding author institution is unspecified, the single institute is classified as the first and the corresponding author institution (Ho, 2014). In a single-author publication with multiple affiliations, only the affiliation in the Latin Americans was considered. In multiple corresponding-authors publications, the Latin American corresponding author, institute, and the country as the corresponding author information first and then the last corresponding author, institute, and country being designated as the corresponding author information (Ho, 2019b). To have accuracy analysis results, affiliations originating from England, Scotland, Northern Ireland, Wales, Anguilla, and British Virgin Isl (British Virgin Islands) were reclassified as being from the UK (United Kingdom) (Chiu & Ho, 2005). Affiliations from Curacao and French Guiana were reclassified as being from Netherlands and France, respectively.
Two citation indicators such as the total number of citations from Web of Science Core Collection in 2020 (TC 2020 ) (Chuang et al., 2011;Wang et al., 2011) and the total number of citations per publication (CPP 2020 = TC 2020 / TP) (Ho, 2012) were used to characterize the publications.

Document types and languages:
In recent years, the number of authors per publication (APP), citations per publication (CPP year ), and the number of publications (TP) were proposed to compare characteristics of document types (Monge-Nájera & Ho, 2017). Using the citation indicator of CPP year gives values more accurate Monge-Nájera & Ho, 2015). In 2020, 3 056 COVID-19 related documents by the Latin Americans were found in SCI-EXPANDED within 11 document types indexed in the Web of Science. Table 1 shows that the most common document type was the articles (46 % of the 3 056 documents). Document type of data papers with three documents had the highest CPP 2020 of 20 followed by proceedings papers with CPP 2020 of 17 respectively. The data papers also had the highest APP of 50. A total of 580 letters were published in 260 different journals. The Travel Medicine and Infectious Disease with IF 2020 of 6.211 published the most of 26 letters (4.5 % of the 580 letters) with a CPP 2020 of 7.9 and an APP of 6.3; followed by the Lancet with IF 2020 of 79.321 (18 letters; 3.1 % of the 580 letters) with a CPP 2020 of 8.9 and an APP of 4.0; the World Neurosurgery with IF 2020 of 2.104 (18; 3.1 %) with a CPP 2020 of 0.39 and an APP of 4.4; the Revista Médica de Chile with IF 2020 of 0.553 (17; 2.9 %) with a CPP 2020 of 0.18 and an APP of 2.5; and the Arquivos Brasileiros de Cardiologia with IF 2020 of 2.000 (15; 2.6 %) with a CPP 2020 of 0.20 and an APP of 7.5. A total of 501 editorial materials were published in 238 different journals, mainly in the Clinics with IF 2020 of 2.365 (47 editorial materials; 9.4 % of the 501 editorial materials) with a CPP 2020 of 1.1 and an APP of 5.8; the Arquivos Brasileiros de Cardiologia with IF 2020 of 2.000 (23; 4.6 %) with a CPP 2020 of 1.3 and an APP of 4.0; and the Revista Médica de Chile with IF 2020 of 0.553 (17; 3.4 %) with a CPP 2020 of 0.24 and an APP of 5.2. A total of 480 reviews were published in 274 different journals. Nineteen percent of the 480 reviews were published in the top three productive journals: the Revista da Associação Médica Brasileira with IF 2020 of 1.209 (23 reviews; 4.8 % of the 480 reviews) with a CPP 2020 of 0.70 and an APP of 5.2; the Clinics with IF 2020 of 2.365 (19; 4.0 %) with a CPP 2020 of 2.9 and an APP of 7.4; and the Frontiers in Immunology with IF 2020 of 7.561 (12; 2.5 %) with a CPP 2020 of 4.7 and an APP of 5.9.
There were seven languages in use. English, as the most popular language, comprised 92 % of the 3 056 COVID-19 related documents by the Latin Americans were found in SCI-EXPANDED, followed by Spanish (196 documents; 6.4 % of 3 056 documents) and Portuguese (47 documents), Other languages that were less used as follows, French (2 documents), and one in German, Polish, and Russian, respectively.

