Practical guides · Practical guide

Periodic review of computerised systems: scope, frequency and follow-up

A system may appear unchanged while its use, interfaces or owners gradually change. A periodic review brings those changes together with incidents and operating evidence. Define the scope, justify the review interval and assign actions for outstanding findings so the team can decide whether the system remains fit for use.

Table of contents

At a glance

How does a periodic review support a decision on continued operation?

Define the system and review period, gather relevant operational evidence and compare current use with the documented validated state. Use changes, incidents and gaps to identify actions. Justify the review interval and any triggers for an earlier review based on risk. Annex 11 does not specify a universal annual interval. Document the decision on continued operation and any conditions attached.

  • The review evaluates the current usage based on operational experience and existing evidence.
  • Intervals, scope, and premature evaluation require a justifiable rationale.
  • Backup, restore, role and supplier information are evaluated in relation to their specific system context.
  • Actions require owners, deadlines, and verifiable completion criteria.

What the Periodic Review is intended to address

The Periodic Review consolidates operational experience of a defined computerised system into a current decision. The key question is: Do existing evidence and controls still support the current intended use? This includes changes to workflows, data pathways, and responsibilities. An earlier approved version does not automatically address this question.

This guide addresses computerised systems within the EU GMP context for human medicinal product manufacturing. Annex 11, Section 11, specifies the periodic assessment of the validated state and GMP compliance. The Commission continues to maintain the January 2011 revision as the current version. For other GxP areas, the applicable basis shall be determined separately. Part 11 is aligned with the relevant FDA documentation and underlying regulations.

Separate system assessment, ongoing audit trail review, product quality report, and technical maintenance. These activities can provide information but relate to different subjects. The review also determines whether additional testing is required; it does not automatically imply a full revalidation of the initial validation.

Sources: [1] [5] [4]

Define the review scope before collecting data

First, define the system boundary, observation period, and cut-off date. Specify version, configuration, locations, interfaces, and functional scope. A simple product designation is insufficient when different modules or tenants support distinct processes. Explicitly document which shared services are being assessed elsewhere.

Recommended responsibility assignment: The process owner evaluates the technical applicability, the system owner compiles the supporting evidence, IT reports on operations and technical events. QA assesses the quality of the conclusion based on the operational procedure. Suppliers provide agreed-upon information. Management is involved when resources or fundamental operational decisions are required.

Create a list of documents in advance with the person responsible and the date. A missing report receives the status "not available" and a consequence for the evaluation. It is not treated tacitly as an inconspicuous finding. For an SME, this clear preparation is often more important than an extensive new report template.

Sources: [1] [3]

Justify the interval and treat events separately

Annex 11, Section 11 does not specify a fixed annual interval. ICH Q9(R1), Section 4.6 links the frequency of risk assessments to the risk level and takes into account new insights as well as planned and unplanned events. From this, a recommended decision-making approach is derived; no universally binding calendar is established.

Describe the criticality of decisions supported by the system, the frequency of system changes, and the reliability of ongoing controls in detecting issues. Additionally, consider unresolved incidents, open actions, supplier changes, and the quality of existing evidence. Justify a meeting date for the next comprehensive evaluation and name the authorized individual responsible for this determination.

A significant incident does not await until this deadline. Define additional triggers for an accelerated review, such as unexplained data loss or a fundamentally altered usage pattern. After a period of stability, the existing interval may remain appropriate; a extension requires a new justification. A overdue review is recorded with impact and catch-up plan, not retrospectively adjusted.

Sources: [1] [2]

Which evidence you should compile

The following work matrix translates the evaluation question into verifiable inputs. Reference approved or otherwise controlled records instead of appending entire document sets. For each source, the time period, system reference, and relevant limitations must be clearly identifiable.

Assess the quality of the data foundation: Are all support channels captured, or does the list only include tickets opened internally? A missing entry may indicate actual event-free periods, but it could also reflect incomplete data capture. Retain this uncertainty in the conclusion.

Recommended entry matrix for a system review
Area to assessAppropriate entryComment on the evaluation
Usage and Status System description, functionality, configuration Compare current usage with the previously assessed scope.
Changes Change register, release evaluations, tests Also consider the combined effect of individual changes.
Incidents Disruptions, deviations, root causes, and corrective actions Recurrent patterns and unresolved implications are being investigated
Roles and Access Roll model, user validations, administration Evaluate evidence and technical appropriateness separately
Data and Retry Security reports, restore proofs, archive access Scope, extent, and currency must align with the current system.
Supplier and Operation Support status, agreements, relevant reports Lack of transparency should be assessed as a gap, not as a warning sign
Open obligations Previous review actions, CAPA, residual deviations Completion and proven efficacy are distinct.

Sources: [1] [2]

Evaluate individual changes and their combined effect

Assign each relevant change to an requirement, a data pathway, or an operating procedure. This includes technical configuration, permissions, infrastructure, interfaces, and altered work processes. Also evaluate changes classified as purely technical when new information challenges this classification.

