How do I know if a research idea is worth starting?

Between you and me, most of us have a folder somewhere with a half-finished draft in it: a spreadsheet last updated two years ago, an abstract, or a poster presented at a meeting but never written up.

This happens to good clinicians with good ideas. It is not a lack of ideas, and it is not a lack of experience. It is a system problem, and it is fixable.

Why projects fail in week one

When a project fails, it usually becomes obvious months later, when we realise the data was never going to take us to submission. But the failure did not happen then. It happened in week one: the decision to start was the decision to fail. Everything in between was an illusion of progress.

The Research Brief exists to catch that decision before it costs months.

Borrowing from the theatre briefing

Before we take a patient to theatre, the surgeons, anaesthetists and theatre nurses run a quick check on the patient, the team and the equipment, to make sure the operation can be done safely before the patient is brought in. A research idea deserves the same check.

The four sections

Fifteen questions, about ten minutes, asked in a strict order. Each section gates the next: if the first answer is no, nothing else matters.

1. Is it worth doing? (four checks)

• Can the question be stated in one sentence? If not, it is still a topic, not a question.
• Have we honestly checked the literature, or are we assuming the field has not answered it? Ten minutes on PubMed is usually enough to tell.
• Does anyone need the answer, even if it adds a new angle or completes a picture?
• Does a real journal in our field publish this kind of question?

2. Can it be answered? (PICO)

• Population with clear inclusion criteria.
• Intervention or exposure that can be measured, not only described.
• Comparator: if there is none, be honest that the study is descriptive.
• Outcome already captured in the data, not something we wish we had recorded.

3. Can we get the data? (feasibility)

• Access to the data source is confirmed.
• A pilot pull of ten records shows the fields we need exist in a usable form.
• The sample size is sufficient.
• The data can be extracted within about two weeks.

Fail any of these and we stop. This is a red line.

4. Can we finish it? (three checks)

• The ethics route is confirmed, or the project is covered under audit or service evaluation, with a realistic timeline.
• At least one colleague in the clinical area agrees the question is worth answering.
• Four protected hours a week are locked in for the next six weeks: not stolen between clinics.

The decision

• 15 out of 15: proceed. Every gate is cleared. Commit the hours.
• 12 to 14: pause. The idea is close, but something is broken, perhaps no comparator or an unclear data collection process. Fix that gap before starting.
• Under 12: let it go. The hardest one. The brief has just shown, in ten minutes, what would otherwise have taken months to discover.

If an idea fails the brief today, that is the brief working: you have just saved yourself months. If it passes all fifteen, you have the green light. The next step is writing it efficiently; see the lesson on why not to write the introduction first.

You can run the brief live and get a decision on the spot, or download the one-page version.