- Published
- Updated
A system version may be assessed for a defined use before every open defect is corrected. Conditional acceptance should not hide those defects or treat a promise of future correction as a substitute for present evidence.
Preserve the meaning of data
Link each open defect to the actual study function. Define affected users, data, triggering conditions and interim controls. Consider detectability as well as likelihood. People with acceptance authority need to understand the residual risk and whether restricted use is practical. A workaround creating uncontrolled workload should not be assumed sustainable simply because it can be described in a document.
Worked example
For example, one report displays an incorrect total under a particular filter combination. The study team restricts that combination, identifies a verified alternative report and clearly informs users. The interface should not imply that the defect is resolved. When a permanent fix arrives, removal of the interim control receives its own considered decision.
Review checks
- Show the triggering condition and affected output clearly.
- Test the workaround with the people expected to use it.
- Assign an owner, review point and reassessment trigger.
Acceptance and boundaries
Acceptance means a stated restricted use is defensible, not that every function is generally suitable. A use failing a critical participant-safety or reliable-results requirement cannot become acceptable merely by relabelling it. Revisit the decision when conditions change. The open-defect register should connect vendor follow-up with the study’s own reasoned risk decision.
Frequently asked questions
- Does every open defect require stopping use?
- Assess impact and context, while avoiding unjustified acceptance of an unmet critical requirement.
- Can training close a workaround?
- Training can be one control, but its actual application and effectiveness still need evidence.