Table of contents
At a glance
When does an IT incident become a GxP deviation or require CAPA?
An IT incident initially identifies an operational event. Annex 11 section 13 calls for reporting and assessing all incidents, not just system failures or data errors. Assess the effect on data, processes and decisions and determine the next quality-system action. For critical incidents, investigate the root cause as the basis for corrective and preventive measures. Technical restoration and quality investigation can close at different times.
Scope and Applicability
Terminology and escalation procedures may be company-specific. This contribution proposes a clear separation of responsibilities; it does not require initiating a full CAPA procedure automatically for every minor deviation.
Three interconnected but distinct responsibilities to understand
An IT incident addresses a technical disruption and the restoration of a service. A GxP deviation assessment evaluates potential impacts on data, processes, and quality. CAPA addresses evaluated root causes and appropriate actions to prevent recurrence. These tasks may relate to the same event, but are not completed through a single technical closure status.
Annex 11 section 13 requires reporting and assessment of all incidents, including system failures and data errors. For critical incidents, identify the root cause as the basis for corrective and preventive action. Chapter 1 and ICH Q10 provide the quality framework. Define implementation in your QMS. This does not mean every minor IT ticket requires three separate documents.
Identify a handover point between IT and the functional department. The technical staff should determine when a potential GxP impact may need to be communicated. The functional department requires information regarding affected systems, time period, and data. Unclear responsibility may result in a restored service being deemed a fully resolved incident.
Maintain a shared incident record with clear questions
Link technical and professional information through a unique reference. This can be achieved via a tool or through appropriate cross-references. It is essential that required tasks do not become disjointed. An closed IT ticket may reference an ongoing data investigation rather than asserting its completion.
Document assumptions explicitly. If you do not yet know when an error began, that affects the scope of the investigation. A successful recovery at 14:00 does not establish that everything was correct before 13:00. Use available logs and data to narrow the period and document the rationale.
| Level | Guiding question | Own work result |
|---|---|---|
| Technical malfunction | What failed and how was it restored? | Timeline, known cause, and operational justification |
| Data impact | Which records may be missing or incorrect? | Justified evaluation of scope and content |
| Impact on quality | Which decisions or processes are involved? | Professional technical assessment responsibility |
| Action addressing the cause | Why could the incident have occurred or have remained undetected? | Appropriate correction and prevention |
| Effectiveness | Does the measure operate as intended in the intended operating condition? | Appropriate testing or observation documentation |
Example: Interface back online, results still missing
A connection between LIMS and ERP fails. IT restores it after two hours and closes the technical incident. However, some messages were not retransmitted. As a result, the ERP contains incomplete results, even though the connection has been restored.
The technical investigation compares expected operations with actual takeover in the affected period. Missing messages are treated by a regulated way. Check repeat and double processing risks. A flat-rate re-dispatch of all messages can create new errors if some have already arrived.
The root cause analysis additionally investigates why the failure or lack of recurrence went undetected. A corrective action may involve monitoring, defect remediation, or restart logic. The closure should distinguish between technical recovery, clearly defined data scope, and action status. A single green connectivity test does not validate all layers.
| Test case | Action | Expected finding |
|---|---|---|
| Compound | Verify the technical channel | Service available |
| Completeness | Compare expected and accepted processes | Missing data identified |
| Repetition | Conduct a scheduled retry | No unresolved duplication |
| Technical Application | Review affected ERP process | Correct result status applied |
| Detection | Test controlled similar failure scenarios | Enhanced control responds |
Determine the root cause and objective of the CAPA
The root cause analysis should extend beyond the immediate technical trigger when necessary for understanding the error sequence. For example, a full storage may have been facilitated by insufficient monitoring, unclear responsibility assignments, or inappropriate storage configuration. Simply releasing storage does not necessarily eliminate these underlying causes.
CAPA actions require a verifiable purpose. The implementation of a new SOP or training is not automatically effective. Define clearly what indicates that the root cause has been controlled and that a relevant recurrence is prevented or detected earlier. Link technical tickets and quality evidence through stable references.
Conduct root cause analysis extending beyond the most recent operator error
The immediately visible error is not always the root cause. A wrong input can be facilitated by an unclear interface, missing validation, or contradictory instructions. Investigate the actual conditions. A one-size-fits-all training session may be inappropriate if the procedure itself is practically unworkable.
Use appropriate methods whose depth aligns with the incident. A minor, well-defined deviation does not require the same level of analysis as a systematic data integrity investigation. Document the hypotheses that were evaluated and the evidence supporting the conclusions. A method name alone does not constitute a sound analysis.
Assess additional affected systems or time periods if a common root cause is possible. The same erroneous mapping component may impact multiple interfaces. Correcting a single message does not adequately address this cross-cutting issue. The scope is justified by the root cause, not arbitrarily set at maximum or minimum.
Actions and effectiveness evaluation relating to the root cause
A correction addresses the specific error. A preventive action alters the condition that enabled the recurrence. Identify the expected effect and evidence for each action. "SOP updated" only confirms the document, not the actual improvement in data flow.
