Posters · Glossary
What is a pilot study?
A pilot study is a small scale version of a planned larger study, or part of it, run first to find out whether the larger study can be done, should be done and how. It tests recruitment, consent, retention, the intervention and the measures, rather than whether the treatment works.
Pilot studies save money and participants by exposing problems before a full trial commits to them. They are also commonly misreported as small efficacy trials, which is exactly what the CONSORT extension for pilot and feasibility trials was written to stop.
Indunil Asanka · Co-founder
4 min read · Published
| Feasibility study | Pilot study | Definitive trial | |
|---|---|---|---|
| Main question | Can this future study be done, should it be, and how | The same, by running the future study or part of it in miniature | Does the intervention work |
| Design | Any design, including surveys and interviews | A smaller version of the planned study, randomised or not | Full scale, usually randomised |
| Sample size based on | Precision of feasibility estimates | Precision of feasibility estimates | Power to detect a meaningful effect |
| Typical outcomes | Willingness, acceptability, recruitment sources | Recruitment rate, retention, adherence, data completeness | Primary clinical outcome |
| Ends with | A decision on whether a pilot or trial is worth planning | Progression criteria: go, go with changes, or stop | An estimate of the effect and its precision |
Pilot and feasibility, defined
Eldridge and colleagues, who led the CONSORT extension published in 2016, built the definitions that most funders now use. A feasibility study asks whether a future study can be done, should be done and, if so, how. A pilot study is a subset of feasibility studies with one extra feature: the future study, or part of it, is actually conducted on a smaller scale. So every pilot is a feasibility study, but a survey of clinicians about whether they would recruit patients is a feasibility study without being a pilot. Thabane and colleagues group the reasons for piloting into four kinds: process, resources, management and scientific.
What a pilot measures
The useful outcomes are the ones the main trial depends on: how many eligible people are found each month, how many consent, how many stay to the end, how many receive the intervention as intended, how complete the outcome data are and how long each step takes. A pilot can also estimate the variability of the main outcome to inform the sample size of the full trial. What it should not do is test whether the intervention works, because it is not designed or powered for that, and a promising or disappointing effect from 40 people is mostly noise.
Progression criteria
A pilot ends with a decision, so the CONSORT extension asks authors to report any criteria set in advance for deciding whether to proceed, proceed with changes or not proceed. They can be written as traffic lights, for example green if recruitment reaches 80 percent of target, amber between 50 and 79 percent with a plan to fix it, red below 50 percent. Setting them before the data arrive protects the team from talking themselves into a trial that the numbers do not support, and it gives funders a clear reason to fund the next stage.
Common mistakes
Calling a small, underpowered trial a pilot after the fact because the result was not significant. Reporting p values for effectiveness outcomes as the main finding. Setting no progression criteria. Choosing the pilot sample size by convention without saying what precision it gives for the feasibility estimates. Skipping the participant flow diagram, which is where recruitment and retention problems show. And leaving out the changes to the main trial protocol, which are the real result of most pilots.
Presenting a pilot on a poster board
A pilot poster is mostly numbers against targets, which suits the key number blocks on 26 of the 40 layouts and the protocol blocks on 38. Ten boards carry a CONSORT style flow block for the participant flow from screening to analysis. Put the progression criteria and their results in the brief as plain text, along with the recruitment, retention and adherence figures, because the generator fills each block from the numbers it is given and leaves out what it is not told.
Questions people ask
How many participants does a pilot study need?
It depends on what the pilot must estimate. Sample sizes are usually set so that feasibility measures, such as a retention rate, are estimated with acceptable precision, or so the variability of the main outcome can be estimated for the full trial. Say how the number was chosen; a figure picked only because it is common is a weak justification.
Can I publish a pilot study?
Yes, and the CONSORT extension exists so that pilot and feasibility trials are reported well. Several journals specialise in them, including Pilot and Feasibility Studies, where the extension was published alongside the BMJ. Report it as a pilot, with feasibility outcomes first and any effect estimates labelled as exploratory.
Is a pilot study the same as a phase 1 trial?
No. A phase 1 trial is a first in human study of safety and dose for a new drug. A pilot study tests the methods of a planned study, and it can come before a trial of any phase or a non drug intervention such as an exercise programme. Some phase 2 studies include feasibility aims, but the terms are not interchangeable.
Do pilot participants count in the main trial?
Usually not, especially if the protocol changes after the pilot. An internal pilot, built into the start of the main trial with no break and no change in protocol, is the exception: its participants can be included in the final analysis if that was planned. External pilots are analysed and reported separately.
Does a pilot study need ethics approval?
Yes, if it involves people, their data or their tissue in the way the main study would. A pilot is research, not a rehearsal, and the CONSORT extension asks for the ethics approval reference to be reported. Participants should be told they are in a pilot and what that means for them.
Make one with posters
The button opens the generator with this use case already described. Change the wording to match your own.
Create a poster with OneCraftRelated questions
- What is an e-poster?An e-poster is a conference poster shown on a screen instead of printed. The usual 16:9 landscape PDF rules, file limits, narration and design for screens.
- What is a hazard ratio?A hazard ratio compares how fast an event happens in two groups over time. What HR 0.6 or 1.5 means, how to read its interval, and how to report it on a poster.
- What is a Kaplan-Meier curve?A Kaplan-Meier curve plots the share of people still event free over time. How to read its steps, censoring ticks, median and the numbers at risk table.
- What is an RCT (randomised controlled trial)?What is an RCT: a study that randomly allocates people to an intervention or a control so groups differ only by chance. How it limits bias, and what it costs.
Step by step in the builder: Choose the template, theme and size.
Written and checked by the OneCraft team. Last checked .