Question 10 of 20 Fundamentals Practical
What should a data management plan cover?
Short answer
A data management plan says how trial data will be collected, checked, coded, reconciled, transferred, locked and archived, and who does each step. It ties those processes to the protocol and the data that are critical to the results, so effort goes where errors would matter.
Why
E6(R3) names the data management plan alongside the protocol, statistical analysis plan and monitoring plan as documents that should be clear, concise and operationally feasible. A useful plan covers the systems used and their validation status, case report form design and completion guidelines, edit checks and how they are tested, query workflows and timelines, medical coding dictionaries and versions, external data such as central laboratory or ePRO data and their transfer agreements, SAE reconciliation, protocol deviation handling, data review and audit trail review, the steps to database lock, and archiving. It should name roles and responsibilities, especially when a CRO and the sponsor share work.
EMA's computerised systems guideline expects trial specific system configuration to be checked against the protocol and the data management plan, so the plan is also a test basis. The plan is a living document. A protocol amendment that adds a visit or endpoint should trigger a plan update.
A plan copied from another study is a warning sign at inspection.
A strong answer covers
- Lifecycle from collection to archiving
- Edit checks, queries, coding and reconciliation
- External data transfers and agreements
- Roles across sponsor and CRO
- Updated with protocol amendments
Common mistakes
- Writing a generic plan that does not mention the trial's critical data.
- Leaving external vendor data out of the plan.
- Never updating the plan after amendments.
The interviewer might ask next
Which sections of the plan would you update first after an amendment adds a new endpoint?