Librarians can assist you with your search, but if you conduct your own extensive sensitive search, here are some recommendations.
- Formulate your question to identify the major concepts. PICO - Patient, Intervention, Comparison, Outcome - is a useful tool.
- Search in as many relevant databases as possible. For clinical topics, we recommend you search at a minimum Ovid Medline (or PubMed), Embase and Cochrane CENTRAL via the Cochrane Library. See the Where To Search tab in this Resource Guide.
- Consider searching the unpublished grey literature which includes clinical trial registries, government and industry reports, dissertations, unpublished practice guidelines, press releases, and conference abstracts. See the Where To Search tab in this Resource Guide.
- Apply as few limits as possible. Common limits often applied but generally not recommended include: year limits (e.g., last 10 years) and language limits (e.g., English Only).
- Use both subject headings where available (e.g., MeSH) and free-text terms (textwords)
- Expand your textword searching (not subject headings) by:
- using similar terms for each concept and combining them with “OR” (e.g., braces OR supports OR orthoses)
- using spelling variations, e.g., orthoses OR orthosis; randomised OR randomized
- searching abbreviations, acronyms as well as their spelled-out versions: EHR OR electronic health records
- truncating or using wild card symbols. For example in PubMed, back pack* includes back pack, back packs, back packer, back packing, and more.
- Document your search process, noting:
- search concepts (e.g., Heart Failure) and terms (heart failure OR congestive heart failure OR CHF). Try using a table or spreadsheet to keep track.
- names of databases and other resources where you searched
- the range of years of the databases (PubMed includes 1946-current)
- dates you conducted the searches (e.g., May 10, 2017 - May 30, 2017) or when you stopped searching ("searches are complete through May 30, 2017")
- the search strategy, including the terms and their combinations (using OR, AND, NOT), and all the sets
- any limits (e.g., date ranges, languages, study designs, age groups) that you applied with justifications
- any citation or systematic review management software utilized (e.g., Endnote, Covidence, RevMan)
- Consider a peer-review of your search strategy. CADTH (Canadian Agency for Drugs and Technologies in Health) developed the PRESS checklist.