An audit trail is a record of user activity in a database. It documents actions such as record creation, data entry, edits, deletions, exports, imports, user rights changes, and project design changes. In clinical research, audit trails support data integrity by allowing the study team to reconstruct who did what, when, and sometimes why.
Audit trails are essential because clinical research data must be attributable and traceable.
If a primary outcome changes from “not hospitalized” to “hospitalized,” the team should be able to identify when the change occurred and which user made it. If a record is deleted, the deletion should be visible. If a dataset is exported, the export should be logged. These records are important during monitoring, audits, inspections, and internal quality reviews.
Audit trails do not prevent all errors, but they make data changes transparent. This transparency changes user behavior and supports accountability. Users are more likely to follow procedures when they understand that changes are recorded. Data managers can also use audit logs to investigate unusual patterns, such as repeated changes to critical fields, exports outside expected workflows, or edits made long after visits occurred.
Audit trails depend on individual user accounts. If multiple staff share one account, the audit trail cannot identify the actual person who made a change. This undermines accountability and is inconsistent with good data integrity practice. Each user should have a unique account, and passwords should not be shared.
Audit log review should be risk-based. It may not be feasible to review every action in a large study. Instead, data managers may review logs for critical variables, deleted records,
late changes, export activity, user rights changes, or unusual site patterns. The review process should be documented, and findings should be followed up through queries, training, or corrective actions.