Table of contents
At a glance
How do monitoring, audit and data review differ in GCP?
Monitoring oversees trial conduct and risks to participants and data. Audit independently examines selected processes outside routine quality control. Operational data review examines specific data and metadata against defined quality questions. Coordinate findings while preserving the different objectives, responsibilities and independence requirements. The trial and its risks determine scope. There is no universally prescribed review percentage.
- An independent audit neither replaces ongoing monitoring nor operational data review.
- Risk-based scope requires a specific justification, not a generic testing percentage.
- A single correction and the effectiveness of a process action address different questions.
The same data, different inspection tasks
Monitoring assesses during the review process, among other things, the protection of participants and the reliability of the data. A GCP audit is an independent, systematic evaluation of appropriate processes and documentation. In contrast, operational data review examines specific data and metadata against predefined quality questions. These activities can complement each other without being interchangeable.
Sources: [1]
Objectives, independence, and decision-making to be defined in advance
ICH E6(R3) is one of the most important quality control activities. The audit should be independent and separate from it. Independence means a clear demarcation from the audited activities; it cannot be derived solely from another job designation.
Monitoring also has specific requirements regarding the role. It is not an audit simply because an external party is involved. The following matrix outlines possible inspection tasks. It serves as an independent reference tool and is not a binding organizational plan for every company.
| Activity | Typical question | Important distinction |
|---|---|---|
| Monitoring | Is the inspection at the site carried out in accordance with the intended requirements? | May be performed on-site, remotely, or centrally; no automated full comparison of all data. |
| Operative data review | Have missing, inconsistent, or suspicious data been appropriately investigated? | Part of the ongoing quality process; not solely achieved through regular independent execution. |
| Audit Trail Review | Are relevant changes and temporal workflows traceable and plausible? | Testing scope to be defined on a study-specific and risk-based basis. |
| Independent Audit | Are the selected processes appropriate and are they reliably executed? | Independent of the validated process and routine quality control. |
Sources: [1]
Why a 100% data comparison is not automatically superior
The monitoring scope depends on risks and the trial's critical-to-quality factors. ICH describes centralised and site-based approaches. Article 48 of the CTR requires appropriate monitoring, with its extent and nature determined by the trial's characteristics. This does not impose universal 100% source data verification.
First, describe potential impacts on participant rights, data security, or reliable results. Derive data fields, time points, and control measures from these risks. A justified sample may form part of a concept; however, it is not sufficient for every process without further evaluation. Conversely, an extensive testing scope must not be used as a substitute for an undefined testing objective.
Plan audit trail review as a specific inquiry point
An audit trail includes indications of data creation and modification. The EMA expects a risk-based, audit-specific review and describes its ongoing, proactive application, provided no other justification is given. Therefore, exporting all protocol lines without defined analysis remains an inadequate review.
Appropriate questions may include: Are post-implementation changes to specific critical fields accumulating? Do data capture and transfer times align with the intended workflow? Are changes attributable to a specific individual and supported by clear, traceable justification? Identify which anomalies warrant a detailed investigation and ensure blind spots and confidentiality are maintained. Not every technically logged administrative action requires the same level of technical review.
Example: Six open fields are not six identical problems
In the exercise dataset, six expected time entries are missing. The operational review matches them against the intended sources and transfer documentation. Three values were locally entered but have not yet been transferred. Two values are actually absent from the intended source. One value is present but was incorrectly assigned to the wrong target field during import.
The six visible gaps have three different causes. A blanket request to fill them in would conceal those differences. Monitoring can investigate implementation at the site. An independent process audit can later assess the overall suitability of handovers and controls. An individual finding need not trigger its own audit, and a technical correction alone does not verify the whole process.
| Finding in this example | First specific testing assignment | Potential for further action required |
|---|---|---|
| 3 values were recorded but not transferred | Compare local sources and pending transfers. | Investigate transmission failure; assess relevant impacts at the site. |
| 2 values without specified source | Investigate the facts and assess allowable avenues for clarification. | Do not invent values; document handling of missing data in accordance with study specifications. |
| 1 Value in incorrect target field | Verify source-to-destination mapping and define affected data. | Correction and justification for the scope of additional assignment reviews. |
| Recurrent pattern despite completion | Assess the effectiveness of the actions taken. | Plan an independent review of the end-to-end process, taking into account appropriate risk. |
Find, data correction, and process improvement separate
A query can resolve a single point of uncertainty. A data correction can rectify the affected data record. Neither necessarily addresses why the error recurs or whether the existing control is adequately functioning. Therefore, for relevant patterns, also investigate the underlying processes.
Not every query automatically constitutes a CAPA. It would also be incorrect to address recurring or significant process issues solely through repeated individual corrections. Define escalation criteria based on significance, frequency, root cause, and impact. The selected corrective action should later be evaluated for effectiveness through a suitable observation or testing procedure.
Sources: [1]
Share results, do not mix inspection roles
Agree which findings data management, monitoring, the trial site and quality assurance should receive and who makes each decision. A shared findings list can help if the purpose, owner and status are clear. A monitor can flag a data anomaly without having to make every resulting statistical decision.
Audit findings can trigger improvements but do not replace ongoing monitoring until the next audit cycle. Similarly, a closed operational finding does not automatically indicate that a previously identified systemic issue has been permanently resolved. Therefore, document the closure at the level at which the respective question was raised.
