Clinical Sample Reconciliation Across CROs and Labs: Why It Breaks and How to Fix It

Why the sources disagree

• Sample IDs are re-labelled when samples move between vendors, or aliquots are created and given new IDs.
• Visit dates in the EDC and collection dates on the manifest differ by a day, or a visit is rescheduled after collection.
• Shipments are logged as sent and never logged as received.
• Consent is withdrawn after samples have been collected, and the inventory does not know.

None of these is unusual. Each one, found late, turns a simple correction into a deviation.

What good reconciliation looks like

Match every sample to a subject, a visit, a consent scope and a location, every time new data arrives rather than once a quarter. Show conflicts with both values rather than letting one source overwrite another. Send each vendor a discrepancy list with the evidence attached, and keep the query and its resolution on record.

Use the schedule of assessments as the expected baseline, so a visit that should have produced a sample and did not is visible as soon as the visit is recorded.

The checks that prevent findings

• Samples from subjects who withdrew consent that are still stored or in testing.
• Breaks in chain of custody between vendors.
• Samples approaching their stability limit.
• Results that belong to samples not in the inventory.

These should run continuously and go to a named owner, not wait for database lock.

Where this fits

Study Intelligence brings each vendor's data into one view of the study and runs these checks as data arrives. See Sample Intelligence for the inventory and vendor and lab reconciliation for how discrepancies go back to vendors.