Web of Science Category and Journal:
Journal Citation Reports (JCR) indexed 9 531 journals with citation references across 178 Web of Science categories in SCI-EXPANDED in 2020. The 3 056 COVID-19 related documents by the Latin Americans were published by 949 journals among the 138 Web of Science categories in SCI-EXPANDED. All 3 056 COVID-19 related publications by Latin Americans were analyzed for further study.
The top three productive categories were: including general and internal medicine (473 documents; 15 %) with an APP of 7.4 and a CPP 2020 of 7.5; public, environmental, and occupational health (380; 12 %) with an APP of 7.0 and a CPP 2020 of 4.9; and infectious diseases (232; 7.6 %) with an APP of 7.3 and a CPP 2020 of 8.0. Other categories published more than 100 documents were: tropical medicine (147; 4.8 %), clinical neurology (135; 4.4 %), immunology (129; 4.2 %), surgery (127; 4.2 %), environmental sciences (123; 4.0 %), multidisciplinary sciences (119; 3.9 %), research and experimental medicine (110; 3.6 %), cardiac and cardiovascular systems (101; 3.3 %), and psychiatry (101; 3.3 %). One document published in categories of soil science and electrochemistry had the highest APP of 26 and CPP 2020 of 30, respectively. The top ten productive journals published in 40 or more documents were listed in Table  2. Total number of documents, percentages of articles, letters and editorial materials, and reviews, journal impact factor in 2020 (IF 2020 ), number of authors per publication (APP), and citations per publication (CPP 2020 ) were also compared. Clinics with IF 2020 of 2.365 (75 documents; 2.5 % of 3 056 documents) published the most documents including 9.3 % articles, 64 % of letters and editorial materials, and 25 % reviews. Compared to the top ten productive journals, documents published in PLoS One had the highest APP of 10 while documents published in Travel Medicine and Infectious Disease had the highest CPP 2020 of 25 (Table 2). The journal with the highest IF 2020 of 91.245 was New England Journal of Medicine with eight documents including four articles, three letters, and one editorial material, with an APP of 20, and a CPP 2020 of 167, followed by Lancet (IF 2020 = 79.321) with 29 documents including 18 letters, seven editorial materials, and four articles, with an APP of 8.0, and a CPP 2020 of 21, and JAMA-Journal of the American Medical Association (IF 2020 = 56.272) with two articles with an APP of 36, and a CPP 2020 of 64. These journals are the top three journals with IF 2020 in the category of general and internal medicine.

Publication Performances: Countries:
A total of 3 056 COVID-19 related publications by the 34 Latin American countries were found in SCI-EXPANDED, 1 685 publications (55 % of 3 056 publications) were singlecountry publications from 23 Latin American countries and 1 371 (45 %) were internationally collaborative publications from 33 Latin American countries and other 121 countries in the world. Eleven of the 34 countries had no single-country publications. Only Belize had no internationally collaborative publication. Six, seven, and twenty countries had no first author, corresponding author, and first-author publications, respectively. The top 14 productive countries with 20 publications or more are listed in Table 3 with six publication indicators (Ho & Kahn, 2014). Others were Honduras (18 publications), Guatemala (16), Dominican Republic (14), French Guiana (12), Barbados (11), Trinidad and Tobago (9), Jamaica (8),    Table 4 shows the 10 most productive Latin American institutions with 55 or more publications with seven publication indicators (Ho & Fu, 2016) and citation indicator, CPP 2020 of first-author and corresponding-author publications, respectively. Among these 10 institutions, seven institutions were located in Brazil and one in Chile, Colombia, and Mexico, respectively. The Uni-

Publication Performances: Authors:
Publication indicators: total number of publications (TP), number of first-author publications (FP), number of corresponding-author publications (RP), and number of single-author publications (SAP) as well as citation indicators: citations per publication for first-author publications (FP CPP 2020 ) and correspondingauthor publications (RP CPP 2020 ) were proposed to compare publication performance of authors in a research topic (Giannoudis et al., 2021). Table 5 listed the top 16 most productive authors with 10 or more Covid-19 publications with the four publication indicators and the tow citation indicators. A. J. Rodriguez-Morales at the Technological University of Pereira in Colombia published the most Covid-19 publications in 2020, followed distantly by other Latin America authors. Table  6 lists the 10 most frequently cited Covid-19 publications in 2020 by the Latin Americans in SCI-EXPANDED with the citation indicator TC 2020 . All the top ten publications including four articles, three editorial materials, two reviews, and one letter were an international collaboration by 172 institutions from 37 countries.

