Independent and not affiliated with the FDA, MHRA, ISPE, PDA, or any agency. Get the appgoutham@madhadi.com
madhadi.comData Integrity & GxP Quality
Browse all topics → Articles Templates & Procedures Learning paths GlossaryScenariosToolsRegulatory ReferencesLearning PathsTopics About Start here
Protocol Plug-and-play starting point Clinical & GCP

Protocol: EDC Study-Build Validation

A plug-and-play computer system validation protocol for an electronic data capture (EDC) study build: approval page, scope split between platform and study configuration, risk-based test cases for eCRF pages, edit checks, audit trail, and access control, with a filled specimen test case.

Document type: Protocol

Read and copy the template below into your own quality system. It is a generic starting point for your own internal use, provided as is, with no warranty; see the Terms and License. Adopting it does not by itself create compliance.

This is a ready-to-use validation protocol for the study-specific configuration of an electronic data capture (EDC) system. It assumes the base platform was separately qualified by the sponsor, CRO, or vendor; this protocol tests what your study team actually built on top of it. Replace every <<FILL: ...>> placeholder, scale the test depth to your documented risk assessment, and route it through your normal validation review and approval before first-subject-first-visit. A filled specimen test case follows. Verify each cited regulation against the current source before you rely on it. This content is general educational reference, not legal or regulatory advice.

1. Approval page

RoleNameSignatureDate
Author (Validation / Data Management)<<FILL>>
Clinical Data Management lead<<FILL>>
Clinical QA<<FILL>>
Biostatistics representative<<FILL>>
System owner (EDC platform)<<FILL>>
FieldEntry
Protocol number<<FILL: VAL-EDC-###>>
Study / protocol number<<FILL>>
EDC platform / version<<FILL: e.g. vendor product and release>>
Protocol version / date<<FILL>>

2. Objective

Provide documented evidence that the study-specific EDC configuration performs as intended: the eCRF pages, edit checks, dynamic display logic, randomization or IRT interface, user roles, audit trail, and export mechanism function per the approved specification and preserve data integrity, in line with GCP, 21 CFR Part 11 (for US submissions), and EU GMP/GCP Annex 11 expectations for computerized systems.

3. Scope

In scope: the study-specific build layered on the qualified base platform for <<FILL: study / protocol number>>, covering eCRF structure, edit checks and validation rules, dynamic forms, user roles and access, audit trail behavior, query workflow, randomization/IRT interface (if applicable), and the export format used by biostatistics. Out of scope: base platform qualification (covered by the vendor/platform qualification record <<FILL: reference>>), and downstream systems validated separately, for example the eTMF or the safety database, tested only at the integration seam in section 7.

4. System description

<<FILL: hosting model (vendor-hosted / sponsor-hosted), user roles (site coordinator, investigator, monitor, data manager, medical coder, biostatistician), integrations (IRT, central lab, ePRO, safety database), languages/sites supported>>

5. Prerequisites

  • Approved User Requirements Specification (URS) and risk assessment on file; see User Requirements and Traceability and CSV Risk Assessment Methodology.
  • Base platform qualification record referenced and current.
  • Configuration specification (build spec) approved and matches the final eCRF.
  • Edit check specification approved; see the companion Edit Check Specification template.
  • Test environment representative of production, with the same configuration version deployed.
  • Test accounts provisioned for every role in scope, including a read-only monitor account and a revoked/expired account for negative testing.

6. Roles and acceptance

Each test case is executed by <<FILL: tester role, e.g. Data Management>> and independently reviewed by <<FILL: QA or a second data manager>>. A case passes when the actual result equals the expected result with objective, dated evidence attached (screenshot, export, or audit trail excerpt). Any deviation is logged in section 8 and dispositioned, with high-risk-function failures resolved and retested, before the summary is signed.

7. Test cases

Record for each: actual result, pass/fail, tester, date, evidence reference. Depth should track the risk assessment: primary-endpoint and safety fields get full scripted coverage; low-risk reference fields can use a lighter, documented rationale.

