EUDAMED UDI error library:
Top 25 Returns (Symptom → Cause → Fix → Prevention)
This bug library is intended for Regulatory Affairs & Quality to quickly classify returns, control corrections cleanly and incorporate prevention into the process. Use the structure per error: symptom, probable cause, RA/QA quick fix, prevention control.
Practical tip: Create an “error board” internally (Jira/Excel) that maps each return to exactly one library entry. This way, recurring errors become visible and disappear permanently.
1) Master data & identifiers
1. SRN is missing or inconsistent
- Symptom: Record is rejected or remains under review.
- Cause: SRN not maintained / wrong actor (manufacturer vs. authorized representative).
- Quick Fix: Set SRN source, entry consistent across all relevant records.
- Prevention: Define field owner, mandatory field check before submission.
2. Basic UDI-DI and UDI-DI not cleanly linked
- Symptom: UDI-DI is not accepted or “parent” is missing.
- Cause: Parent code not maintained or spelling/format deviating.
- Quick Fix: Create the master list Basic UDI-DI and use it as a reference.
- Prevention: Lookup validations (only allow existing parent codes).
3. Issuing Agency /Code Format Inconsistent
- Symptom: “Invalid code” / format error.
- Cause: Issuing body (e.g. GS1/HIBCC/ICCBBA) not correctly or code incorrectly structured.
- Quick Fix: Check the allocation point per product line, harmonise code formats.
- Prevention: Drop-down selection + regex/format checks (Excel) and validation before submission.
2) Classification, Legislation, Equipment Characteristics
4. MDR/IVDR legislation set incorrectly
- Symptom: Contradictions in mandatory fields or rejection.
- Cause: Record uses MDR logic in IVDR or vice versa.
- Quick Fix: Check applicable legislation and correct dependent fields.
- Prevention: “Legislation Gate”: release dependent fields only in appropriate logic.
5. Contradictory risk class / characteristics
- Symptom: Data set not plausible (e.g. implant flag vs. class).
- Cause: Classification not coordinated with device flags.
- Quick Fix: Check classification matrix (RA/QA) against device attributes.
- Prevention: Review checklist: Check risk class + central flags together.
3) UDI-DI, Status, Names & Models
6. Device name/model inconsistent across variants
- Symptom: Variants difficult to understand, data seem inconsistent.
- Cause: Naming conventions are missing (marketing vs. engineering vs. labeling).
- Quick Fix: Define naming SOP (Device Name vs. Model vs. Reference).
- Prevention: “Golden Values” from a System of Record; Approval by RA/QA.
7. Device status incorrect or not maintained
- Symptom: Refusal / conflicts with updates.
- Cause: Active/inactive logic not clear; old variants remain active.
- Quick Fix: Set status definition and clean affected records.
- Prevention: Change process: Status change only via release + evidence.
4) Packaging Hierarchy & Child UDIs
8. Packaging levels incomplete
- Symptom: Declines in packaging structures.
- Cause: Child UDI or quantity information missing; hierarchy not consistent.
- Quick Fix: Model packaging levels as a tree (level 0/1/2 …), check quantities per level.
- Prevention: Packaging checklist: level, UDI, quantity, status, relationship.
9. Unit of Use / Secondary Identifiers applied incorrectly
- Symptom: Unclear identifiers, inconsistencies.
- Cause: Unit of Use vs. Packaging not cleanly separated.
- Quick Fix: Clarify terms internally and define rules per product family.
- Prevention: Document field guidance directly in the template (comment/help text).
5) Production Identifiers (PI): Lot/Serial/Expiry/Manufacturing
10. PI fields are “possible” instead of “actually” maintained
- Symptom: Data does not match the label / real production process.
- Cause: Confusion “PI may occur” vs. “PI is on the label”.
- Quick Fix: Set PI decision based on actual labeling.
- Prevention: Define labeling owner as data owner for PI logic.
6) Certificates, NB & Revisions
11. Certificate number/revision number unclear
- Symptom: Queries or inconsistencies when purchasing a certificate.
- Cause: Revisions/extensions not neatly versioned.
- Quick Fix: Set up certificate register (no., rev., type, NB code, runtime).
- Prevention: Change process: Certificate changes trigger UDI review.
7) Process error (the most common “invisible” causes)
12. Data changes without approval
- Symptom: Returns are repeating; audit evidence is missing.
- Cause: No formal review/approval mechanism.
- Quick Fix: Make RA/QA release step mandatory before each submission.
- Prevention: Use workflow with audit trail and roles (portal/SAP add-on).
13. “Submission Success” will not be archived
- Symptom: Audit/inspection lacks evidence, although technically submitted.
- Cause: Result files/government responses are not backed up centrally.
- Quick Fix: Introduce evidence filing for each submission run (zip/reports/logs).
- Prevention: Evidence SOP (monthly/per run) + naming convention.
The remaining 12 errors (short format)
Use this list as a “fast check”. If you wish, I can formulate each point analogous to above (symptom/cause/fix/prevention).
- Direct marking information is missing or contradictory.
- Latex/human/animal tissue flags implausible.
- IVDR specifics (e.g. self-testing) are missing despite IVDR labelling.
- Language not consistent with product information provided.
- Storage & handling conditions incomplete.
- Device labelled as sterile / needs sterilization before use contradictory.
- Reusable surgical instrument flag set incorrectly.
- Reference number/material number duplicate or not unique.
- Packaging status differs from device status without justification.
- Multiple UDI DIs are inadvertently listed as identical.
- Data export/import creates tacit format changes (e.g. leading zeros).
- Updates are submitted as a “new creation” instead of a change.
Next step
If you want to systematically reduce returns, combine
Excel template with the Global Submission Portal
or – in the case of SAP governance – Global UDI Add-on for SAP.
FAQ
How do I interpret technical validation errors as RA/QA?
Always work with the logic “What exactly is affected?” (Field/Section), “Which rule is being violated?” and “What source is Owner?”. Ideally, you will get errors localized in such a way that corrections are possible without XML special knowledge.
How do I make sure an error doesn’t come back?
Every error must have a prevention check: template validation, mandatory field gate, approval process or automated check before submission. Once you standardize this, the return rate decreases significantly.