How-to guide

How to Find Research Papers

Use this guide to move from “I don’t know where to start” to a systematic way of finding the right papers for essays, audits, presentations, and projects.

Step 1: Define what you’re actually looking for

Turn your idea into a clear question

Before you open any database, translate your vague idea into a focused question. You can use a simple framework such as PICO:

  • Population: Who is the group (e.g. adults with type 2 diabetes)?
  • Intervention / exposure: What is being done or studied (e.g. GLP-1 agonist)?
  • Comparison: Usual care, placebo, or another treatment?
  • Outcome: What matters (e.g. HbA1c, mortality, QoL)?

Having this written down stops you from getting lost in irrelevant results and makes it easier to build search terms.

Step 2: Choose the right places to search

Core clinical databases

For most medical questions, start with:

  • PubMed / MEDLINE: Core biomedical literature, good for almost any clinical topic.
  • Cochrane Library: High-quality systematic reviews and meta-analyses.
  • Embase (if available): Particularly strong for pharmacology and European journals.

Broader and grey-literature searches

To widen your net:

  • Google Scholar: Quick way to see a broad range of articles and citation counts.
  • Preprint servers: For very recent work (e.g. medRxiv), but interpret cautiously.
  • Local resources: University library portals, hospital library services, guideline websites, and specialty society pages.

Step 3: Build a search strategy (keywords + subject terms)

List your key concepts

Take your PICO and list synonyms for each part. For example, “heart attack” might also appear as “myocardial infarction” or “acute coronary syndrome”.

Combine terms using Boolean operators

Use simple logic to control your search:

  • AND narrows: breast cancer AND trastuzumab
  • OR broadens: myocardial infarction OR heart attack
  • NOT excludes: stroke NOT haemorrhagic

Use subject headings (e.g. MeSH)

Many databases use controlled vocabulary (like MeSH in PubMed). Look up the subject heading for your topic and combine it with free-text keywords to catch variations in terminology.

Step 4: Apply sensible filters

Time frame and study type

Use filters to focus on what you actually need:

  • Limit by publication date for rapidly evolving fields (e.g. last 5–10 years).
  • Filter by study design: RCTs, cohort studies, systematic reviews, guidelines.
  • Consider language filters if you can only read in one language.

Population filters

Apply filters for adults, children, humans, or specific age groups where available, so your results match the population you care about.

Step 5: Screen the results efficiently

Scan titles and abstracts first

Don’t read everything in full. Start with:

  • Discard clearly irrelevant titles straight away.
  • Read abstracts for “maybes” to decide if they genuinely answer your question.
  • Prioritise higher-quality designs (RCTs, large cohorts, systematic reviews).

Use citation chasing

Once you find a really relevant paper:

  • Look at the reference list for earlier key studies.
  • Use “cited by” features to see more recent papers that referenced it.

Step 6: Access and organise full-text papers

Getting full text

Use your university or hospital library access where possible. If an article is behind a paywall, options include:

  • Institutional log-ins (Athens/Shibboleth/university portal).
  • Checking if there is an open-access version via the journal or repository.
  • Emailing the corresponding author politely to request a copy.

Keeping everything organised

Don’t rely on your Downloads folder. Consider:

  • Saving PDFs into clearly named folders (e.g. “Breast cancer – HER2 – adjuvant”).
  • Using reference managers (e.g. Zotero, Mendeley, EndNote) to store and cite papers.
  • Adding brief notes or tags (e.g. “good methods”, “key RCT”, “guideline-changing”).

Step 7: Check you’ve covered the basics

Before you stop searching, ask:

  • Have I found at least one good systematic review or major guideline?
  • Do I have the main landmark trials or pivotal studies?
  • Are there any obvious gaps in populations, outcomes, or time periods?

If yes, you probably have enough to start writing your summary, presentation, or project – you can always return for more targeted searches as new questions appear.