Co-authorship network analysis of Iranian researchers on osteoporosis.

Abstract

This study aims to investigate the pattern and characteristics of cooperation between Iranian researchers in the field of osteoporosis through the co-authorship social network analysis. All papers on osteoporosis with at least one Iranian author were retrieved from medical databases of PubMed, Scopus, and Web of Science from 2009 to 2019. After the removal of duplicates, the title and abstract of the papers were reviewed by two independent reviewers, and screening was performed according to the inclusion and exclusion criteria. The data were entered into the BibExecl software, and the different spelling forms in the author names were manually merged. The authors' co-occurrence matrix was then developed and entered into the UCINET software and the cohesion indexes (density, diameter, and average distance) and centrality indexes (degree, betweenness, closeness, and eigenvector) for the co-authorship network were estimated. The institutions and countries of the authors of the entered papers were also used in the network analysis and their socio-graphs were drawn. We used 1328 documents in the analysis. The co-authorship network was constructed only for authors with at least 5 papers. The network had 183 nodes (authors) in 13 clusters. Its density was 0.063 and its number of components was 2. The large component encompassed 95.6% and the small component 4.4% of authors. The average distance in the main component of the network was 3 and its diameter was 6. Larijani B was ranked first in the network in terms of degree, betweenness, closeness, and eigenvector centrality. In terms of the contribution of organizations in osteoporosis publications, Tehran University of Medical Sciences (with 35.5% of papers published in WoS database), Shahid Beheshti University of Medical Sciences (14.7%), and Shiraz University of Medical Sciences (9.3%) retained in the first to third place, respectively. In papers published on the WoS, most Iranian authors have collaborated with authors from the USA, Belgium, Canada, and the UK respectively. The co-authorship network had low cohesion with slow information flow between its members. The collaboration with young researchers by the network's active, efficient, and broker authors will lead to the maintenance and development of the network.