EU Pharmacovigilance Inspection Findings: Benefit–Risk Assessment and Decision-Making

Purpose and Scope

Pharmacovigilance ultimately exists to support decisions about the safe and appropriate use of medicines. Safety information is collected and analysed, but the regulatory value of that work depends on how the resulting evidence is interpreted and translated into action. An inspection may therefore examine not only whether pharmacovigilance activities were performed, but whether the organisation can demonstrate a controlled and scientifically justified process for reaching material safety decisions.

This article examines that inspection perspective. It focuses on the process and evidence supporting pharmacovigilance decision-making, rather than attempting to judge the scientific merits of a particular medicinal product or reproduce the detailed methodology of benefit–risk assessment.

The central distinction is between a decision that is scientifically debatable and a process that is inadequately controlled. Pharmacovigilance frequently involves uncertainty, incomplete information and reasonable differences of scientific judgement. An inspection does not necessarily establish a deficiency merely because another expert might have reached a different conclusion. The more relevant questions are whether the applicable process was followed, whether the material evidence was considered, whether important uncertainty was recognised, whether the decision was appropriately governed and whether the reasoning can be reconstructed.

The inspection model developed in the preceding articles can therefore be extended as follows:

requirement → evidence → evaluation → uncertainty → decision → action → governance → subsequent review

Why Decision-Making Is an Inspection Subject

A pharmacovigilance system can produce large quantities of data without necessarily producing reliable decisions. Individual case reports, aggregate analyses, signal assessments, epidemiological evidence, literature findings and other sources must be interpreted in context.

The resulting decision may concern whether a signal requires further evaluation, whether a risk-management measure should be introduced or modified, whether product information should change, whether additional evidence is required, or whether the available evidence supports continued use under the existing conditions.

Because these decisions can have important consequences, an inspector may need to establish how the organisation moved from information to conclusion. A final statement such as "no further action required" is much less informative if the organisation cannot demonstrate what evidence was considered and why that conclusion was reached.

This does not mean that every decision requires an extensive written scientific essay. Documentation should be proportionate to the decision, the uncertainty and the applicable process. The inspection issue arises when the retained evidence is insufficient to reconstruct a material decision.

Regulatory Framework

The EU pharmacovigilance framework establishes requirements for the collection, evaluation and management of safety information and for the quality system supporting those activities. GVP Modules I, VI, VII, IX, XII and XVI address different components of this system, while the applicable EU legislation provides the legal framework for pharmacovigilance obligations.

Benefit–risk assessment is also embedded in the broader regulatory lifecycle. Safety findings can lead to changes in product information, additional monitoring, risk-minimisation measures, post-authorisation studies, regulatory referrals or other actions depending on the evidence and applicable procedure.

The regulatory framework should therefore be read as an interconnected system. A signal is not synonymous with a regulatory decision, and a risk-management measure is not synonymous with proof that the underlying risk has been eliminated. The decision process connects these stages.

The specific regulatory procedure applicable to a decision must always be considered. This article does not establish a universal decision-making procedure for every medicinal product.

From Safety Evidence to Regulatory Decision

A useful conceptual model is:

Safety information
       ↓
Validation and characterisation
       ↓
Scientific evaluation
       ↓
Assessment of uncertainty
       ↓
Benefit–risk interpretation
       ↓
Decision
       ↓
Risk-management / regulatory action
       ↓
Follow-up and reassessment

Different pharmacovigilance processes may enter this pathway at different points. A new signal may begin with an emerging safety concern. An aggregate report may provide a broader assessment. A post-authorisation study may provide new epidemiological evidence. A regulatory authority may request an assessment in response to information from another source.

The common feature is that information must be converted into an evidence-based decision through a controlled process.

Scientific Judgement and Inspection Judgement

Scientific judgement is inherent to pharmacovigilance. The available evidence may support several plausible interpretations, particularly when the event is rare, the data are confounded or the available studies have limitations.

Inspection judgement is different. The inspector is primarily evaluating whether the organisation's pharmacovigilance system is compliant and effective and whether the evidence demonstrates that the organisation operated its processes appropriately.

This distinction is fundamental. An inspector may challenge an assessment because the underlying evidence appears incomplete or because the documented reasoning does not explain an important issue. That is not equivalent to an inspector substituting a different scientific opinion for the company's expert judgement.

The stronger the evidence trail, the easier it is to distinguish legitimate scientific uncertainty from inadequate process control.

Defining the Decision Question

Good decision-making begins with a clear question. A vague question such as "Is there a safety issue?" may not be sufficient for a complex assessment.

A decision question might instead concern whether the evidence supports a causal association, whether an identified risk has changed in frequency or severity, whether a current risk-minimisation measure remains appropriate, or whether additional evidence is needed before a regulatory action can be considered.