Effectiveness assessment may utilize appropriate tests, real-life post-facto data, or a combination thereof. A controlled failure demonstrates detectability; real-world application demonstrates practical utility. The duration and scope shall be determined based on risk and event frequency. A universal observation period is not specified herein.
Assess adverse effects. An automated retry mechanism can prevent missing data, but may result in duplicate activities. Therefore, the measure must be evaluated within the complete relevant workflow. Effective CAPA addresses root causes without inadvertently introducing a new critical gap.
Operate, monitor trends, and align with strategy appropriately
Recurrent incidents can provide insights for structural improvements. A planned system decommissioning may constitute part of a sustainable solution. However, it does not replace the current risk mitigation and necessary data evaluation. Conduct immediate handling and long-term improvements as interconnected, distinct levels.
Provide technical and professional criteria to the relevant teams. An IT staff member should understand the information required by QA; QA should not infer the technical status of a system solely from a ticket status. A joint exercise simulating a system interface failure can practically demonstrate this handover.
Draft: Technical restoration and data cleansing to be completed separately
The table separates technical restarts from professional clarification. Prior to repeating any process, compare expected versus actually completed activities to ensure that a consolidated restart does not generate duplicates. Maintain both closures under the same incident reference. Recurring root causes may require additional corrective actions; such actions must not delay the immediate containment of the current risk.
| Incident field | Example entry | Purpose of the evidence |
|---|---|---|
| Technology | Connection failed during the specified period. | Operational status and restoration. |
| Data | Expected result messages during the period. | Completeness and content. |
| Quality | Decisions are based on the data. | Possible impact on the process. |
| Cause | Detection or retry was inadequate. | Action addresses actual condition. |
| Correction | Missing procedures are reviewed and corrected post-hoc. | No unresolved dual processing. |
| Effectiveness | A similar failure scenario will be appropriately detected. | Improvement has been practically demonstrated. |
Next steps
- Stabilize the disturbance and ensure the affected scope.
- Assess GxP implications in conjunction with the respective department and QA.
- Define causes and necessary corrective or preventive actions.
- Assess efficacy and link related supporting evidence.
Project experience and collaboration
Link service recovery and data investigation under one reference
When an incident occurs, IT and process owners investigate both the technical fault and the affected transactions. Which data have reached the target? What is missing? What might a retry duplicate? A limited outage may need targeted correction, while a recurring cause may warrant broader action. We record technical closure separately from confirmation that the business transactions are complete. The responsible Quality staff use those facts to assess the deviation and any CAPA.
Frequently asked questions about Incidents, deviations and CAPA
Can IT alone assess the product's efficacy?
Is training always an appropriate CAPA?
Must every IT incident trigger a CAPA?
Not at all. Evaluate possible GxP effect, cause, repetition and existing procedures. A clearly limited disorder can be adequately treated without extensive CAPA. A systematic or critical finding may require further measures. The decision should be comprehensible and not depend solely on the ticket type.
Can IT close the incident even though the data investigation is still ongoing?
The technical completion can be separately documented if its scope is clearly defined and the open technical task is reliably continued by the responsible party. Link both references and responsible parties. A closed technical status must not imply that all quality and data issues have also been resolved.
Why is training not sufficient for every root cause?
A failure may arise from an unclear interface, missing functionality, inconsistent procedure, or insufficient capacity. Training does not address all of these conditions. Investigate the actual root cause and anticipated impact of the corrective action. Completion of a training seminar alone does not validate improved data handling.
How do we differentiate a correction from a performance qualification?
The correction addresses the specific defect or condition. The effectiveness verification assesses whether the evaluated root cause is adequately controlled and whether the corrective action maintains its intended effect under normal operating conditions. Both may have distinct forms of evidence and timing. A technically repaired individual incident does not automatically confirm sustained effectiveness.
Sources and review date
Annex 11 covers incidents. Chapter 1 and ICH Q10 address root cause analysis and improvement. Our example incident record and tests show how to establish technical recovery and resolve questions about the data separately.
- European Commission: EU GMP Annex 11, Computerised Systems (2011)§ 13: Incidents and Evaluation; Linking to CAPA according to Q10 and Chapter 1. Current version as of EudraLex on 14.10.2026; no universal project or deadline specification.
- European Commission: EU GMP Chapter 1, Pharmaceutical Quality SystemQuality system, root cause analysis, CAPA and continuous improvement.
- EMA / ICH: Q10 Pharmaceutical Quality System, Step 5 (2008)Section 3.2.2: Root Cause Analysis, CAPA, and Effectiveness Evaluation. Specific action plans and exercises provided in the contribution are standalone methodological aids.
- PIC/S: PI 041-1, Good Practices for Data Management and Integrity (2021)Inspection Guidance on Data Integrity; in particular electronic data, transmission, access, and review.
Sources checked: .
Combine technical incidents and quality investigations
We support you in assessing the current situation, clarifying responsibilities, and implementing appropriate evidence within operations. The type of support that is most effective depends on your system, your team, and the specific objective.
- 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.

