Table of contents
At a glance
What should continuity arrangements for critical systems achieve under Annex 11?
Annex 11 section 16 addresses continuity of critical processes when a system fails. Document and test suitable fallback arrangements, with activation time based on risk and process needs. Define permitted activities, required information, responsibilities and return to normal operation. A backup supports restoration but is not a complete continuity procedure. Where prerequisites are missing, a controlled stop may be appropriate for particular activities.
- Plan the critical process including its information dependencies.
- Alternate procedures require authorized activities and clear boundaries.
- The retry must consolidate intermediate records and electronic states.
Align continuity with the critical process
Annex 11 January 2011 will continue to be listed in the Human Medicines GMP section on 7 October 2026. § 16 refers to systems that support critical processes. A possible arrangement is the passage of manual or alternative systems without requiring a single technical solution.
Start by asking what is actually at risk during an outage. Is it ongoing process control, access to the current material status, or the recording of an already initiated activity? These distinctions determine which work can proceed and the conditions that must be met. The general desire for rapid availability is still not a substantiated process requirement.
Failure scenarios and dependent information determine
An application may be accessible while its identity service or a critical interface is unavailable. Conversely, the user interface may not be available, even though data remains securely stored. Therefore, describe plausible scenarios and the respective capability gaps, rather than relying solely on the state "IT down."
Consider shared dependencies as well. An alternative procedure that relies on the same failed file repository for its forms may not be available. Similarly, a printed list provides limited value if it does not reliably reflect the current status. The suitability of backup information must align with the specific risk and timing context.
Define alternate operation with authorized activities
Define activation, authority, required resources, and boundaries. What information must be available, who is authorized to initiate the alternate procedure, and when is escalation or suspension required? The required time to achieve usability shall be derived from the process. Annex 11 does not specify a universal time threshold for this.
Our own matrix distinguishes prerequisites and stop points. It is not a recommendation to continue working on paper in every case of failure. A manual approach is only helpful when it adequately captures the required controls. For example, if reliable assignment of a material is missing, further work may lead to a false sense of security.
| Process requirement | Prerequisite for continuation | Limit or stop point |
|---|---|---|
| Use materials | Reliable identity and authorized status | Status or assignment unresolved |
| Document the procedure | Controlled accessible documentation and responsible person | Recording not unambiguously assignable |
| Availability of quantity | Known starting material and recorded intermediate movements | Availability is no longer reliably determinable |
| Investigate and resolve disturbance | Named technical and professional responsibility | Unclear impacts on data or process |
| Resume normal operation | Comparison of Intermediate Processes and System Status | Discrepancies not yet evaluated |
Example: Exercise on inventory during a system outage
Before the outage, exactly 100 units of batch B-010 are available with confirmed permission for use. An approved fallback procedure allows documented issues from that known stock only. Transaction N-01 issues 12 units and N-02 issues 8 units. The fallback in this example does not permit new status decisions.
The independent requirement states: 100 − 12 − 8 = 80 units remain. Following a restart, both processes must be accounted for exactly once. The table describes expected results from a planned exercise, not a recommendation for a general quantity limit.
| Test step | Expected result | Error that should be visible |
|---|---|---|
| Start | 100 units; Batch and status confirmed | Unknown or outdated starting version |
| N-01: 12 units | 88 remain; process assigned | Withdrawal without clear documentation |
| N-02: 8 units | 80 remain; second step assigned | Mix-up or omitted movement |
| Transfer back to the production system | N-01 and N-02 each considered once | Double booking of N-01 would result in 68 instead of 80 |
| Closure | Electronic status and interim copies are consistent | System start is prematurely considered as process completion. |
Intermediate records are controlled.
Plan how records can be reached, identifiable and protected against confusion. For paper, for example, this means suitable forms, clear process references and comprehensible corrections. In the case of electronic tools, their actual access options and data paths are to be considered.
Chapter 4 addresses real-time documentation and controlled changes. For the backup operation, this leads to a practical requirement: the subsequent takeover must not obscure the original temporal and professional relationship. Document the actual process and distinguish it from the later entry into the restored system. A retrospectively prepared summary report does not automatically clarify these differences.
Plan the retry as a separate transition
Please clarify upfront who assesses the technical status and who evaluates the professional continuity. Partially processed messages or bookings must not inadvertently re-impact during re-entry. For reconciliation, appropriate references and knowledge of the most recent reliable system state are required.
Therefore, in this example, each intermediate step is compared against the electronic inventory record. A visible remaining balance of 80 alone would not constitute complete verification, as two erroneous movements could accidentally cancel each other out. The individual transactions and their respective assignments remain critical. Only after a reviewed concordance is a decision made to resume the established normal workflow.
