Medline and Embase complement each other in literature searches

EDITOR—Greenhalgh briefly mentions some important differences between the Medline and Embase databases.1 To compare the usefulness of the databases in the drug information setting we carried out a prospective study of all requests for drug information that required a literature search during July and August 1996. For each inquiry both databases were searched back to 1992 with current CD ROMs. Standard search strategies were used, based on indexed subject headings and textwords. Relevant articles were selected on the basis of the information in the abstract title or subject headings, or both. When an article seemed to be unique to a database this was verified by cross checking in the other database, using an author or journal name search.

For 11 of 32 questions no relevant information was retrieved by either database, leaving 21 questions for comparison. The same articles were retrieved by both databases for only four questions. Embase retrieved articles that were unique to that database for 14 questions, while Medline retrieved unique articles for two questions. For one question both databases retrieved a unique article.

Embase was able to answer adequately 20 of the 21 questions, whereas Medline answered only 17 adequately; both databases were required to answer all 21 questions. Articles in Embase are assigned more index terms than those in Medline, and consequently people using Embase may be less likely to miss an important article but must spend more time browsing through irrelevant material.

The overlap of journals in the two databases is only about 34%,2 and other studies have shown that using them both improves the coverage of the literature.3^–5 Despite this, Embase seems to be underused. On the assumption that the words “Embase” and “Medline” would be used in many publications to describe the search strategy, we did a textword search of Medline (1993 to August 1997) and retrieved 66 and 1669 hits respectively.

Our study shows some important differences between the databases. Firstly, although Medline adequately answered most of the questions, a considerable number could be answered only by using Embase. Secondly, Embase complements Medline and vice versa. Finally, to retrieve important articles that are not indexed in Medline it is essential to use Embase in searches done for a comprehensive review or to find rare case reports. Embase is more expensive, more time consuming to use, and perhaps less accessible than Medline. We believe, however, that all practitioners searching for information about drugs and therapeutics should use Embase when Medline has not retrieved sufficient information or when more comprehensive coverage of the literature is required.