What Should Be Tested Before the Main Study? Pilot and Feasibility Design

Purpose of a pilot study

A pilot is a small-scale rehearsal of a future definitive study. Its primary question is whether the main study can be conducted as planned—not whether the intervention is effective. Recruitment, consent, randomisation, measurement, adherence, blinding and attrition are typical feasibility outcomes.

Prespecify progression criteria

Define “go”, “modify” and “stop” thresholds before analysing results. Criteria should reflect resources and what the definitive study requires.

Sample size

Pilot sample size should estimate critical feasibility parameters with useful precision. There is no universal rule that 30 participants are sufficient. Do not power a pilot as if it were the definitive effectiveness study.

Avoid unstable effect estimates

Effect estimates from small pilots are imprecise and can be optimistic. Base the definitive sample size on a clinically meaningful target, external evidence and more stable pilot information such as variance and attrition.

Reporting

Use the CONSORT pilot and feasibility extension. Feasibility outcomes should be primary, while clinical outcome analyses are exploratory.

References

• CONSORT pilot and feasibility extension.

• Eldridge SM et al. CONSORT extension. 2016.

• Thabane L et al. Tutorial on pilot studies. 2010.

Accessed 20 June 2026.