Confirms what UAT already tested
Reconciliation checks that the logic verified in IxRS UAT is holding up in the live system, subject by subject.
What's included
What goes into randomization reconciliation
- Allocation and stratification data compared between IxRS and the clinical database on a defined recurring cadence
- Discrepancies in treatment assignment, enrollment status, or stratum flagged and routed for resolution
- Reconciliation performed by the blinded team, with any unblinded follow-up handled per the DMC charter's access controls
- Findings tracked to closure and summarized ahead of interim analyses or database lock
- Reconciliation logic revisited whenever the randomization specification is amended mid-study
Clean before lock, not scrambled at lock
Discrepancies caught during conduct don't accumulate into last-minute issues ahead of statistical QC of the database.