PubMed vs. Embase vs. Scopus: Choosing Your Search Databases

Searching only PubMed is one of the most common shortcuts that gets a systematic review flagged during peer review — different databases index different journals, and no single one covers everything relevant to most topics.

PubMed / MEDLINE

Free, biomedical and life sciences focused, uses MeSH controlled vocabulary. A near-universal starting point for health research, but not comprehensive on its own — it under-indexes some non-US and non-English-language journals.

Embase

Subscription-based, strong European and pharmaceutical/drug-related coverage, uses its own Emtree vocabulary. Includes many journals not indexed in PubMed, and is often specifically required for Cochrane and health technology assessment reviews.

Scopus

Broad multidisciplinary coverage (not just biomedical), strong for citation tracking and forward/backward reference searching. Useful as a supplementary database, especially outside pure clinical medicine.

Other commonly used databases

• CINAHL — nursing and allied health, essential for nursing systematic reviews.
• PsycINFO — psychology and behavioral science.
• Cochrane CENTRAL — specifically for randomized controlled trials.
• Web of Science — broad multidisciplinary, strong citation indexing.

How many databases do you actually need?

Cochrane and most methodological guidance recommend searching at least two or three databases relevant to your field, plus a trial registry for clinical questions and a grey literature search where publication bias is a concern. The right combination depends on your topic — a nursing review needs CINAHL; a drug-efficacy review needs Embase.