Skip to content
Scintria
ToolsGrant writerFunding
HomeLiteratureClinical trialsEvidence LabProtocol packData analysisLearnProjectsLibraryReferencesApplications
Start free
ScintriaStart free

Learn / Clinical Research Management / Data management and trial statistics

Clinical data management

13 minute read · 6-question quiz

Download this lesson as a colour Word or PDF study guide with the figures and answers. Included in paid plans; reading and quizzes stay free.Sign in

What you will learn

  • Describe the stages of the clinical data management process

  • Design a simple CRF and explain the role of the data management plan

  • Explain queries, coding and reconciliation

Clinical data management (CDM) collects, cleans and prepares trial data so that statisticians can analyse them reliably and regulators can trust them. The aim is a database that is complete, accurate, consistent and verifiable.

Stages

  1. Planning: write the data management plan (DMP), covering CRF design, data flow, validation checks, coding, reconciliation and lock procedures.
  2. CRF design: create forms (paper or electronic) that collect exactly what the protocol needs, no more and no less.
  3. Database build and validation: build the electronic data capture (EDC) system with edit checks and test it (user acceptance testing) before use.
  4. Data entry or capture: staff enter data at the site into the EDC, or data are entered from paper CRFs.
  5. Data cleaning: automatic and manual checks find missing, inconsistent or implausible values, and queries go to the site for resolution.
  6. Coding: adverse events are coded with MedDRA and medicines with the WHO Drug Dictionary.
  7. Reconciliation: safety data in the clinical database are compared with the safety database; laboratory and other external data are matched to the main database.
  8. Database lock: the database is closed to changes before analysis.
  9. Archiving and transfer.

CRF design principles

  • Collect only data needed for the protocol's endpoints, safety and compliance.
  • Use clear questions and standard formats for dates and units.
  • Prefer coded choices to free text.
  • Order fields to match the visit flow.
  • Limit duplication.
  • Provide completion guidelines.
1PlanDMP, CRF design2Build andtestEDC, edit checks3Capture andcleanEntry, queries4Code andreconcileMedDRA, safety,lab5LockFreeze foranalysis
The data management process from planning to database lock.

Data cleaning and queries

Edit checks run automatically when data are entered, flagging for example a heart rate of 300, a visit date before consent, or a missing mandatory field. When a check fails, a query is raised to the site, which responds by correcting or confirming the value, with an audit trail. Data managers also run manual review and listings to find patterns the system cannot catch, such as a site with unusually uniform values.

Coding

Verbatim terms recorded by sites ("tummy ache", "stomach pain") are mapped to standard dictionary terms (MedDRA for events) so they can be grouped and counted.

Database lock

Lock happens only when all queries are closed, coding and reconciliation are complete, protocol deviations are classified, and sign-offs are obtained. After lock, unblinding and analysis proceed. Unlocking is exceptional and fully documented.

Worked example: query metrics

A trial with 50,000 data points has generated 1,500 queries by month 8.

  • Query rate = 1,500 / 50,000 = 3 per 100 data points.
  • Open queries: 180, so 88% closed.
  • Average resolution time: 9 days, above the 5-day target.

Site 4 has 600 of the queries, 40% of all, mostly for missing dates. The team retrains Site 4 on date formats and reviews the form design. The query rate falls to 1.8 per 100 within two months.

3 per 100Query rate88%Queries closed9 daysMean resolution time
Typical data cleaning metrics monitored by the data manager.

Prevention beats cleaning

A well-designed CRF with clear instructions and built-in checks prevents many queries. Every avoidable query costs time at the site and the data centre.

Changing data after lock

Unplanned changes after lock undermine credibility and may force reanalysis. Make sure all issues are resolved before lock, and document any exceptions rigorously.

Check yourself

Look at a CRF page. Is every field needed for an endpoint, safety, or compliance? Which ones could be removed?

Summary

  • CDM aims for complete, accurate, consistent and verifiable data, following a data management plan.

  • CRFs collect only what is needed; edit checks and queries clean data with an audit trail.

  • MedDRA and the WHO Drug Dictionary standardise adverse event and medicine terms.

  • Database lock follows closed queries, coding and reconciliation; changes after lock are exceptional.

Check your understanding

Question 1
1

The data management plan describes:

Question 2
2

MedDRA is used to:

Question 3
3

1,500 queries from 50,000 data points is a query rate of:

Question 4
4

A query is:

Question 5
5

Database lock should occur:

Question 6
6

An edit check is:

Draft study guide. Spotted an error or something unclear? Tell us at hello@scintria.com. This is educational material, not clinical or regulatory advice.

The trial master file and essential documentsRandomisation, blinding and the statistical analysis plan