IDRequirementTest stepExpected result
TC-01eCRF page structure matches build specOpen every page for a fresh subjectFields, labels, units, and page sequence match the approved configuration spec
TC-02Required-field enforcementAttempt to save a page with a required field blankSave is blocked with a clear message; no partial record is created
TC-03Range / logic edit check firesEnter a value outside the specified valid rangeThe specified query fires with the specified text, at the specified severity
TC-04Edit check does not over-fireEnter a valid boundary value from the edit check specNo query fires
TC-05Dynamic display / branching logicEnter a value that should trigger a conditional field or pageThe conditional field/page appears (or hides) exactly as specified
TC-06Query lifecycleRaise, answer, and close a queryQuery state transitions correctly; audit trail captures raise, answer, and close with user and timestamp
TC-07Audit trail: initial entryEnter a new valueTrail records user, timestamp, and the value entered
TC-08Audit trail: modificationChange a previously saved valueTrail records user, timestamp, old value, new value, and reason for change
TC-09Audit trail cannot be disabledAttempt to suppress or turn off the trail as a non-privileged userNot possible; no user-facing control exists to do so
TC-10Role-based access: siteLog in as a site userSees only that site’s subjects; cannot access other sites’ data
TC-11Role-based access: monitorLog in as a monitor accountRead-only; cannot edit source-equivalent fields
TC-12Revoked accountAttempt login with a deactivated accountAccess denied; attempt is logged
TC-13Randomization / IRT interfaceTrigger a randomization call from the EDCCorrect treatment arm/kit assignment returned and recorded; no duplicate or missing calls
TC-14External data reconciliation fieldLoad a sample central-lab or ePRO transferData lands in the correct fields, matches the transfer specification, and any mismatch triggers a reconciliation flag
TC-15Export formatGenerate the dataset export used by biostatisticsExport structure, variable names, and record counts match the export specification
TC-16Time zone / timestamp integrityEnter data from a non-UTC siteTimestamps are captured and displayed consistently with the system’s documented time-handling design

8. Deviation handling

Log each deviation with description, impact, root cause, and resolution. A high-risk-function failure (audit trail, access control, randomization interface, a primary-endpoint edit check) must be resolved and re-tested before the summary is signed.

Dev #Test caseDescriptionDisposition
<<FILL>><<FILL>><<FILL>><<FILL>>

9. Acceptance criteria (protocol level)

  • Every eCRF page and field matches the approved configuration specification.
  • Every edit check in the approved specification fires correctly and does not over-fire on valid boundary values.
  • The audit trail captures initial entry and every modification with user, timestamp, old value, new value, and reason, and cannot be disabled by an end user.
  • Role-based access enforces real restrictions for every tested role, including a negative test of a revoked account.
  • Randomization/IRT and external data interfaces reconcile completely, with no duplicate or dropped records.
  • The export matches the specification biostatistics will actually use.
  • Every URS requirement traces to at least one executed test case with no orphans; see the Requirements/Test Traceability Matrix.

10. Summary and conclusion

<<FILL: statement that the study-specific EDC configuration is validated for its intended use, or the conditions/restrictions on release; signed by Data Management and QA>>

11. Attachments

  • Requirements traceability matrix (URS to test case).
  • Executed test evidence (screenshots, exports, audit-trail excerpts).
  • Deviation records and resolutions.
  • Edit check specification (referenced, not reproduced).

Filled specimen: TC-08 (audit trail: modification)

The following shows one executed high-risk test case, so you can see the evidence level expected. Details are illustrative.

FieldEntry
Test caseTC-08, audit trail modification capture
PreconditionSubject 0142 has a saved systolic blood pressure value of 128 mmHg on the Vitals page
Step 1As a site user, change the value from 128 to 138 mmHg and select the reason “Transcription error, corrected against source”
Step 2Save the page, then export the audit trail for this field
ExpectedTrail shows old value 128, new value 138, user, timestamp, and the selected reason, with the original value still viewable
ActualAudit trail export confirmed old=128, new=138, user J.Kim, timestamp 2026-08-05 14:12 UTC, reason “Transcription error, corrected against source”; original value remained visible in the trail
ResultPass
EvidenceAudit trail export AT-TC08; screenshot SC-TC08
Tester / QAJ. Kim / A. Ferreira, 05 August 2026

Because audit trail integrity is the control every downstream review depends on, this case is executed with full scripted rigor rather than sampled. A vendor’s platform-level Part 11 claim does not substitute for testing that the study-specific build actually behaves this way.

Common inspection findings this protocol prevents

  • An edit check described in the specification but never actually built or tested, discovered only when a bad value reaches the database.
  • A “validated” claim resting entirely on vendor documentation, with no evidence the sponsor tested its own study configuration.
  • Role-based access that looks correct in the specification but was never negative-tested (a monitor account that can, in practice, edit a field).
  • An audit trail that captures the new value but not the reason, or that can be suppressed under some administrative role no one thought to test.
  • A randomization interface tested once at build time and never re-verified after a mid-study configuration change.

How to adapt this protocol

  1. Scale test depth to your risk assessment; primary-endpoint, safety, and randomization functions get full scripted testing, low-risk reference fields can use lighter, documented coverage.
  2. Add test cases for any study-specific integration not listed here (imaging, genomics, wearables).
  3. Point section 7’s interface and export tests at your real downstream systems and the export specification biostatistics actually uses.
  4. If the study amends mid-conduct, re-run the affected test cases under change control rather than treating the original protocol as still fully representative.
  5. Confirm every cited regulation against the current source before execution.
Use madhadi.com as an app Full screen, works offline, one tap from your home screen.