Plan: Clinical Data Management Plan (DMP)

This is a ready-to-use Clinical Data Management Plan (DMP). It is the controlling document for how data is captured, cleaned, coded, reconciled, standardized, and locked for one study. Replace every <<FILL: ...>> placeholder with your own specifics, set your document numbers and dates, and route it through your normal review and approval before the study starts. A worked filled specimen follows. Confirm each cited regulation against the current source before you rely on it. This is general educational reference, not regulatory or professional advice; adapt it to your protocol, systems, and quality system.

Document control header

1. Purpose and scope

This DMP defines the data management activities, systems, standards, and responsibilities for Study <<FILL: protocol>>. It covers CRF and EDC design, edit checks, user acceptance testing, data review and query management, medical coding, external-data and safety reconciliation, CDISC standardization, and database lock. It applies from study build through lock and archival. It does not restate the protocol or the statistical analysis plan; it references them.

2. Systems and standards

3. Roles and the sponsor/CRO split

State who performs and who has accountability for each activity. In an outsourced model the sponsor retains oversight accountability under ICH E6 even where a CRO executes.

4. Data flow and standards

Describe the path from source to submission: source to CRF (eCRF) to SDTM to ADaM to tables/listings/figures, with the annotated CRF linking collected fields to SDTM targets. State the mapping and conformance approach (for example, SDTM mapped per the SDTMIG and validated with a conformance checker; ADaM derived from SDTM with traceability; Define-XML and the reviewer’s guides produced). Reference: clinical-data-management-cdisc. Note the current submission transport format (XPT today; confirm whether Dataset-JSON is accepted in the Data Standards Catalog at submission time).

5. Data cleaning strategy

• Edit checks: the philosophy and severity model (hard stop reserved for format/required; most consistency issues raise a query; soft warnings inform only). The full check inventory lives in the Edit Check Specification, <<FILL: ECS-ID>>. See Edit Check Specification template.
• Manual data review: scope, reviewer roles, and cadence; listings reviewed and by whom.
• Query management: routing, response expectations, aging thresholds and escalation, and the rule that queries are never bulk-closed to hit a lock date without genuine resolution.
• Central/statistical monitoring signals feeding risk-based monitoring.

6. Medical coding

State the dictionaries and versions, the auto-code then manual-code workflow, the coding conventions and decision log, the medical review of coded terms, the query approach for uncodable verbatim terms, and the up-versioning decision. Coding is complete and reconciled before lock.

7. Reconciliation plan

Reconciliation is scheduled through conduct, not deferred to the week of lock.

8. Quality and metrics

• Quality tolerance limits and key metrics: open-query count and median age, queries per CRF page, missing-pages count, overdue coding, reconciliation status, and critical-variable completeness. See quality-metrics-and-kpis.
• Risk-based monitoring linkage: how CDM signals feed the monitoring plan. See risk-based-monitoring-clinical.

9. Database lock and archival

State the lock prerequisites (all critical-variable queries closed, coding complete and reviewed, all reconciliations complete, deviations classified, functional sign-offs), the two-phase or single lock model, the unlock-under-change-control rule, retention period, and archival format. The operational checklist is the Database Lock Checklist.

10. References

ICH E6(R2)/E6(R3), Good Clinical Practice (data handling, computerized systems, sponsor oversight). 21 CFR Part 11 (Electronic Records; Electronic Signatures); EudraLex Volume 4, Annex 11. FDA Study Data Technical Conformance Guide and FDA Data Standards Catalog (CDISC formats). ICH E2A/E2B (safety data), for SAE reconciliation context.

Confirm the current version of each reference before issue.

11. Revision history

Filled specimen (excerpt)

The following shows sections 2 and 3 completed for an illustrative Phase 2 study. Details are illustrative.

Study: ABC-201, Phase 2, 240 subjects. Sponsor: illustrative biotech. CRO: illustrative full-service CRO (data management outsourced; sponsor retains oversight).

Section 2, systems and standards:

Section 3, roles (excerpt):

The oversight column is what an inspector checks in an outsourced trial: execution can move to the CRO, accountability does not leave the sponsor.

Common inspection findings this plan prevents

• A study running without a current, approved DMP, so the process is undefined.
• The sponsor/CRO responsibility split unstated, so oversight gaps appear at inspection.
• Reconciliation and coding treated as end-of-study tasks with no planned cadence.
• Dictionary and CDISC versions not fixed, so submission metadata does not match the data.
• No defined query-aging or bulk-close rule, so a rushed lock contaminates the dataset.

How to adapt this plan

1. Fill the header, systems, and CDISC/dictionary versions in sections 1 and 2.
2. Complete the roles table honestly for your sponsor/CRO model; name the accountable function for each activity.
3. Point the cross-references (ECS, lock checklist, SAP, monitoring plan) to your real document numbers.
4. Set the query-aging thresholds and reconciliation cadence to your study’s size and risk.
5. Confirm every regulation in section 10 against the current published version, and confirm the accepted submission format in the Data Standards Catalog, before issue.