Sources: [1]
Checklist for a verifiable review request
A good review planning is short enough for the application and concrete enough for comprehensible decisions. Use the points for a single data review as well as for coordination with monitoring and quality assurance.
- Objectives and relevant risks are identified.
- Monitoring, operational review, and independent audit are conducted as separate tasks.
- Scope, data sources, and where applicable metadata are defined.
- The auditor possesses the required knowledge, appropriate access, and necessary role segregation.
- Disorders lead to defined clarification and escalation routes.
- Changes remain source-tracked, traceable, and verifiable.
- Assess closure of the individual case separately from any process action.
- Results and open issues are incorporated into further risk-based planning.
Project experience and collaboration
Route a data finding for follow-up without duplicating work
With the responsible teams, we follow a sample finding from operational data review to possible escalation. Which question does Data Management resolve, when is monitoring needed, and when would independent process review help? An isolated, explainable deviation may warrant a focused inquiry. Recurring patterns may call for wider investigation. We agree ownership and feedback so the same finding is not handled repeatedly and necessary independent assessment is preserved.
Frequently asked questions about GCP monitoring and data review
Is a data review the same as monitoring?
Not necessarily. An operational data review can be part of data management or monitoring and has its own defined scope. Monitoring also covers trial conduct and relevant site processes. Identify the objective, scope and responsible people for each activity. The screen they use does not distinguish their roles.
Sources: [1]
When is a review considered an independent GCP audit?
When conducted as an independent, systematic review with appropriate authorization and clearly defined role boundaries. E6(R3) distinguishes auditing from routine monitoring and quality control. The use of an external service provider or the term "audit" alone does not establish this independence.
Sources: [1]
Does GCP require a complete comparison of all data with the source?
No. In clinical trials within the scope of Regulation (EU) No. 536/2014, the sponsor determines the extent and type of monitoring based on the characteristics of the trial (Article 48). ICH E6(R3) 3.11.4 guides the selection of monitoring approaches according to identified risks and takes into account on-site, remote, and central monitoring. Therefore, a complete source data verification for each data point is not a general obligation. Justify which controls address the key risks of your trial.
Is each sample automatically sufficient?
Which changes should be included in the audit trail review?
Changes and metadata relevant to data integrity and study progress, as defined by your risk-based, study-specific concept. This includes unexplained value alterations or suspicious temporal patterns. A one-size-fits-all treatment of every technical protocol line is equally unhelpful, and a complete disregard for relevant history is inappropriate.
Is an audit trail export sufficient evidence of review?
No. The export provides the material. An review traceability record additionally demonstrates what was assessed, the applicable criteria, and how anomalies were handled. Maintain the required reference to the inspected inventory and to the documented decision.
Sources: [4]
Must every data query trigger a CAPA?
No. Individual clarifications can be addressed within the established data process. However, significance, recurring patterns, and potential process root causes are relevant. If individual corrections do not adequately resolve the issue, appropriate further actions must be evaluated.
Sources: [1]
Can a monitor also perform the statistical evaluation?
Not solely by virtue of being the monitor. Define the authority and qualifications needed for the decision. A monitor can identify and escalate anomalies, while the actual assessment may involve data management, statistics or the trial site.
Sources: [1]
When should a recurring finding lead to a process investigation?
Consider it when the risk, recurrence or poor effectiveness of previous actions indicates that correcting one case is insufficient. Define what needs examining. This may require a more detailed operational investigation or an independent audit. Different findings can call for different responses.
Sources: [1]
How do we prevent duplicate inspections without gaps in validation evidence?
Review audit objectives, sources, responsibilities, and the use of existing results. A shared overview can reduce duplicate efforts. However, maintain the necessary independence of the audit and verify that an existing result actually addresses the same question and covers the same scope.
Sources: [1]
Sources and review date
Sources checked on 7 October 2026. ICH E6(R3) Principles and Annex 1 have applied in the EU since 23 July 2025. The final Annex 2 has been published, with an EU effective date of 15 January 2027. The 2023 EMA computerised systems guideline explains GCP expectations and is not an EU regulation. This article covers clinical trials of human medicines. The examples are our working materials. Applicable regional law and trial documents remain decisive.
- EMA/ICH: E6(R3), Principles and Annex 1, Step 5Principles and Annex 1 (Step 5), EU effective from 23 July 2025. Sections 3.10, 3.11.2–3.11.4, 3.16.2 and 4.2.2–4.2.5.
- EMA: ICH E6 Good Clinical Practice – current versionStatus page, accessed on 7 October 2026. Principles and Annex 1: EU-effective from 23 July 2025; Annex 2: EU-effective from 15 January 2027.
- EU: Regulation (EU) No. 536/2014 on clinical trialsConsolidated version as of 5 December 2022, version status verified on 7 October 2026. Article 48 on monitoring.
- EMA: Computerised systems and electronic data in clinical trialsFinal Guideline dated 9 March 2023. Sections 6.1.2 and 6.2.1–6.2.2.
Sources checked: .
Integrate validation objectives and system justification effectively
We support you with clear review assignments, appropriate data and metadata controls, and traceable handovers between functional areas, IT, and quality assurance.
- 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.

