oa Health and biomedicine in e-Marefa and IMEMR databases: Comparative study of their content and subject analysis

e-Marefa is an Arab electronic database managed by the Knowledge World Company for Digital Content. e-Marefa Database includes the full text of journal articles and other types of publications from the Arab countries. The database covers health and biomedicine in addition to science, engineering, humanities and social sciences. It includes a total of 250,000 records (articles) of which 5,6494 are in health and biomedicine. The Database was launched in December 2011 and is offered for subscription. IMEMR (Index Medicus for the Eastern Mediterranean Region) is one of the major projects of the Virtual Health Sciences Library (VHSL), of the World Health Organization Regional Office for the Eastern Mediterranean (EMRO). The vast majority of the journals of the IMEMR are neither indexed nor abstracted in international indexing and abstracting services. The Index currently indexes 612 health and biomedical journals and includes 164,864 records from the 22 countries of the Region of which 101,252 records are from Arab countries. The database was established in 1987 and is offered free of charge. Study sample: A set of 200 articles common between the two data bases was identified, retrieved and a database of these was created for analysis. The 200 records were published in 14 journals from countries commonly covered by the two databases. Results: The two databases use international standards and tools for managing bibliographic data and organizing knowledge. They both use the Dublin Core (based on MARC 21 and UNESCO CCF) for metadata, the Anglo-American Cataloguing Rules, NLM Subject Headings and LC Subject Headings. The IMEMR uses more indexing terms than the e-Marefa. The average number of terms used in IMEMR is 6.31 while it is 2.5 terms per article in e-Marefa. e-Marefa Database does not use the NLM publication types as one of its indexing terms. IMEMR uses post-coordinated terms based on the NLM Subject Headings, while e-Mareaf uses pre-coordinated terms from the LC Subject Headings enriched by Arabic local headings. There was high level of compatibility between the two databases as far as using the MeSH terms used as there was minor discrepancy between the two databases in the selection of main subject headings assigned to articles. The study identified the top MeSH terms used by each of the two databases and the common terms used by both. The top five journals contributing to the 200 articles are from Egypt, Oman and Qatar. Recommended future actions include urging the e-Marefa Database to increase the number of terms assigned per article and apply a post-coordination indexing and search methodology to allow for more findability and accessibility. IMEMR can benefit from providing multilingual indexing and searching facility. The two databases may collaborate to provide the full bibliographic data and full text in a public-private partnership model. Recommended future research should focus on measuring recall and precision based on the information retrieval systems used by the two databases. The paper includes four annexes of countries covered in the two databases, the descriptors/ subject headings used, discrepancies in the main headings used and the source journals.