Research Focuses:
A total of 331 publications (11 % of 3 056 publications), 122 publications (6.9 % of 1 781 publications with author keyword information), and 94 publications (6.3 % of 1 484 publications with Key-Words Plus information) included 'patients', 'coronavirus infections', and 'infection', respectively. The results showed that 'health' including, for example, public health, mental health, telehealth, health care, and health policy and 'infection' including, for example, infections, co-infection, and infection control were the research focuses by the Latin Americans after COVID-19 outbreak for a year. Furthermore, 'pandemic' and 'Brazil' ranked not only in the top 20 words in publication title but also in the top 20 author keywords.

DISCUSSION
In 2020, 3 056 COVID-19 publications in SCI-EXPANDED by Latin Americans were found in eleven document types. The COVID-19 publication pattern in the first year of its outbreak suggests that more letters and editorial materials and fewer articles were published. It has been reported that document types of news, editorial materials, and letters were published more when SARS suddenly happened (Chiu et al., 2004). A higher percentage of Spanish publications was found. The Portuguese language was also used clearly. Three articles entitled "Protocol for a partially nested randomized controlled trial to evaluate the effectiveness of the scleroderma patient-centered intervention network COVID-19 home-isolation activities together (SPIN-CHAT) program to reduce anxiety among at-risk scleroderma patients" (Thombs et al., 2020a); "Inborn errors of type I IFN immunity in patients with life-threatening COVID-19" ; and "Changes in mental health symptoms from pre-COVID-19 to COVID-19 among participants with systemic sclerosis from four countries: A Scleroderma Patient-centered Intervention Network (SPIN) Cohort study" (Thombs et al., 2020b) were published by more than 100 authors with 139, 129, and 101 respectively. An editorial material entitled "Fair allocation of scarce medical resources in the time of Covid-19" (Emanuel et al., 2020) Garfield (2002) noted that a review with 350 citations within one year after publication was the highest citation recorded. A review with 738 citations in the year of publication and 1 189 citations in the year after publication was reported (Ho & Kahn, 2014). Spanish is the main language in Latin Americans and the Portuguese language is mainly used in Brazil that is the most productive country in Latin Americans. Non-English documents had fewer citations, with a CPP 2020 of 0.91, while English documents had a CPP 2020 of 5.7. The non-English document with the highest TC 2020 was an article entitled "Food (in) security in the context of the SARS-CoV-2 pandemic"  published in Cadernos de Saúde Pública in Portuguese, with a TC 2020 of 15. The lower percentage of articles (46 % of all 3 056 documents) compared with medical topics, for example, 69 % in breast reconstruction , 70 % in Ebola (Pouris & Ho, 2016), 75 % in human papillomavirus (Lin et al., 2011), and 89 % in fracture nonunion (Giannoudis et al., 2021) was found. Similarly, it was reported that a lower percentage of articles were published in the beginning stage of the SARS breakout in 2003 (Chiu et al. 2004).
Clinics was the most popular journal. Documents published in Travel Medicine and Infectious Disease had the higher citations per publication. The top two most frequently cited papers were published in the New England Journal of Medicine with the highest IF 2020 of 91.245.
