Comparison of Using Scopus and PubMed for Compiling Publications with Siriraj Affiliation

Abstract

Objective: To compare the commercial citation database, Scopus, with the open access, PubMed, in compiling publications with Siriraj Affiliation. Methods: The affiliation term “Siriraj” was used as a keyword for searching in Scopus and PubMed. Two programs, developed in Python, were used to extract information from each database’s outputs. ISSN (International Standard Serial Number) was used to determine the overlapping sets of journal titles. Finally, articles comparison with multiple algorithms in the overlapping sets of journals was performed. Results: The total numbers of articles with the affiliation term “Siriraj” as retrieved on May 2, 2008 were 3,315 from Scopus and 2,040 from PubMed. Determined by ISSN, 48 journals (54 articles) were unique in PubMed, and 299 journals (484 articles) in Scopus. The number of common journal titles was 329. However, the numbers of articles in the common journal titles were different. There were 1,600 common articles (54%), 1,049 Scopus-unique articles (35%), and 315 PubMed-unique articles (11%). The combination of articles from both databases resulted in 3,502 articles, which were 369 articles (10.5%) increase from the total articles retrieved from Scopus alone. Conclusion: The Scopus database collected data from a wider range of journal titles and contained more articles especially from the earlier period (year 1967-1986). However, PubMed contained some unique journal titles and articles, which were not included in Scopus. This study showed that the searching affiliation term “Siriraj” in Scopus did not cover all publications with Siriraj Affiliation. The combination of articles retrieved from both databases resulted in more coverage.