.jpg){kind=link}
- Published
- Updated
- Sources checked
A control group needs a stronger rationale than convenience or customary use. An E10-focused decision note considers ethical acceptability alongside the study’s ability to detect a real treatment difference, making the scientific implications of the comparator explicit before the design is finalized.
Clarify the study objective
ICH E10 addresses scientific and ethical considerations for placebo, active and other control approaches. Demonstrating superiority and excluding an unacceptable loss of effect are different design questions. Selecting a comparator before defining the intended claim makes the later justification harder to assess and defend scientifically.
Review comparator evidence
When considering an active control, compare historical effect evidence, populations, treatment standards and endpoint measurement. Assay sensitivity is the ability to distinguish an effective from an ineffective treatment. Naming a current comparator is insufficient without examining whether the earlier evidence can reasonably inform the new study.
Apply the design checklist
Poor adherence, rescue therapy, short follow-up or insensitive measurement can make treatments look artificially similar. Connect these risks to protocol safeguards. Where a noninferiority margin is needed, assess its scientific rationale, clinical meaning and analytical assumptions together rather than treating the number as a statistical convention.
- State the intended superiority or noninferiority claim.
- Assess transferability of the comparator’s historical effect.
- Control conduct problems that could create artificial similarity.
State interpretive limits
Consider ethics through access to effective treatment and participant safety. Do not claim that one control type suits every indication. Record the selected approach, alternatives rejected and limits on interpretation; where regional regulatory advice is needed, include that interaction in the development plan before locking the design.
Frequently asked questions
- Is an active control always simpler?
- No. Historical evidence and assay sensitivity are particularly important for noninferiority interpretation.
- Does similarity between groups prove efficacy?
- Not alone. Consider design, margin, data quality and assumptions together.