The first author and corresponding author are the two most contributed authors in a publication (Riesenberg & Lundberg, 1990). At the institutional level, the determined institution of the corresponding author might be a home base of the study or origin of the paper (Ho, 2012). Six publication indicators: total number of publications (TP), single-country publications (SP), collaborative publications (CP), firstauthor publications (FP), corresponding-author publications (RP), and single-author publications (SAP) (Ho & Kahn, 2014) were applied to compare publication performances. Brazil ranked top in all six publication indicators followed distantly by other Latin Americans. Domination in Covid-19 publications in the Latin Americans by Brazil was not surprising since this pattern occurs in other medical topics, for example, stroke (Alarcon-Ruiz et al., 2019), neurosciences (Forero et al., 2020), and bariatric surgery (Toro-Huamanchumo et al., 2021). Colombia ranked the top in two citation indicators, CPP 2020 of FP and RP with 7.6 and 9.2, respectively. The USA had not only active collaborative research in Covid-19 with the Latin Americans (730 publications; 53 % of 1 371 internationally collaborative publications) but also joined the top four highly cited Covid-19 publications with the Latin Americans including an article by Beigel et al. (2020) with TC 2020 of 637, an editorial material by Emanuel et al. (2020) with TC 2020 of 561, a review by Rodriguez-Morales et al. (2020) with TC 2020 of 490, and an article by Van Bavel et al. (2020) with TC 2020 of 357. In 2002, Garfield (2002) noted that a review with 350 citations within one year after publication was the highest citation recorded. Furthermore, a review by Lander et al. (2001) had 738 citations in the year of publication (Ho & Kahn, 2014).
The University of São Paulo in Brazil dominated all the six publication indicators while the Technological University of Pereira in Colombia had a higher impact for their first-and corresponding-author publications. Compare the top ten institutes, publications with first authors and corresponding authors from the Technological University of Pereira (Colombia) had the highest FP CPP 2020 of 85 and RP CPP 2020 of 35, respectively. Publications with first authors from the Pontifical Catholic University of Chile (Chile) had much lower FP CPP 2020 of 1.4 and publications with corresponding authors from the Federal University of Minas Gerais (Brazil) had RP CPP 2020 of 1.5.
Rodriguez-Morales also had the most 40 corresponding-author publications. C.D.F. de Souza at Federal University of Alagoas in Brazil had the most 13 first-author publications. Compare the top 16 authors in Table 5, Rodriguez-Morales had the greatest CPP 2020 for his first-author publications with 135 while D.K. Bonilla-Aldana had the greatest CPP 2020 for his corresponding-author publications with 56. When the names of different authors are the same, or an author uses different names in his/ her publications (such as maiden surnames), there may be deviations in authorship analysis (Zhang et al., 2012).
The citation index of a publication might not be a direct measure of its quality or importance; it is a measure of recognition that may suggest its visibility or impact on the scientific community (Furlan & Fehlings, 2006). Total citations (TC) from the Web of Science Core Collection are one of the popular bibliometric indicators used by authors in literature. However, it can be changed over time. Ho proposed a citation indicator of TC year to compare the most cited publications in a field (Ho, 2012;Ho, 2013). The advantage of TC year is that it is invariable and ensures repeatability in comparison to the index of citation from the Web of Science Core Collection (Fu et al., 2012).
USA was the most popular collaborative partner with nine publications, followed by the UK with six publications and Brazil and China with five publications respectively. Eleven Latin American countries published the top ten publications including Brazil with five publications, Mexico with two and one published by each of Argentina, Bolivia, Colombia, French Guiana, Ecuador, Honduras, Panama, Peru, and Venezuela respectively. Ten institutes had two of the top ten publications including the University of Oxford (the UK), Tan Tock Seng Hospital (Singapore), University of Southampton (UK), Harvard University (USA), University of California San Diego (USA), University of Washington (USA), University of Pennsylvania (USA), Peking University (China), Seoul National University Hospital (South Korea), and National Institute of Allergy and Infectious Diseases (NIAID) (USA). International collaborations with authors from outside of the Latin Americans had the top citations.
Analysis of used words in publication titles, abstracts, author keywords, and Key-Words Plus was proposed for the research focuses (Zhang et al., 2010). In SCI-EXPAND-ED, 1 781 (58 % of 3 056 publications) and 1 484 (49 %) publications included author keywords and KeyWords Plus information, respectively. The analysis of words in publication titles, author keywords, and KeyWords Plus revealed that health and infection were the main research focuses.
Ethical statement: the authors declare that they all agree with this publication and made significant contributions; that there is no conflict of interest of any kind; and that we followed all pertinent ethical and legal procedures and requirements. All financial sources are fully and clearly stated in the acknowledgements section. A signed document has been filed in the journal archives.