Regulations and compliance · Practical guide

EMA’s 2023 guideline on computerised systems in clinical trials

A tested study platform can still apply a changed visit window at the wrong site. Trace standard software, study configuration and actual activation. The 2023 EMA guideline helps sponsors, sites and IT connect the intended trial workflow to suitable evidence and access arrangements.

Table of contents

At a glance

Which system and study evidence does the EMA’s 2023 guideline call for?

EMA’s guideline expects evidence for the system and the study-specific configuration. Annex 2 covers approved user requirements, a predefined configuration checked against the protocol and data management plan, traceability to tests, validation and test plans, documented results and an approved validation report. It also covers release before operational use, change control and periodic assessment. Use supplier documents where a sufficiently detailed assessment shows they cover the intended use, and supplement missing evidence.

  • Demonstrate standard software and study-specific configuration separately.
  • Protocol changes require a verifiable activation status.
  • Investigator site monitoring and access should not be solely delegated to the Sponsor.
  • Verify data significance and context during transfers and subsequent handovers.

1. Establish the scope and applicable version

Final guideline EMA/INS/GCP/112288/2023 was adopted on 7 March 2023 and dated 9 March 2023. The cover page gives an application date six months after publication. It replaces the 2010 Reflection Paper on electronic source data and concerns clinical trials with investigational medicinal products.

As of 6 October 2026, the ICH E6(R3) Principles and Annex 1 in the EU are effective from 23 July 2025. Annex 2 has been adopted but will become effective in the EU from 15 January 2027. Therefore, older references in the EMA guidelines to E6(R2) should be read in light of this updated status. EU GMP Annex 11 is not an automatic substitute for the clinical regulatory domain.

Sources: [1] [2]

2. From Study Decision to System Boundary

The scope encompasses more than one eCRF. It includes, among other things, the capture, transfer, and processing of clinical data, as well as processes that may influence participant protection and the reliability of results. Therefore, laboratory imports or the assignment of a visit can be equally relevant as a visible data entry form.

Describe the specific workflow: Where is the first required documentation generated, who supplements it, which automated step alters its significance, and who uses the result? From this arise requirements for data, interfaces, and roles. A simple list of product names does not illustrate these dependencies.

Own traceability chain for a study configuration
LevelQuestion to askRelevant evidence
Study protocol Which rule applies to which participants and trial sites? Approved site and documented scope of validity
Configuration What software configuration does this rule require? Versioned assignment of requirement and closure
Verification Do boundary and error cases function? Expected, actual result, and evaluated finding
Activation May the trial site use this version? Required Approvals and Activation Evidence
Data transfer Are value, identity, and context understandable? Defined export and appropriate reconciliation

Sources: [1] [3]

3. Combine supplier documentation and own configuration

Annex 2 of the EMA guidelines allows the use of evaluated, appropriate validation documentation provided by a supplier. However, the responsible party must understand the extent to which the supplier's documentation addresses its own requirements and intended use. Gaps or inadequacies may necessitate additional activities. A simple supplier confirmation does not fully validate the scope of the documentation.

Therefore, the standard version, study configuration, and customer-specific enhancements should be maintained separately. Differences between test and production environments shall be evaluated based on their significance. The platform functionality can be well justified for a new visit plan, while the specific daily boundary or site-dependent activation remains to be verified.

Sources: [1] [3]

4. Example: two allowable configuration states

The example is not a medical instruction. In protocol version V1, the visit window—including boundaries—is defined on study days 5 to 9. Version V2 extends this to days 4 to 10. For inspection site A, the required prerequisites for V2 have been met and the configuration is activated; inspection site B is still operating under V1. The study day is provided as a clearly defined whole number during the exercise.

Four separate test data sets now evaluate the assignment. Set A at Day 4 falls within the active window. Set B at Day 4 and Day 10 falls outside; Set B at Day 5 falls within. Thus, two out of four data sets are classified as within the window. A global activation of V2 would evaluate all four as within, thereby incorrectly classifying P02 and P04.

An external visit must not simply disappear from the record. The defined expected outcome is accurate documentation and the prescribed clarification procedure, not a prohibition of necessary medical care or the invention of an alternative visit date. Historical data will not be unreviewed and rewritten according to a new logic.

Expected results linking the test site's version to the correct limit
Test data setTrial site and versionStudy DayExpected classification
P01 A / V2 4 Within 4–10
P02 B / V1 4 Outside 5–9; Clarification
P03 B / V1 5 Within 5–9
P04 B / V1 10 Outside 5–9; Clarification

Sources: [1] [3]

5. Demonstrate specific access and reviewer control

The EMA guideline expressly addresses trial site control of relevant data. A central server on which the sponsor alone decides who may access the data does not automatically provide that control. Suitable arrangements for independent copies or justified technical alternatives must preserve the required independence and access.

Test normal use and prohibited actions. Can a user see data from another trial site? Is blinding preserved? What happens when a role changes? Role names such as Monitor or Administrator are insufficient evidence that effective permissions fit the trial.

Sources: [1] [3]

6. Data transfer requires more than a line count

A transfer must retain value and significance. Therefore, identifiers, units, time references, versions, and relevant change context should be present in an appropriate review. Identical row counts do not preclude duplicates or misassigned participant allocations. The specific scope shall align with the data purpose and risk.

Edit checks are controlled functions. A logically plausible input value may still be factually incorrect; a technical query does not automatically correct the observation. Describe who verifies the source, who is authorized to perform a justified correction, and how the original entry and subsequent changes remain traceable.

Sources: [1] [3]

7. Plan change, disruption, and study end in advance

For a protocol change, identify the affected participants and trial sites and its effect on existing data. Retesting may include dependent functions that are themselves unchanged. A new version name does not establish correct activation.

