Insights into bibliometric analyses

The past decade has witnessed an exponential increase in scientific articles presenting bibliometric analyses, indicating the growing significance of understanding and measuring the structure and dynamics of academic knowledge. These analyses provide scientific metrics on the publications, which is necessary in order to grasp a holistic view of a domain, as this becomes more difficult considering the expansiveness of academic productions. Bibliometry, first used in French as “bibliometrie” in 1934, is defined as “a coordinated set of measures relating to the books and the documents. […] Bibliometry summarizes statistics and provides comparison indices”1 (author’s translation). Currently, bibliometric analyses refers to methodological approaches based on statistics and mathematics, usually completed with visual diagrams or maps, performed on scholarly literature. These analyses provide objective data on research hotspots; academic collaborations between authors, institutions, and countries; knowledge diffusion; and the interconnections between research fields and subfields. Moreover, they aid researchers and clinicians in identifying emerging topics and practices, monitoring the evolution of keywords used on specific subjects, assessing the frontiers in the field and gaps in the literature, and proposing directions for future research. Consequently, bibliometrics offers valuable insights into the rapidly evolving scientific landscape and has been used in various disciplines. The recent interest in bibliometric analyses reflects the joint development of rapid calculation methods and computer tools enabling researchers to map networks of connections. A simplistic approach for examining bibliographic information, such as focusing on authors, affiliations, and publication year, can be taken using Web of Science. However, articles on specific topics are often indexed on different platforms, so citation searching on a single platform is insufficient to obtain an overview of research in the field. Software programs specifically designed for bibliometric analyses differ in several respects, including ease of use, customization capabilities, imported data, and some outputs. The major and most commonly used software packages are VOSviewer, Bibliometrix, and CiteSpace, which are free and open source. VOSviewer2 can be downloaded or used online and is simple to use. Complementary to this package is Bibliometrix, which is a package within the statistical program R.3 As with R, using Bibliometrix requires slight computational competencies, although another application with a web interface, Biblioshiny, has been designed to assist with this. VOSviewer and Bibliometrix provide results in the form of network visualization maps for word co-occurrence (conceptual structure), co-citation (intellectual structure), co-authorship (social structure), and thematic evolution.3,4 CiteSpace5 is more focused on monitoring and mapping dynamic landscapes of knowledge and identification of emergent research areas. Bibliometric analyses are also called bibliometric reviews. As for other kinds of reviews, such as systematic reviews, scoping reviews, or diagnostic test accuracy reviews, their conduct must be rigorous, reproducible, and transparent. To this end, Montazeri et al.6 have released a reporting guideline for bibliometric reviews of the biomedical literature (BIBLIO), based on guidelines for other reviews. A review can be conducted with bibliometry as the sole methodological approach. For instance, a recent article using bibliometrics mapped the links between articles, words, and authors on the subject of barriers and facilitators to implementing guidelines in health care.7 The authors identified topic hotspots, such as public health, evidence-based medicine, and quality promotion, as well as promising research directions.7 Alternatively, some authors use these bibliometric data analyses in addition to other types of reviews; for example, both bibliometrics and scoping review approaches were employed in a scoping review aiming to provide a comprehensive view of the positive aspects of caregiving in dementia.8 Similarly, bibliometry has been performed in conjunction with a systematic review on the effect of teachers’ behaviors (eg, gestures, expressions) on learning outcomes9 and in conjunction with an umbrella review.10 For authors conducting bibliometric analyses, several challenges can complicate the interpretation and accuracy of their findings. First, as for other reviews, whether systematic, scoping, or otherwise, results depend on the bibliographic databases used. Databases such as Web of Science, Scopus, or PubMed vary in coverage and indexing practices, potentially leading to incomplete or biased datasets. Second, some problems exist in review typology where the terms scoping review or systematic review may be inappropriately used.11 In this context, the terminology bibliometric review may make things more confusing. Indeed, bibliometric analyses are sometimes called scoping reviews or systematic reviews (eg, in the review by Tzitiridou-Chatzopoulou and Zournatzidou12). Another challenge relates to confusion over terminology. Indeed, in reviews, the term bibliometric database is sometimes used instead of bibliographic database in which the scientific literature is searched. Finally, a more general challenge is that of open access, essential for improving the completeness of bibliometric analyses, which rely on bibliographic metadata. By making scientific publications freely available, open access promotes more equitable research practices, as it enables researchers from underfunded institutions or developing regions to participate more actively in global scientific endeavors. In addition, open access promotes the visibility of research, which can improve citation rates, and increases the availability of reference lists, leading to more accurate assessments of scientific impact and connections. While bibliometric analyses enable researchers to gain a holistic overview of a field and identify potential intra- or inter-professional collaborators, they also provide precious insights for other professionals. Funding agencies and policymakers can also benefit from bibliometric data by identifying research strengths and ensuring that investments are, for instance, aligned with innovative areas or research gaps. These analyses should additionally help public health leaders or academic institutions shape science policies and guide decisions on the promotion of research fields, whether at the national, regional, or institutional level. In today’s highly interconnected and interdisciplinary research landscape, fostered by remote working, the need for bibliometric analyses to understand research dynamics and organize and interpret vast quantities of data will certainly continue to grow. However, bibliometric analyses do not replace other forms of reviews (eg, scoping and systematic reviews), and they should be seen as a key methodological tool with a complementary perspective. Sandrine Ding School of Health Sciences (HESAV), HES-SO University of Applied Sciences and Arts, Western Switzerland,SwitzerlandBureau d’Echange des Savoirs pour des praTiques exemplaires de soins (BEST): A JBI Center of Excellence, SwitzerlandThe author declares no conflict of interest.Corresponding author: Sandrine Ding [email protected]DOI: 10.11124/JBIES-24-00522