Choosing the Research Question Overview
We’re working on this section. For now, have a look at:
- As Dave Sackett has been quoted as saying ‘one-third of a trial’s time between the germ of your idea and its publication in the New England Journal of Medicine should be spent fighting about the research question’ (see Riva and colleagues).
- Your research question should match what you think the future users of your trial results need, or better what they have told you they need, and the PRECIS-2 tool might be a useful way of doing that.
Top 5 Choosing the Research Question Tips
01.
The research question drives your trial; mess this up and you might spend five years doing a trial nobody cares about. Give it some time.
02.
The research question should start life in a systematic review, a guideline group, or some other prioritisation process that highlights a gap that needs to be filled.
03.
Trial questions generally have a PICO (Participant–Intervention–Comparator–Outcome) format. The choice of all of these should involve the people whose decisions you are trying to support with your trial results.
04.
Check trial registers (e.g.hclinicaltrials.gov) to see that someone isn’t already doing your trial.
05.
More to come: keep checking the site…
More About Choosing the Research Question
Choosing the research question may seem like a quick and simple task; in reality we need to invest time to really get to grips with, and craft, the research question. A clear and appropriate research question, or sometimes a set of linked questions, provides a foundation for good research. It can take weeks of ironing out the finer details but once you’ve carefully considered your patient population, the intervention you’re testing and its comparator, and you’ve decided on your outcome measures you will be in a position to select the study design that allows you to answer your research question effectively and efficiently. A mistake here can cost you years of wasted effort.
The attrition of trial participants can impact negatively on the trial’s results. Bias can be introduced if participants drop-out as a result of experiencing severe side effects, and the study can be underpowered if attrition rates are substantial.