Align exercises to weakness identification
A review can clarify roles and responsibilities; it does not automatically demonstrate that forms, permissions, or recovery functions operate effectively in practice. Select an appropriate combination of approval and actual testing. During this process, the production environment must be protected against unintended interventions.
Intentionally vary relevant conditions: absence of representation, uncontactable backup location, or a partially booked process. Document observations and assess their significance. A training error is an opportunity for targeted improvement, not a justification for retrospectively adjusting the expected outcome based on the observation. Changes to the procedure may require appropriate revalidation evidence.
Anchor responsibility and maintenance in operation
An emergency plan must align with the current system and existing organizational structure. Changes to interfaces, data repositories, personnel, or work processes may alter its assumptions. Therefore, its maintenance should be integrated with the change management process and periodic review, rather than being retained as a static document.
For the initial evaluation, a clearly defined and critical scenario is sufficient. The department and QA determine whether continuation is authorized, while IT outlines outage and recovery options. The following checklist supports this joint review process. It does not constitute a guarantee that a specific alternative procedure is officially approved or suitable for every disruption.
- Critical processes and specific failure scenarios described.
- Dependencies of the alternative procedure have been reviewed.
- Activation, responsible roles, and stop points defined.
- Intermediate records and temporal assignment prepared.
- Return-inclusion including partial reprocessing and duplicate booking is governed.
- Appropriate training, examination, and maintenance reasons determined.
RTO and RPO: Process objectives verified with a calculated case scenario
RTO defines the target time to restore operational capability as specified in the current incident response concept. RPO defines the acceptable time gap between the restored data state and system outage. Both values are derived from the specific process context. Annex 11 does not specify universal timeframes in hours or minutes. Even a technically rapid restore may result in a data set that is not sufficiently current from a functional standpoint.
In the example, the last valid backup is at 08:00, and the system failure begins at 10:00. Without further restoration options, data loss of up to two hours occurs. If the process only tolerates a 15-minute downtime, the recovery objective is not substantiated. Additionally, it must be assessed when actual registration, interface connectivity, status synchronization, and authorized resumption activities are truly available. A simple database restart does not necessarily equate to full operational recovery.
| Characteristic | Example | To be evaluated statement |
|---|---|---|
| Last usable version | 08:00 Uhr | Which data are missing thereafter? |
| Failure onset | 10:00 AM | Up to two hours of data gap |
| Own RPO Objective | 15 minutes | Not supported solely by this safeguard |
| Technical Start | Database is back online | Not yet fully operational |
| Technical Alignment | Status and outstanding reconciliations verified | Approved further processing confirmed |
| RTO Measurement | Defined start up to appropriate endpoint | Total required reprocessing effort |
Sources: [1]
Conduct the emergency drill at the actual end point
Define prior to the exercise the conditions under which the process is considered professionally successful. If the process requires current block statuses, specific interfaces, and authorized users, these conditions shall be incorporated. Review interim manual records and their controlled transfer. A seemingly rapid reconciliation may later lead to duplicate entries or inconsistent statuses if the reconciliation is omitted.
Document measured times and data states separately from target values. A target is not a previously validated result. Findings are reviewed and addressed together with responsible parties and improvements. Also assess dependent services, personnel availability, and access rights. The emergency procedure should enable continuation of the process within authorized limits; it does not need to restore all comforts immediately. The required functions shall be professionally justified and defined.
Project experience and collaboration
Plan for the outage, interim records and return to normal operation
What happens to transactions processed manually during an outage? We examine a defined scenario with Operations, QA and IT. Alongside restoration and application availability, we check reconciliation of interim records, duplicate processing and outstanding decisions. A short outage may need a different response from an extended period of fallback operation. In each case, your procedure should identify who assesses the state of the business process and decides when it can resume. A planned rehearsal can expose handover gaps that a server check misses.
Frequent Questions on Emergency Operation and Resumption
Is Business Continuity the same as Backup?
No. A backup supports data recovery. Business continuity additionally addresses how critical processes are supported and controlled in the event of a disruption and how they return to normal operations. Personnel, information, alternate records, and authorized activities can be just as essential as technical recovery.
Sources: [1]
Must manual follow-up be performed for each failure?
No. A manual workaround is a possible measure, not a general obligation. If necessary information or controls are missing, a structured pause in specific activities may be appropriate. The decision must align with the specific process risk and the pre-established approach.