Example: A system interface receives a different field format, a downstream report adopts a new filter rule, and a departmental function introduces an additional release step. Each change can be individually reviewed. The review further assesses whether the entire data pathway remains clearly traceable and justifiable in light of the current process. Whether an additional test is required depends on the evaluation of the data flow.

Compare the approved, tested, and actually implemented version. If the connection is missing, clearly identify which decision remains uncertain. A subsequently updated document does not automatically eliminate a potential gap in operations. Determine whether a document correction, a technical assessment, or a targeted re-evaluation is required.

Sources: [1] [3]

Review incidents, responsibilities, and supplier information together

Group incidents by impact on the process rather than exclusively by technical category. Multiple minor interface failures may reveal the same gap in reconciliation. Distinguish between temporary repairs, root cause remediation, and evidence that the corrective action is effective in day-to-day operations. EU GMP Chapter 1 links CAPA to an assessment of its effectiveness.

During role changes, verify not only deactivated users but also new responsibilities, deputies, privileged accounts, and administrative service providers. The question is whether actual access and professional accountability still align. Leverage existing controlled access reviews as entry points; justify additional sampling in cases of discrepancies.

For the supplier, consider agreed-upon capabilities and actually available information: supported version, announced changes, relevant outages, and point of contact. A general certificate does not address any open questions regarding your configuration. Where information is missing, document the specific impact on your evaluation and agree on a feasible response or an alternative appropriate mitigation.

Sources: [3] [1]

Assess professionally the access to backup, restore, and archiving functions.

Annex 11 distinguishes between regular backup, verified recoverability, and accessible archived data. A green backup protocol alone does not address whether the technical process can be resumed following a failure.

Verify the most recent appropriate restoration record: Which dataset was restored, in which environment, and with what outcome? Is the record still aligned with the current architecture? Are the required configurations, permissions, keys, or associated services included in the considered restoration? Missing components must be evaluated; they cannot be attributed to the age of a successful individual test.

For archive reviews, assess the availability and understanding of complete records including necessary contextual relationships. Separate operational backup from long-term retention. Plan controlled testing with appropriate test data while protecting the production environment. During review, reference existing robust testing; however, the suitability of such testing remains to be justified.

Sources: [1]

Example: Decisions are made based on observations

A laboratory uses a LIMS with device integration and electronic approvals. Since the last review, roles and the transfer of a result type have been modified. A failure has recurred; the most recent restore trace originates from a previous architecture. The matrix provides decision recommendations for these assumed findings.

No general criticality or deadline is specified. The responsible personnel must assess the impact on results, records, and previously made quality decisions. A review report may explicitly acknowledge uncertainty, but must clarify any further decisions and, if applicable, the continued permissibility of operation.

Commented Practical Matrix: Findings, Conclusion and Final Documentation
Observation in this exampleRecommended decisionWhere completion is evident
New result transfer without a clear test reference Assess affected data pathways and already transferred data Rationale assessment of impact and appropriate supplementary testing
Recurrent disruption despite closed ticket Investigate the root cause and previous corrective actions again Validated effectiveness criterion based on appropriate observation
No longer required privileged account Review access and potential use immediately Controlled purification and documented follow-up evaluation
Restore only addresses previous architecture Clarify the suitability of the existing validation; plan required tests Evidence of restart covers the scope relevant today
Supplier information for the version is missing Request specific information from the point of contact Response evaluated or alternative safeguarding decided
No relevant gap identified after robust assessment Document the continued operation Approved conclusion and justified next scheduled date

Sources: [1] [3] [2]

Maintain a robust conclusion and actions taken

State the outcome for the defined scope of use. Identify the supporting evidence, open uncertainties, and required actions. A general statement such as "the system is validated" is of limited value if underlying limitations in equipment remain unaddressed. Clearly specify whether and under what conditions the assessed process may continue to be operated.

Give each action an owner, a justified due date and a verifiable completion criterion. Identify who accepts the result. A significant gap may call for immediate safeguards or restricted use. Follow the quality system to make this decision. Approving the review document alone cannot settle it.

A formally completed review may still include follow-up actions, provided the significance and handling of such actions are clearly defined. Incorporate these into the relevant action plan and assess their effectiveness at the next appropriate opportunity. Simply repeating the same open list does not indicate progress.

Sources: [1] [3]

Checklist for Preparation, Execution, and Closure

This checklist is intended for use during joint meetings involving the department, system owner, IT, and QA. It aims to make decisions visible; the number of completed items is not a quality metric. For non-relevant inspection areas, a clear and justified rationale within the agreed procedure is sufficient.

For ongoing planning, it is recommended to maintain a central overview of systems including the responsible person, next scheduled review, outstanding actions, and upcoming evaluation triggers. This ensures that systems remain visible, even when a review has not been completed due to missing documentation.

