# Literature search strategy: databases, Boolean queries, and deduplication

> Chapter 3 of 8 of The Systematic Review E-book by AIPRA. Search design determines what evidence a systematic review can possibly find. Database choice should follow the specialty and scope of the review, and a common recommendation is to search at least two independent bibliographic databases to reduce retrieval bias. Building a query means identifying the core concepts implied by the research question, expanding each concept with synonyms and related terms to maximize sensitivity, then expressing that as a database-specific query using Boolean operators, phrase and proximity rules, and controlled vocabulary such as MeSH in PubMed or Emtree in Embase.

- Source: https://aipra.co/systematic-review-ebook/searching-for-articles
- Topics: systematic review search strategy; Boolean search PubMed; search string systematic review; MeSH Emtree; deduplication references; grey literature searching
- Last updated: 2026-09-05
- Licence: free to read; cite as shown at the end of this file.

## Key points

- Search at least two independent bibliographic databases to reduce retrieval bias; in medicine, PubMed is commonly paired with Embase or Web of Science.
- Query construction is a three-step process: identify core concepts, expand each with synonyms, then translate into each database's native syntax.
- Controlled vocabulary differs by platform — MeSH in PubMed, Emtree in Embase — so a query cannot simply be copy-pasted between databases.
- OpenAlex offers broad multidisciplinary coverage and easier programmatic access, and surfaces substantial grey literature such as preprints, which matters for fast-moving fields.
- Deduplication using DOI and PMID catches most duplicate records, but identifier matching is imperfect; plan to resolve residual duplicates during screening.

## Chapter outline

1. Choosing literature databases
2. Practical selection guidance
3. From concepts to a search query
4. Running searches and importing exports
5. Deduplication across databases

## Questions this chapter answers

### How many databases should a systematic review search?

A common recommendation is at least two independent bibliographic databases, so that indexing gaps in any single source do not bias retrieval. The specific combination should follow the review's specialty and scope, and institutional access often constrains the choice.

### How do I build a Boolean search string for a systematic review?

Start by identifying the main concepts in your research question. Expand each concept with synonyms and related terms that authors might have used. Then combine synonyms within a concept using OR, combine concepts using AND, and add field limits and controlled vocabulary appropriate to each database.

### Why does my PubMed search not work in Embase?

Each database has its own syntax, field tags, and controlled vocabulary. PubMed uses MeSH terms while Embase uses Emtree, and phrase and proximity operators differ. Queries must be translated per database rather than reused verbatim.

### How are duplicate references removed in a systematic review?

Deduplication normally relies on unique identifiers such as DOI and PMID, which resolves the majority of duplicates automatically. Because metadata can differ between exports, a small number of duplicates typically survive and should be removed during title/abstract or full-text screening and reflected in the PRISMA flow diagram.

### Should systematic reviews include grey literature?

Including grey literature such as preprints, theses, and reports can reduce publication bias and matters most in fast-moving fields where peer-reviewed publication lags. Sources like OpenAlex surface a large share of preprints. Whether to include grey literature should be prespecified in the protocol.

## References cited

- Cochrane Handbook for Systematic Reviews of Interventions, Chapter 4 (searching for and selecting studies), section on duplicate reports and deduplication. https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04

## How to cite this chapter

AIPRA. Literature search strategy: databases, Boolean queries, and deduplication. In: The Systematic Review E-book. Updated 2026-09-05. Available from: https://aipra.co/systematic-review-ebook/searching-for-articles