How to Generate Research Ideas
Good projects almost always start with a good question. This guide walks you through how to spot researchable problems, turn them into clear questions, and choose ideas that are realistic for your stage.
Step 1: Start from real problems
The best ideas usually come from everyday clinical practice, not from trying to “be original” in abstract. Look for things that make you think:
- “Why do we do it this way?”
- “This seems inconsistent between teams or hospitals.”
- “We’re not very good at <x> – could it be improved?”
Jot these thoughts down during placements, ward rounds, clinics, MDTs, or teaching sessions. Don’t judge the ideas yet – just capture them.
Step 2: Use patients, guidelines and audits as prompts
Patients and cases
Notice patterns: delayed diagnoses, frequent readmissions, common prescribing errors, or recurrent practical problems for patients or staff.
Guidelines and local practice
Compare local practice to NICE / national guidelines and trust protocols. Gaps between “what we should do” and “what actually happens” are classic starting points for audits and QI projects.
Previous audits and projects
Ask your department if there are old audits that need re-auditing, or projects that never got written up. Improving or repeating an existing project is often more realistic than starting from scratch.
Step 3: Scan the literature for gaps
Once you have a rough topic (e.g. VTE prophylaxis on your ward), do a quick search on PubMed or guidelines:
- What has already been studied?
- Are there controversies, conflicting results, or unanswered questions?
- Are most studies from settings very different to yours?
You’re looking for places where the evidence is thin, outdated, or not applied well in real life.
Step 4: Turn problems into clear questions (PICO)
Use PICO to shape an idea into something researchable:
- Population: who exactly (e.g. adult inpatients on the respiratory ward)?
- Intervention: what change, exposure or tool (e.g. new VTE checklist)?
- Comparison: current practice, previous year, another ward?
- Outcome: what you’ll measure (e.g. % appropriately prescribed LMWH)?
If you can’t define at least Population and Outcome clearly, the idea probably needs more work.
Step 5: Check feasibility
A brilliant idea that you can’t finish is less useful than a simple project that gets completed and presented. Sense-check:
- Can the data be collected from notes / EPR in the time you have?
- Do you need ethics approval, or is it an audit / service evaluation?
- Will you have support from a supervisor who can access data and sign forms?
- Can you realistically analyse the data with your current skills (or help)?
Step 6: Shortlist and get feedback
Pick 2–3 ideas you’re most interested in. For each, write a one-line PICO question plus 3–4 bullet points on why it matters and how you would do it.
Take this mini-list to a consultant, registrar, or academic at your own university or hospital. People are much more likely to help if you arrive with specific, thought-through options rather than “do you have any projects?”.
Step 7: Capture ideas continuously
Keep a simple “ideas list” on your phone or laptop. Many good projects come from revisiting half-formed thoughts after you’ve seen the same problem a few more times. Add:
- Brief description of the issue.
- Where you saw it (ward, clinic, specialty).
- Any guideline or paper you think links to it.
Final point
Remember that not every research project has to be the next ground breaking discovery. Science is built off lots of small steps that lead to greater change, audits, prospective analysis and literature reviews are a great way to help contribute to the wider scientific community!