FDA's data integrity guidance for drug CGMP sets out expectations such as controlled access, audit trails that can't be bypassed, and records that are original, accurate and retained. The ALCOA+ principles describe what good data looks like. In the EU, the 2025 draft revisions of Chapter 4 and Annex 11 expand expectations on data governance across paper, electronic and hybrid systems.
Why it matters for labeling
An approved master, review comments, comparison results and approvals are all data that supports a release decision. Integrity problems in labeling usually look like missing or overwritten history: approvals that can't be tied to a specific version, edits after sign-off, or records scattered across email and shared drives. A complete audit trail is the main protection.
Common mistakes
- Shared logins or accounts that make actions unattributable
- Records that can be edited or deleted without a trace
- Evidence held in several places, with no single reliable record
- Backups and retention that were never tested