The exact question determines which evidence is relevant. It also affects how uncertainty should be described and which outcomes should be considered.

From an inspection perspective, a clearly defined decision question helps demonstrate why particular analyses were performed and why other information was not central to the decision.

Evidence Identification and Completeness

A material safety decision should be based on evidence appropriate to the question. The relevant evidence may come from multiple sources, and the organisation should have processes for identifying information that could materially affect the assessment.

Completeness does not mean that every conceivable source must be included in every assessment. It means that the process should be capable of identifying and considering relevant information according to the applicable methodology.

An inspector may therefore examine whether the organisation's conclusion is consistent with information available to it at the relevant time. The ability to explain why particular evidence was included, excluded or considered non-material can be important when the evidence base is complex.

Individual Cases and Aggregate Evidence

Individual case reports are often an important component of safety assessment, but an individual case rarely provides the entire basis for a broader benefit–risk conclusion.

Aggregate evidence can reveal patterns that individual cases cannot establish. Conversely, aggregate analyses can obscure clinically important details that are visible in individual reports. A robust assessment therefore considers the appropriate level of evidence for the decision question.

Inspection testing may involve tracing a conclusion from an aggregate assessment back to selected underlying cases or other data sources. The purpose is not necessarily to reproduce the entire analysis but to establish that the relationship between the source evidence and conclusion is credible and controlled.

Signal Assessment Is Not the Same as Benefit–Risk Decision

A signal assessment asks whether a new or changed safety concern warrants further investigation or action within the signal-management process. A benefit–risk decision is broader and considers the safety issue in the context of the medicine's benefits, alternatives, disease characteristics and other relevant evidence.

The two processes can therefore produce different outputs without being inconsistent. A potential signal may be investigated without immediately changing the overall benefit–risk conclusion. Conversely, an accumulating body of evidence may alter the benefit–risk balance even when no single signal determines the outcome.

This distinction is useful during inspection because the existence of a signal does not automatically establish that a regulatory action was required. The relevant question is whether the organisation applied the appropriate process to the information available.

Characterising Uncertainty

Uncertainty is not evidence of a deficient assessment. It is often an intrinsic characteristic of pharmacovigilance evidence.

A strong assessment identifies important uncertainties and explains their potential influence on the conclusion. These may arise from limited exposure, missing information, confounding, reporting bias, inconsistent findings, limitations of study design or other factors.

The inspection question is whether material uncertainty was recognised and appropriately incorporated into the decision. An assessment that presents a highly uncertain conclusion as though it were established fact may indicate a weakness in scientific reasoning or governance.

Evidence Weighting

Not all evidence carries the same interpretive significance. A spontaneous-reporting pattern, controlled study, observational study, literature review and mechanistic evidence can contribute differently to a safety assessment.

The organisation should use an appropriate scientific framework for considering the evidence. It should also be able to explain material apparent conflicts between evidence sources where those conflicts affect the decision.

Inspection does not require that every assessment use a single prescribed weighting formula. The relevant requirement is that the assessment is scientifically appropriate, documented to an extent proportionate to its significance and consistent with the applicable pharmacovigilance process.

Conflicting Evidence

Conflicting evidence is common in safety science. One study may suggest an association while another does not. Reporting patterns may change without a corresponding change in underlying incidence. A mechanistic concern may exist without sufficient epidemiological evidence to quantify risk.

The existence of conflicting evidence is therefore not itself a deficiency. The inspection concern is whether the organisation recognised the conflict and considered its implications.

A decision record should make it possible to understand why the available evidence was considered sufficient, insufficient or inconclusive for the decision being made.

Benefit–Risk Context

Safety cannot normally be interpreted in isolation from therapeutic benefit. The same safety finding can have different regulatory significance depending on disease severity, treatment alternatives, magnitude and certainty of benefit, patient population and availability of risk-minimisation options.

The exact benefit–risk methodology varies by regulatory context and decision. The inspection focus is whether the organisation applied the appropriate framework and considered material factors relevant to the decision.

A safety assessment that accurately describes a risk but does not connect that risk to the decision it is intended to inform may therefore be incomplete from a governance perspective.

Decision Options

A pharmacovigilance assessment does not always lead to a binary decision of "action" or "no action". Depending on the situation, options can include continued monitoring, additional analysis, further data generation, changes to risk-minimisation measures, product-information changes or escalation through an appropriate regulatory pathway.

The important inspection issue is whether the options were considered appropriately and whether the rationale for the selected course can be reconstructed.

The absence of a documented alternative does not automatically demonstrate a deficiency. Documentation should be proportionate to the decision and the applicable process.

Materiality and Proportionality

Not every pharmacovigilance decision requires the same degree of governance. A routine operational assessment and a decision with potential consequences for the overall benefit–risk profile are materially different.

