_(39288504530).jpg){kind=link}
- Published
- Updated
Treating randomisation as an unavoidable system step can conceal incomplete eligibility decisions. Determine which information must be available before assignment by considering both the protocol and the actual site workflow.
Define the question and assumptions
Distinguish pending laboratory findings, unconfirmed medical history and expired assessments. A completed data entry screen is not the same as an authorised clinical decision. The record should show who determined eligibility and on what evidence; system permission to proceed is not itself evidence of eligibility.
Hypothetical study example
In a hypothetical study, a screening result becomes available outside normal hours. A coordinator proposes randomising earlier to meet a planned date. The team checks decision authority and required information instead of treating uncertainty as a box that can be completed retrospectively.
Design review checks
- Check the validity window of every assessment supporting the eligibility decision.
- Define decision authority when findings are missing, delayed or internally inconsistent.
- Separate safety management from analysis handling for an inadvertently randomised participant.
Record the decision and its limits
An acceptable workflow demonstrates the clinical decision, not merely populated fields. If an error occurs, do not silently erase an assignment or record. Protect the participant, document the event and assess analysis implications under prospectively defined principles with the relevant scientific and clinical personnel.
Frequently asked questions
- Do completed EDC fields establish eligibility?
- No. Valid assessments and an authorised clinical determination must also be demonstrated.
- Does incorrect randomisation automatically mean exclusion?
- No. Participant safety and statistical analysis address different questions.