Regulations and compliance · Practical guide

GCP source data: access and control at the investigator site

For GCP source data at a trial site, an investigator needs appropriate access to relevant information during the trial and arrangements for required records afterwards. A sponsor portal can provide access while leaving control of the data unresolved. A worked example connects ongoing access, blinding and the eventual handover.

Table of contents

At a glance

How can the investigator site retain the required access to and control over source data?

Define which source data and essential records the investigator site needs, who controls them and how they remain available during and after the trial. Relevant data must not be solely under sponsor control. Plan suitable independent copies or justified alternative controls, plus a complete, understandable handover. Give the site enough opportunity to review the data before access is withdrawn. Assess blinding and applicable retention requirements separately.

  • Access and actual data control differ.
  • Determine the required content and blinding limits prior to handover.
  • Connect account closure with verified, usable transition.

1. Assign source data to the study context

The article concerns clinical trials of human medicines. ICH E6(R3) Annex 1 addresses investigator records and responsibilities, data and computerised systems. The 2023 EMA guideline gives further detail on electronic systems and trial site access to data.

Principles and Annex 1 are effective in the EU from 23 July 2025. Annex 2 will become effective according to the current EMA status on 15 January 2027. For the access scenario discussed herein, the already effective foundations are applied. A general transfer of GMP Annex 11 requirements or a blanket retention period to all study data does not apply.

Sources: [1] [2] [3]

2. Determine which data the trial site needs

Document pre-emptively identify the source data that arise, where they are recorded, and how they are accessible. This may include internal records, central laboratory results, or data directly entered by participants. Particular relevance lies in determining which information influences the assessment of suitability, treatment, or safety.

Not every party requires unrestricted access to all study data. Blinding and protocol-based exceptions must be considered. Therefore, an access concept should clearly define which user receives which information at any given time. The necessary control over one's own and key records must not be substituted by a general reference to the sponsor portal.

Sources: [1] [3]

3. Distinguish access, copy, and control

A login may work temporarily while the trial site lacks sufficient control over its records. The EMA guideline addresses the risk of relevant source data being solely under sponsor control. Use suitable independent copies or justified technical alternatives to prevent that situation.

The internal matrix serves as a basis for evaluating a specific model. It is not a rule mandating that every file must always be dynamic or immutable. What is essential are content, meaning, required metadata, availability, and justified control throughout the relevant period.

Own work matrix for data access and control
SituationWhat a login alone does not clarifyCoordinated Control
Ongoing data capture Can the Sponsor control the data exclusively or arbitrarily revoke access? Appropriate independent data source or justified alternative measures
Central Laboratory Does the investigator receive relevant findings in time? Immediate access in accordance with the protocol
Blind information Which contents may still not be visible? Roll and time-based access restriction
Read-only Phase Are changes, relevant history, and context understandable? Required data view and subsequent batch release determination
After study completion Is recording retained usable without active service access? Appropriate handover, review and continued retention

Sources: [1] [3]

4. Ensure availability during the study

ICH E6(R3) specifies the need for the investigator to have timely access to relevant data, including data from external sources. This is particularly important when information may affect participant confidentiality or safety. A subsequent archive copy does not resolve an existing access gap that existed during the study.

Therefore, also plan to deal with failures, blocked accounts and changes of employees. The concrete solution depends on the system and the importance of the data. A replacement route does not have to replicate all comfort functions, but should allow the necessary action. Data protection and blinding remain to be observed even in the event of a malfunction; an uncontrolled common account would not be an appropriate standard solution.

Sources: [1] [3]

5. Example: Transfer prior to account closure

At close-out, a trial site expects a copy of the records for twelve participants. The agreed scope includes the relevant electronic data entered, subsequent changes and required attachments. The check finds twelve participant folders, but one is missing a relevant laboratory attachment. The folder count alone cannot establish completeness.

The site also verifies whether it can access and assign the data without the previous portal access. The table describes an individual intended workflow. It does not establish a universal export format for all studies.

Example: completed handover matrix
StepFind or specified conditionDecisionEvidence needed
Compare contents 12 participant order books available; 1 relevant annex is missing Handover is still incomplete Identify and supplement missing content
Verify assignment The attachment must belong to the correct participant and timing. File name alone is not sufficient Control content and context
Assess usability Release at intended work location without opening the portal Assess readability and required context Result of practical testing
End access Investigation of complete release documentation still open Keep access in place until the review is complete Approved transfer following an appropriate procedure
Ongoing operation Storage responsibility and access rights regulated upon handover Maintain required availability Accountability and appropriate storage documentation

Sources: [1] [3]

6. Determine the required output based on content

A PDF file may be suitable for certain information. However, it is not automatically a complete copy of all required electronic information. Assess which data, relevant history, metadata, and relationships are needed. Conversely, every output does not need to replicate all functions of the original system.

Before withdrawing read-only access, the EMA guideline describes providing a complete, understandable copy of the required data, including the relevant audit trail context. The trial site should not need advanced technical skills to use it. Agree the format and test usability before shutdown.

Sources: [3] [1]

Further reading:

7. Explicitly control exit and transition procedures

The agreements should clarify which party provides each deliverable and ensure that the investigator site maintains control over the monitoring. For data hosted by providers, an appropriate transfer procedure must be established; the EMA explicitly excludes exclusive sponsor control. A change of provider or termination of the contract shall not trigger this issue.

