Protocols and Prespecified Analysis

#A protocol records what is planned

SPIRIT 2025 describes the randomized-trial protocol as a foundation for planning, conduct, reporting and external review. It provides minimum items to address in a trial protocol. Cochrane likewise describes a review protocol as a record of the intended question and methods before study results are fully known.

Read the plan alongside the final report. It helps distinguish what was intended from what was eventually done, but its existence does not prove that the design was appropriate or that the work followed it. Trials and reviews need different protocol content.

Evidence: SPIRIT 2025: reporting randomized-trial protocols / Cochrane Handbook: reporting protocols and completed reviews

#Advance choices limit hidden flexibility

Cochrane calls for planning searches, risk-of-bias assessment and synthesis methods in advance. It also calls for specifying a limited, reasoned set of subgroup analyses and choosing important outcomes before seeing which results are largest or most available.

Read whether key decisions were fixed before the findings could influence them. Prespecification helps make selective analysis or presentation easier to detect; it does not remove every source of bias or forbid useful exploratory work. An analysis developed after examining data needs to be described as such.

Evidence: Cochrane MECIR: planning review methods at protocol stage / Cochrane Handbook: reporting protocols and completed reviews

#The analysis plan adds operational detail

CONSORT 2025 explains that a trial protocol may contain a complete statistical analysis plan or describe main principles while linking to a separate, more detailed plan. Its explanation notes that the plan can cover analysis of primary and secondary outcomes. SPIRIT also includes access to the analysis plan among its protocol-reporting items.

Read whether enough detail is available to understand the planned comparison and outcome analysis. A short registration record may not settle all analytical choices. This page does not supply statistical methods for a particular dataset or a complete analysis-plan checklist.

Evidence: CONSORT 2025 explanation and elaboration / SPIRIT 2025: reporting randomized-trial protocols

#Changes need their reasons and timing

CONSORT notes that legitimate protocol or analysis-plan changes can occur after a trial starts, for example when unexpected problems or new evidence arise. It calls for iterations to record the changes, rationale and timing. Cochrane also asks authors to explain differences between a protocol and completed review.

Read what changed, when it changed and how it might affect interpretation. An amendment is not automatically misconduct, but an unexplained change prevents readers from judging whether findings influenced the decision. Mynd claims no conducted trial, registered protocol or clinical review on this page.

Evidence: CONSORT 2025 explanation and elaboration / Cochrane Handbook: reporting protocols and completed reviews

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.