Finding research means locating studies that answer a question and obtaining enough of their methods and results to judge them. A search result, abstract, or confident summary is a starting point. It is not a substitute for the paper when a consequential decision depends on details the abstract omits.
PubMed provides citations and abstracts and often links to full text. Access to a citation does not mean the full article is freely available. Institutional libraries may offer additional databases and document access; the best search route depends on the question and available resources.
Search concepts, not just a sentence
Break the question into major concepts. For an illustrative question about discharge education and understanding, try terms for discharge or transition from hospital together with education or teaching. Inspect relevant records to see which words and subject headings their authors and indexers use.
PubMed supports Medical Subject Headings, known as MeSH, alongside words in titles and abstracts. Its help explains automatic term mapping and ways to inspect how a search was translated. Check that translation when results seem unexpectedly narrow or unrelated. A familiar abbreviation can mean something different in another field.
A practical search log should include the database, exact search, date, filters, and number of results. This record is particularly useful when a colleague needs to reproduce the search or update it later.
Avoid convenient exclusions
A free-full-text filter answers an access question, not an evidence-quality question. Restricting results to easily available articles can leave out relevant studies. Similarly, a recent-publication filter may miss a foundational study, while an old article may no longer reflect current practice. Use date limits deliberately and explain them.
When a study looks important, seek the complete report, its supplements, and linked corrections. Multiple publications can describe the same study. Do not count an initial paper and its follow-up report as independent replications without checking the participants and trial identifiers.
An illustrative search handoff
A unit council asks whether a teaching approach improves understanding after discharge. A helpful librarian handoff would include the population, proposed approach, comparator, desired outcome and follow-up, a few relevant papers already found, and the decision deadline. It would also identify whether the council needs a quick background search or a formal, reproducible evidence review.
After retrieval, the nurse might discover that a promising article studied student learning rather than patient understanding. That record can be excluded with a reason. Its similar vocabulary does not make it direct evidence for the ward's question.
Save a usable result
Store a citation and stable link with a brief note explaining relevance and limitations. Separate “not found,” “abstract only,” “full text read,” and “appraised.” These labels prevent a bibliography from being mistaken for a completed evidence assessment.
Continue with systematic reviews for nurses when considering an existing synthesis.
If a search turns up an unfamiliar publication, review how to assess a journal before relying on it.
Sources
- National Library of Medicine: PubMed User Guide (accessed 2026-09-17).
- Cochrane Handbook, Chapter 4: Searching for and selecting studies (accessed 2026-09-17).