values recorded multiple times for a visit

Out-of-Window Edits: Data entered or modified after protocol-defined timeframes

Unauthorized Access: Users making changes outside their assigned roles

Retrospective Entries: Backdated entries without justification

Frequent Value Changes: Fields modified multiple times without clear rationale

Deleted Records: Data removed without explanation or traceability

Consider the following audit trail excerpt that helped uncover an unreported protocol deviation:

Subject	Field	Old Value	New Value	User	Date/Time	Reason
SUBJ103	Dose Administered	100 mg	200 mg	CRC_Jason	2025-05-22 15:05 UTC	Dose correction after error noticed

While the value was corrected, the audit trail revealed no deviation was filed, and the PI had not signed off. Without the trail, this event might have gone unnoticed.

Steps to Investigate Data Discrepancies Using Audit Trails

When an inconsistency is detected — either through monitoring, data management review, or statistical checks — audit trail analysis should follow a systematic approach:

1. Identify the anomaly: Determine which subject or form has the discrepancy.
2. Pull the audit log: Extract the audit trail for the specific field or visit.
3. Trace modification history: Review timestamps, user IDs, and reasons for changes.
4. Cross-check source documents: Validate data against site records or EHR screenshots.
5. Interview involved personnel: Understand the rationale behind any unexpected changes.
6. Document the investigation: Log the findings and any resulting CAPAs or protocol deviations.

These steps ensure both transparency and defensibility during regulatory inspections.

System Features That Support Effective Discrepancy Investigations

Modern EDC systems often include built-in features that simplify audit trail review and facilitate data investigations:

• 🔍 Filtered Audit Logs: Ability to isolate logs by subject, user, or field
• 📋 Color-coded Change Logs: Visual highlighting of changes for quick identification
• 📂 Export Functions: Downloadable logs for documentation and inspection
• 👥 User Role Mapping: Assigns changes to specific personnel roles for accountability
• 📎 Source Document Upload: Attachments to justify corrections

These functionalities are critical for preparing inspection-ready documentation and resolving discrepancies before database lock.

Regulatory Expectations for Audit Trail Use in Discrepancy Management

Both the FDA and EMA expect that sponsors have systems and SOPs in place for audit trail review, especially in response to data discrepancies. In FDA inspections, examples of key expectations include:

• ✔️ Sponsors must demonstrate timely detection and resolution of discrepancies.
• ✔️ Audit logs must be reviewed by trained personnel and stored in the TMF.
• ✔️ Investigations must be documented and linked to protocol deviations if applicable.
• ✔️ Systems must prevent retrospective tampering of audit records.

Refer to Japan’s PMDA Clinical Trial Portal for additional global perspectives on audit trail use and data traceability requirements.

Inspection Findings Involving Audit Trail Investigations

Here are examples of actual inspection findings related to audit trail investigations:

Finding 1: Inadequate Documentation of Correction

The sponsor failed to document the reason behind repeated changes to SAE classification in the EDC system. The audit trail existed but lacked detailed rationale.

Regulatory Response: Issued a 483 citing lack of documentation and absence of QA oversight.

Finding 2: No Training on Audit Log Review

CRAs were unaware of how to access or interpret audit trails, resulting in missed data discrepancies at multiple sites.

Regulatory Response: Warning letter issued and training program overhaul mandated.

Best Practices for Site and CRA Involvement

Investigating discrepancies isn’t just a data management function. CRAs and site personnel play critical roles. Recommendations include:

• ✔️ Integrate audit log checks into routine monitoring visits
• ✔️ Train site staff on documentation requirements for post-entry changes
• ✔️ Use centralized monitoring to flag unusual data patterns
• ✔️ Maintain logs of all investigations and resolutions in the eTMF

Conclusion

Audit trails in EDC systems are more than digital footprints — they’re the backbone of any data discrepancy investigation. By building systems that support detailed, tamper-proof audit logs and by training teams to use them effectively, sponsors and CROs can significantly reduce the risk of undetected data issues and inspection findings.

Establishing SOPs, using automated alerts, and conducting routine reviews will ensure that your audit trails aren’t just available — they’re actionable. In the complex world of clinical data management, that makes all the difference.