Plan for a change of responsible person or institution. ICH E6(R3) addresses control and retention of essential records and communication of changes to those arrangements. Determine the retention period from the applicable framework and the responsibilities agreed for those records. A single period for every type of data would be too imprecise.

Sources: [1] [3]

Further reading:

8. Access Approval Checklist

The transition is considered complete when the required data are available in an appropriate format, their use has been reviewed, and subsequent responsibilities are clearly defined. A shipping confirmation for a download link alone does not constitute evidence of this. Similarly, the number of folders does not substitute for verification that each contains the required contents.

The following internal assistance document can help structure the closing phase. It leaves the study-specific data selection open, while requiring a clear decision on that matter. Thus, ongoing access and subsequent retention are treated as a continuous process.

  • Data source types, storage locations, and responsible parties defined.
  • External relevant data made available in a timely manner.
  • Blind spots and roles in the access concept are considered.
  • Exclusive sponsor control effectively avoided.
  • Required output including context and history agreed.
  • Completeness, assignment, and practical usability verified.
  • Account closure and further responsibilities are only traceable thereafter.

Sources: [1] [3]

Project experience and collaboration

Plan access through to a usable handover

With study owners and the system team, we identify the data the investigator site needs during the trial and the records it must retain afterwards. Controlled examples test access, blinding, complete handover and later retrieval. An appropriate independent dataset may need a different technical solution from an active portal. Before access ends, we support acceptance of the content handed over. A successful download does not establish completeness.

Frequently asked questions about GCP source data access and control at the investigator site

Is an active portal access available for data control?

Not automatically. A login is a means of access. The trial site must retain the required control and availability of its records. Check whether the sponsor alone controls the data and what happens when access is withdrawn.

Sources: [3] [1]

Can the inspection site have access to all study data?

No. Access is based on responsibilities, relevant data, and the protocol. Blinding may limit certain information. However, the required data for participant monitoring, safety, and own records must remain available in an appropriate manner.

Sources: [1] [3]

Are central laboratory test results also relevant?

Yes, if they are important for the assessment of suitability, treatment, or safety and for the required evidence. ICH E6(R3) considers timely access to relevant external data. A copy provided only after study completion cannot substitute for this ongoing need.

Sources: [1]

Must every copy be dynamic?

No. A static copy may suffice for static content if it preserves the required information, including context, content and structure. For dynamic files, the EMA calls for the dynamic properties of the original to be captured as well (2023 guideline, section 6.5). Test what the trial site will later need to understand and reconstruct for the actual data type.

Sources: [1] [3]

Is PDF automatically an appropriate complete copy?

No. Verify that the required data, relationships, and relevant changes are clearly included. A concise PDF report may omit essential detail information. The format alone does not guarantee completeness or sufficient future usability.

Sources: [3]

When may the read-only access be terminated?

The EMA guideline combines withdrawal with a pre-provided complete, understandable copy and sufficient time for testing it. Access should not be withdrawn before completion of this review. Plan content, format and acceptance therefore before the shutdown date.

Sources: [3]

Why are twelve participant logs insufficient in the example of this article?

The number indicates only that twelve folders are present. A relevant annex may still be missing or incorrectly assigned. The reconciliation must verify the agreed-upon required contents and their interrelationships. Quantity and content control address different questions.

Sources: [1] [3]

Can the Sponsor be the only controlling party?

The EMA guideline limits control over relevant source data to the sponsor alone. Appropriate independent copies or justified alternative technical measures should be considered. The specific solution must be documented and actually enable the required level of control.

Sources: [3]

Does a general retention period of 25 years apply to all data?

No. The 25 years do not apply to all data. For a clinical trial of a human medicine within Regulation (EU) No. 536/2014, Article 58 requires the investigator and sponsor to retain the contents of their trial master files for at least 25 years after the trial ends, unless other Union law requires longer. National law governs patient medical files. Classify the records before determining the retention period.

Sources: [4] [1]

Who should perform the practical handover verification?

Involve people who can assess the content and how it will be used later. Technical support can assist. Define responsibilities in the applicable procedure. A successful download does not establish that the trial site can use its required records.

Sources: [1] [3]

Sources and review date

Source date: 7 October 2026. Regulatory requirements, guidelines, and industry practices must be categorized separately. Tables, figures, and case examples are to be considered standalone reference materials; they do not constitute completed product testing.

  1. EMA: ICH E6(R3), Principles and Annex 1Principles and Annex 1; in force in the EU since 23.07.2025.
  2. EMA: Status ICH E6 Good Clinical PracticeAs of 07.10.2026: Principles/Annex 1 in force; Annex 2 to become effective from 15.01.2027.
  3. EMA: Computerised Systems and Electronic Data in Clinical TrialsEMA/INS/GCP/112288/2023, 09.03.2023; Guidelines for clinical trials.
  4. EU: Regulation (EU) No. 536/2014 on clinical trials on medicinal products for human useConsolidated version as of 5 December 2022, reviewed against the current EUR-Lex version on 18 October 2026. Article 1 on scope of application and Article 58 on archiving of the Master File and on the demarcation of patient records.

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Jointly develop the specific evidence process

We support the investigational site, sponsor, and IT in planning specific data access, required controls, and a practical transition following study completion.

  • 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 →