a survey on traditional overlap, relative overlap and synthetic degrees of freedom between pubmed and scopus in the cardiovascular disease field

introduction: determining the overlap rates is a quantitative method to compare database contents. in this survey, traditional overlap, relative overlap and synthetic degrees of freedom between pubmed as a free database and scopus as a subscription-based one in the cardiovascular disease field have been determined. methods: this descriptive applied research was done during january to june 2009. first, the descriptors related to three cardiovascular diseases, namely stroke, high blood pressure and cerebrovascular accidents, were obtained from emtree thesaurus. then, original articles on the three mentioned problems were retrieved from pubmed and scopus by searching these descriptors. primary raw data needed for overlap calculation were obtained using reference manager software. data analysis was done in microsoft excel. results: a total of 9899 articles on cardiovascular diseases were retrieved from both databases, including 8593 and 7713 articles from pubmed and scopus, respectively. there were 6407 common titles while 2186 and 1306 titles belonged only to pubmed and scopus, respectively. the traditional overlaps were 68%, 66.7% and 60.7% for high blood pressure, cerebrovascular accidents and stroke, respectively. relative overlap between pubmed and scopus for stroke, high blood pressure and cerebrovascular accidents were 75.8% vs. 75.3%, 89.2% vs. 74.1% and 86.7% vs. 74.3%, respectively. synthetic degrees of freedom were smaller than 0.5 for the three considered diseases. conclusion: due to pubmed's higher relative overlaps versus scopus's, more user and librarians are led to use pubmed. larger numbers of retrieved unique articles for each cardiovascular disease from pubmed as compared with scopus, along the fact that it is free, makes pubmed preferable.