Section 1.17 min read

The Research Database Landscape

Core summary

No single database indexes every published paper. Understanding which databases exist, what they cover, and when to use each one is the first step in any literature search.

Detailed explanation

Think of research databases as libraries — each one has a different collection. If you only visit one library, you will miss books that only exist in the others. Here are the major biomedical databases: PubMed: Free, maintained by the U.S. National Library of Medicine. Contains over 36 million citations from MEDLINE and life science journals. This is the starting point for most clinical researchers. Its controlled vocabulary (MeSH terms) makes precise searching possible. Scopus: The largest abstract and citation database, covering over 27,000 journals across all disciplines. Excellent for citation tracking and finding research outside medicine (engineering, social sciences). Subscription required. Web of Science: Another major multidisciplinary database with strong citation analysis tools. Particularly useful for tracking how often a paper has been cited and by whom. Cochrane Library: The gold standard for systematic reviews. Contains the Cochrane Database of Systematic Reviews, CENTRAL (clinical trial registry), and other evidence synthesis databases. Embase: Strong coverage of pharmacology, drug research, and European journals. Overlap with PubMed is about 60%, meaning 40% of Embase content is unique. Essential for systematic reviews. Other important databases include CINAHL (nursing and allied health), PsycINFO (psychology), and regional databases like LILACS (Latin America) and IMEMR (Eastern Mediterranean). For a standard literature search, PubMed alone catches roughly 65-70% of relevant studies. For a systematic review, you need at least 2-3 databases to meet PRISMA standards.

Clinical example

Dr. Layla is planning a systematic review on acupuncture for chronic lower back pain. If she searches only PubMed, she would miss approximately 30% of relevant trials indexed exclusively in Embase, CINAHL, or AMED (a complementary medicine database). Her search must span at least four databases to be comprehensive.

Research example

A landmark study by Suarez-Almazor et al. found that searching only MEDLINE retrieved 55% of eligible RCTs for a musculoskeletal systematic review. Adding Embase increased retrieval to 80%, and adding hand searching of key journals brought it to 92%.

Knowledge check

Q1. What approximate percentage of relevant studies does PubMed alone typically capture?

Q2. Which database is considered the gold standard for finding systematic reviews?

Q3. What is the approximate overlap between PubMed and Embase content?