- Published
- Updated
A protocol changing eligibility, visit timing or endpoint definitions affects more than operational schedules. It can alter comparability, planning assumptions and interpretation. Include a statistical impact assessment rather than treating the amendment only as a document-distribution task.
Define the question and assumptions
Connect the changed element to the scientific question, information collected and planned analysis. Examine whether participants enrolled under old and new versions can be evaluated consistently. Record what information was visible when the decision was made, distinguishing blinded review from changes informed by treatment-result knowledge.
Hypothetical study example
In a hypothetical trial, the primary visit window is widened. The team does not regard this solely as a fix for scheduling delays. It reviews measurement-time meaning, mapping of participants across versions and assumptions about endpoint availability in the sample size calculation with the responsible statistician.
Design review checks
- Map all endpoint, population and analysis elements affected by the amendment.
- Record decision timing and the level of information available to decision makers.
- Test whether new rules can be applied to earlier observations using example participants.
Record the decision and its limits
An accepted impact record explains rationale, potential bias and reporting consequences, alongside applicable review and approval steps. Producing a more favourable result cannot justify an analytical change scientifically. Preserve previous rules and transition decisions, and reconcile versions of the protocol, statistical plan and final report so the history remains reconstructable.
Frequently asked questions
- Are operational changes independent of analysis?
- Not always. Timing or measurement changes can alter endpoint meaning.
- Should pre-amendment data be deleted?
- No. Preserve source observations and explain their analytical use.