- Published
- Updated
Excluding a paper as a failed study simply because it does not show the expected benefit can distort the evidence. Favourable and unfavourable results should pass through the same scientific eligibility framework. Result direction and study reliability are different assessments.
Scope of the decision
An unfavourable finding may involve no demonstrated difference, an important unwanted event or lower performance in a subgroup. Interpret these possibilities alongside design and uncertainty. A statistically nonsignificant result does not automatically establish equivalence and may require appropriate methodological expertise.
Worked review
For example, favourable reports show strong short-term performance while a new study reports reduced benefit at longer follow-up. The team does not dismiss it as an outlier. It examines population, methods and follow-up differences and assesses effects on the time boundary of the claim and remaining evidence needs.
Practical checks
- Keep result direction separate from scientific eligibility and quality decisions.
- Explain the population, outcome and time horizon of the unfavourable finding.
- Refer implications for claims and benefit-risk assessment to appropriate personnel.
Acceptance and limits
Acceptance requires visible inclusion and a reasoned assessment of the finding’s weight. One unfavourable study does not automatically invalidate every earlier result. Conversely, a conclusion assembled only from favourable papers is not balanced; address conflict and uncertainty explicitly.
Frequently asked questions
- Does no significant difference mean the devices are equal?
- No. Assess design, sample and uncertainty before making such an inference.
- Can a small negative study be excluded?
- Decisions should follow eligibility and methodological rationale rather than the direction of the result.