Sources: [1]
What is the required revalidation period specified in Annex 11?
§ 16 does not specify a general time limit. The time required to activate alternative arrangements depends on risk and the extent of process support. Therefore, justify your specific requirement and assess whether available resources and responsibilities are actually capable of fulfilling it.
Sources: [1]
Does a printed status list serve as a valid alternative?
Only if it is reliable, available, and sufficiently up-to-date for the intended use. Specifically assess temporary status changes and movements. A readable list alone cannot substitute for an unclear material usage authorization, merely because it was printed prior to system failure.
Must the alternative process be validated?
The suitability of specific provisions must be adequately documented and validated. Risk-based evidence should be provided for actual manual or technical steps. A specific document label alone does not confirm that the alternate procedure is reliable and traceable.
Sources: [1]
How do we prevent duplicate postings?
Assign appropriate procedural references and reconcile intermediate records with previously processed system states. In the example provided, N-01 and N-02 may each only occur once. The specific technical or organizational implementation must align with your application and be validated during routine practice.
Is a meeting considered an adequate emergency test?
She can evaluate decisions and roles, but cannot provide evidence of an unmonitored technical function. Select the training scope based on the number of questions to be answered. Where actual availability or restoration is critical, appropriate practical evidence is required rather than solely a documented discussion.
Sources: [1]
How are corrections to emergency records handled?
Maintain original information and context in an appropriate controlled procedure. For paper records, refer to Chapter 4 for clearly identifiable original information, date, and signature, together with the rationale if applicable. For electronic records, consider the relevant controls for changes and traceability.
Who restores the normal operation?
The emergency procedure must specify who is authorized to make decisions. A possible allocation could be: IT verifies restoration and technical status, the process owner evaluates the technical data reconciliation, and QA assesses relevant deviations. This serves as an organizational guidance, not a mandated universal list of signatures; Annex 11, Section 2 requires close collaboration and clearly defined responsibilities. The mere restart of a server does not establish that the process can continue operating with complete data.
Sources: [1]
When must we update the emergency plan?
If relevant assumptions change, for example data pathways, operating procedures, backup measures, or responsibilities, a supplementary, regular review may be appropriate. A universal test frequency is not specified in § 16; therefore, maintenance and validation activities should be justified based on risk and actual changes.
Sources: [1]
Does Annex 11 specify a fixed RTO or RPO?
Annex 11 does not specify a universal restart time or data loss threshold applicable to all systems. Derive appropriate objectives from process characteristics, criticality, and dependencies and assess their achievability. Document the specific final business state that constitutes restored operational capability.
Sources: [1]
Why might a faster restore still fail to meet the RPO objective?
The recovery time and currency of the restored data set are distinct attributes. A rapidly restored database may be based on an outdated backup. Therefore, separately assess the most recent viable data point and any necessary supplemental data transfers. Technical success alone does not establish adequate data currency.
Sources and review date
Source date: 07.10.2026. The relevant area here is human medicinal product GMP. EudraLex incorporates Annex 11 and Chapter 4 as of January 2011; the consulted draft from 2025 is not considered a current regulatory requirement. The completed examples and tables are internal work aids and do not constitute official template documents.
- EU GMP Annex 11: Computerised SystemsRevision January 2011; Effective date 30.06.2011; Access date 07.10.2026. Citations: §§1–2, 4.3, 7.2, 10–13, 16
- EU GMP Chapter 4: DocumentationRevision January 2011; Effective date 30.06.2011; Access date 07.10.2026. References: §§4.1, 4.3–4.5, 4.8–4.10
- PIC/S PI 041-1: Good Practices for Data Management and IntegrityPI 041-1, 01.07.2021; Access date 07.10.2026. Citations: §§9.1.5–9.1.6, 9.4; §9.9 Storage and Recovery
- European Commission: EudraLex Volume 4Status as of 07.10.2026: Human medicinal product section implements Annex 11 January 2011 and Chapter 4 January 2011; Access date 07.10.2026. Relevant provisions: Human medicines / Annex 11
- European Commission: Consultation 2025 on Chapter 4, Annex 11 and Annex 22Status as of 07.10.2026: closed; 07.07.–07.10.2025. Occurrences: details; revision of Annex 11; links explicitly marked as draft guidelines. Serves solely for status demarcation and is not considered an authoritative source of requirements.
Sources checked: .
Address the specific control and verification requirements
Daniel Herrmann Consulting supports you in adapting this procedure to your system portfolio, evaluating gaps, and preparing appropriate validation evidence. The scope and level of involvement are determined by your actual process.
- 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.