You should also plan for system outages and subsequent handover. The responsible reviewer must be able to appropriately use and retain required data. Prior to revoking access, an appropriate handover must be verified; a ZIP file without clear context is insufficient alone. Specific deadlines shall be derived from the actually applicable requirements, not from a universal software rule.

Sources: [1] [3]

8. Checklist for Sponsor QA, Data Management, and IT

Start with a critical trial workflow and the people who perform it. Clinical data management describes the rule, the trial site explains its practical use, and IT and the supplier explain implementation. QA connects evidence to the decision. A small project still needs clear ownership for each important boundary.

The following list serves as an independent reference tool. It is applicable for a new study as well as for the targeted evaluation of a protocol amendment.

  • Clearly define the protocol status, study segment, and system boundary.
  • Assign standard functions and study configurations separately.
  • Verify supplier documentation against specific coverage.
  • Edge cases, incorrect roles, and unexpected data to test.
  • Consider the required prerequisites prior to activation for each testing site.
  • Trace transfers, audit trails, and justified corrections.
  • Check investigator oversight, downtime procedures and subsequent handover.
  • Document actual usage decisions and any open findings.

Sources: [1] [3]

Project experience and collaboration

Link the platform release to the configuration for the specific study

With study owners, Data Management and the system team, we trace a selected protocol rule through to configuration and its test. This shows what supplier documentation covers and what arises from your particular study setup. An unchanged standard function calls for a different assessment from a new calculation or a site-specific version change. We support closure of the gaps and a documented release decision. Responsibility for the study and its data remains with the designated parties.

Frequently asked questions about EMA guidance on computerised systems in clinical trials

Is the guideline only applicable to electronic CRFs?

No. Your scope includes additional computerized systems and processes that may influence clinical data, participant protection, or result reliability. This can include data transfers, laboratory integrations, and study-specific documentation systems. The specific activity is decisive, not merely the product name.

Sources: [1]

Does ICH E6(R3) automatically replace the EMA technical guideline 2023?

No. ICH E6(R3) and the EMA Computerised Systems Guidelines address different levels: ICH E6(R3) outlines the overarching GCP responsibilities and data handling requirements, while the 2023 guideline provides deeper insights into computerised systems and electronic data. ICH E6(R3) does not supersede the guideline. Review the requirements together and additionally determine which version applies to your study and region.

Sources: [1] [2] [3]

Does ICH E6(R3) Annex 2 already apply?

As of 6 October 2026, the EMA has approved the guidelines; however, they become effective within the EU from 15 January 2027. The Principles and Annex 1 have already been effective in the EU since 23 July 2025. Always specify the exact section and its current status rather than referring generically to E6(R3) as a general timeframe.

Sources: [2]

Does the platform provider's validation suffice?

It can provide appropriate evidence, but must be evaluated. Study-specific configurations and actual usage may require additional assessment. Particularly, protocol rules, interfaces, and roles should be visibly linked to their respective documentation.

Sources: [1] [3]

Does every study require the complete repetition of all baseline tests?

Not automatically. Suitable and appropriate documentation can still be utilized following evaluation. However, the specific study configuration and changes still require sufficient substantiation. A blanket acceptance and a blanket full reperformance are both inferior to a justified scope.

Sources: [1] [3]

Can a visit outside the window become technically invisible?

In this article example, the actual visit remains recorded and appropriately documented. This example illustrates software behavior, not a medical procedural instruction. A system should not replace real observation with artificially generated entries.

Sources: [1] [3]

Must an Amendment be immediately activated at all testing sites?

Match the scope to the protocol and prerequisites at each trial site. ICH E6(R3) addresses activation of trial-specific systems and updates once the relevant approvals are in place. A global technical deployment does not establish suitability for every site.

Sources: [1] [3]

Why does an export with the same number of rows not suffice?

The number of records identifies certain errors but does not ensure complete identity and semantic accuracy. A missing data entry may be numerically balanced by a duplicate. Therefore, verify appropriate keys, content, and required context of the transferred inventory.

Sources: [1] [3]

Can the sponsor alone decide how the trial site accesses its records?

The auditor's role and control over relevant data must be ensured separately. Agreements, hosting, and transfer responsibilities must address these requirements. A access that solely depends on a later, voluntary release by the sponsor does not adequately resolve the issue.

Sources: [1] [3]

Which documents should small study teams prioritize compiling first?

Start with protocol status, critical requirements, configuration, vendor documentation, and appropriate tests. Expand to include roles, activation, and data flows. From this, it becomes evident which evidence is already provided and where specific gaps exist; a general standalone handbook alone does not deliver this mapping.

Sources: [1] [3]

Sources and review date

Source date: 6 October 2026. EMA Guidelines 2023, together with ICH E6(R3) Principles and Annex 1, effective in the EU from 23 July 2025. Annex 2 will become effective in the EU from 15 January 2027. Case numbers are derived from internal work aids.

  1. EMA: Guideline on computerised systems and electronic data in clinical trialsEMA/INS/GCP/112288/2023; Document 09.03.2023, final acceptance 07.03.2023; Application six months after publication
  2. EMA: ICH E6 Good Clinical Practice – Status of Adoption and ApplicationRetrieval date: 06.10.2026: Principles / Annex 1 current as of 23.07.2025; Annex 2 EU implementation 15.01.2027
  3. ICH E6(R3): Principles and Annex 1 – EMA Step 5EMA/CHMP/ICH/135/1995; EU Application 23.07.2025

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Traceable link between study processes and system validations

We support you in defining your system usage, providing appropriate evidence, and establishing clear responsibilities among functional areas, QA, IT, and external service providers.

  • Computer system validation →

    Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

Send a project enquiry Customer testimonials →