Designing a Pilot Evaluation

#Explain the larger study and the pilot objective

The 2016 CONSORT extension addresses randomised pilot and feasibility trials conducted before a future definitive randomised trial. Their primary aim is feasibility of the future trial, not estimating effectiveness in the same way as the definitive study. The extension does not directly apply to internal pilots or non-randomised feasibility studies.

State what future work the pilot informs and which uncertainty it examines. The word pilot alone does not identify the design or purpose. This is guidance for reading and planning questions, not a complete protocol, ethics approval or Mynd trial-design service.

Evidence: CONSORT pilot and feasibility extension (2016) / Eldridge et al. (2016): feasibility and pilot definitions

#Tie measurements and sample size to feasibility

The extension's methodological principles say that measurements and analysis should address each pilot objective. They may include outcomes planned for the definitive trial, but need not. Participant numbers should have a rationale based on feasibility objectives. Formal hypothesis testing for effectiveness is not recommended because that is not the pilot's aim and it will usually be underpowered.

Ask whether the proposed measurements actually answer the preparatory question. A small study with a favourable result is not automatically an effectiveness demonstration, and a non-significant result is not proof of no benefit. No sample-size formula or individual study calculation is supplied here.

Evidence: CONSORT pilot and feasibility extension (2016)

#Plan how the findings will affect the next decision

CONSORT's pilot principles call for a decision process about proceeding, potentially using criteria for proceeding, amending or stopping. NIHR also asks for a clear route to later research. A 2022 study of NIHR funding applications examined how progression criteria were included, but could not establish whether clearly defined criteria led to future definitive trials.

Describe what finding would change the plan and explain the basis for that decision. Do not treat a checklist or threshold as a guarantee of later success. Report unresolved uncertainties and the reasons for changes. These sources concern trial preparation, not automatic permission for routine deployment or a Mynd evaluation already performed.

Evidence: CONSORT pilot and feasibility extension (2016) / NIHR: guidance on feasibility studies / Trials (2022): progression criteria in funding applications

Source note

The sections above were checked against the linked sources. No clinical review has been performed. This is general research education, not a clinical guideline.