  • System boundary, evaluation period, and actual usage are clearly defined.
  • The interval and trigger for an accelerated assessment are justified.
  • Changes, incidents, and prior actions are fully sufficient for decision-making.
  • Roles, supplier information, safeguarding, and restart are evaluated with appropriate documentation.
  • Missing documentation and inconsistent results remain explicitly visible.
  • The continued operation is technically assessed; necessary restrictions have been implemented.
  • Actions have responsible parties, deadlines, and verifiable completion criteria.
  • The conclusion and the next meeting date have been approved within the operational procedure.

Sources: [1] [2]

Project experience and collaboration

Assess the current state of the system with your team

For a periodic review, we establish what has actually changed since the previous assessment with the system owners, the process team and IT. We check changes, incidents and outstanding actions against the operational records. A focused review may be sufficient when the evidence is sound. If the overview is missing, we build it with the team first. The report should support a reasoned decision about the system’s condition, actions required and continued operation.

Frequently asked questions about Periodic Review

Must a Periodic Review take place annually?

Annex 11 Section 11 does not specify a general annual cycle. Determine and justify the schedule for your system based on risk, changes, deviations, and the reliability of ongoing controls. An internally established annual cycle remains a business decision, not a general annual obligation under Annex 11.

Sources: [1] [2]

Which systems should be included in the review plan?

Begin in the considered GMP area with the defined system inventory and supported GMP functions. Capture external applications and relevant shared services. Assign a lead evidence document for overlaps to ensure neither responsibilities are lost nor the same service is inconsistently assessed.

Sources: [1]

Who should perform and approve the review?

It is recommended that a joint evaluation be conducted by process responsibility, system responsibility, IT, and QA personnel, based on their respective areas of responsibility. A supplier may contribute technical information. The preparation, assessment, and approval are determined by the quality system. The technical operational decision requires individuals with the appropriate authority, not merely access to a reporting template.

Sources: [1] [3]

Does the review replace a revalidation?

The review evaluates which evidence supports current usage and where additional work is required. It does not replace any actual required inspection. Conversely, the mere presence of a review date does not imply that all original tests must be repeated. Derive supplementary evidence from specific changes, risks, and knowledge gaps.

Sources: [1] [2]

Is an audit trail review the same as a periodic review?

No. An audit trail review examines relevant records of events and data modifications. A system assessment further evaluates usage, changes, operational experience, and proof of status. Use the results of appropriate ongoing reviews as input and assess their coverage, without equating different review assignments.

Sources: [1]

What should be done with a system having no reported incidents?

First assess whether the data capture was feasible. No alert can demonstrate a stable operation, nor can the absence of alerting pathways. Review appropriate sources to determine whether support cases, professional deviations, and relevant technical events were considered. Only thereafter can the absence of incidents be meaningfully interpreted.

Sources: [1] [2]

Must a restore be performed during every review?

Not automatically. It is essential to determine whether existing evidence adequately demonstrates the recoverability of the current system. Annex 11 addresses the evaluation and periodic monitoring of such evidence. Assess relevance and system alignment. An additional controlled test may be required if architectural changes or gaps limit the validity of existing assessments.

Sources: [1]

Does a supplier report suffice for a cloud-based application?

The report may provide relevant operational information. However, verify whether it actually covers your specific configuration, interfaces, roles, and technical workflows. Link it to your supporting documentation. A general statement about the platform does not address specific questions regarding the intended use.

Sources: [1]

Can a review with open actions be considered complete?

The review report may be closed with regularly scheduled open actions, provided that the significance and impact of these actions have been assessed and further processing has been assigned to responsible parties with defined timelines. The closure of the review report should not be conflated with a decision regarding continued operation. An open critical uncertainty is not considered acceptable if addressed solely by a scheduled item in the action plan. Necessary protective measures and relevant professional authorizations must be explicitly documented.

Sources: [1] [3]

What do when a review is overdue?

Capture the deviation from the original plan and assess the resulting uncertainty. Prioritize based on current system risk, obtain missing inputs, and agree on a catch-up plan. Additionally, evaluate whether any acute events require immediate actions. Backdating does not provide evidence of a prior evaluation.

Sources: [1] [2]

Sources and review date

Source date: 6 October 2026. This guide considers EU-GMP for human medicinal products. The work matrix, audit questions, and LIMS example are standalone planning recommendations based on the referenced sources; they do not establish general deadlines or release decisions.

  1. European Commission: EU GMP Annex 11, Computerised Systems
  2. ICH Q9(R1): Quality Risk Management
  3. European Commission: EU GMP Chapter 1, Pharmaceutical Quality System
  4. FDA: Part 11, Electronic Records; Electronic Signatures – Scope and Application
  5. European Commission: EudraLex Volume 4 – Current Documents

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

Prepare the next system review in detail.

When evidence, responsibilities, or open actions do not align, Daniel Herrmann Consulting supports you in defining and evaluating the boundaries and gaps. Together, we determine which information is missing for your next decision.

  • 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 →