A risk-based system should therefore identify decisions that warrant enhanced review, escalation or governance. Materiality may depend on factors such as potential patient impact, novelty of the concern, uncertainty, scope of the affected population and possible regulatory consequences.

This does not mean that an organisation must invent a universal scoring system. The control should be appropriate to its pharmacovigilance system and the applicable requirements.

Governance of Safety Decisions

Scientific assessment and governance are complementary. The people performing an assessment need appropriate expertise, while the organisation needs a defined mechanism for review, challenge and escalation of material safety issues.

Governance should be proportionate to the decision. A routine assessment may follow an established workflow, whereas a potentially consequential change to the benefit–risk profile may require review by additional functions or governance bodies.

From an inspection perspective, the important evidence is not simply that a meeting occurred. It is whether the relevant decision-makers received the material information, whether appropriate expertise was available, whether significant disagreement or uncertainty was considered and whether the resulting decision was documented.

The Role of the QPPV

The QPPV has an important oversight role within the pharmacovigilance system. That role does not mean that the QPPV must personally make every scientific assessment or regulatory decision.

For material safety issues, however, the QPPV should have appropriate visibility of significant risks and of the processes used to assess and manage them. The evidence of oversight will depend on the organisation's governance structure.

An inspector may therefore examine whether important safety information reaches the QPPV, whether the QPPV can escalate concerns, and whether the QPPV has sufficient authority and access to relevant information to perform the assigned role.

Cross-Functional Challenge

Benefit–risk decisions can involve pharmacovigilance, regulatory, medical, clinical, epidemiological, statistical, quality and other expertise. Different functions may legitimately emphasise different aspects of the evidence.

A disagreement is not necessarily a quality failure. In some circumstances, constructive challenge is evidence of a functioning decision process. The inspection concern arises when disagreement about a material issue is hidden, unresolved without rationale or excluded from the documented decision process.

A mature governance process should allow relevant scientific disagreement to be surfaced and resolved or appropriately escalated.

Decision Records

A decision record should allow a knowledgeable reviewer to understand the question considered, the material evidence, the conclusion and the resulting action at a level proportionate to the decision.

The appropriate format varies. It may be a formal assessment, committee record, regulatory submission, signal decision document, aggregate-report conclusion or another controlled record.

The record should not be judged by length. A concise record can be adequate if it captures the material reasoning. A long record can remain inadequate if it contains extensive background but does not explain the actual decision.

Decisions Under Uncertainty

Regulatory and pharmacovigilance decisions frequently have to be made before all uncertainty can be resolved. Waiting for perfect evidence may not be an available option when a potential serious risk requires timely management.

The quality of the decision therefore depends partly on how uncertainty is handled. An organisation may decide to act despite uncertainty, continue monitoring while generating additional evidence, or maintain the existing position where the evidence does not justify a change.

Inspection should focus on whether the decision was reasonable within the applicable process and evidence available at the time, rather than judging it solely with hindsight.

The Importance of Timing

A scientifically sound conclusion reached too late may not provide effective pharmacovigilance. Conversely, a rapid decision based on inadequate information may also be problematic.

The relevant process should therefore connect the urgency of the issue with the time required for appropriate assessment. A serious emerging concern may require accelerated escalation, while a complex low-urgency assessment may allow more extensive evidence generation.

Inspection evidence can include dates of awareness, assessment, escalation, decision and action. These dates help establish whether the organisation responded appropriately to the circumstances.

Regulatory and Internal Decision-Making

Not every internal pharmacovigilance conclusion is itself a regulatory decision. An MAH may perform an internal assessment that leads to continued monitoring, while a competent authority or scientific committee subsequently reaches a regulatory conclusion under a formal procedure.

The distinction should be preserved in documentation. An internal assessment should not be represented as though it were a competent-authority decision, and an external regulatory conclusion should not be attributed to the MAH's internal process.

This is particularly important when inspection evidence includes correspondence with regulatory authorities or records of formal procedures.

When New Evidence Arrives

Benefit–risk assessment is not necessarily a one-time exercise. New cases, studies, literature, post-authorisation data, regulatory information or other evidence can change the assessment.

A controlled pharmacovigilance system should have mechanisms for identifying information that may warrant reassessment. The threshold for reassessment depends on the process and significance of the new information.

Inspection may therefore examine whether a material new finding was appropriately connected to the existing safety assessment and whether the organisation could demonstrate why reassessment was or was not required.

Historical Reconstruction

Inspection frequently involves events that occurred before the current personnel or current version of a procedure were in place. The organisation should therefore retain sufficient evidence to reconstruct material historical decisions.

A current conclusion about what the organisation would decide today is not necessarily evidence of what was reasonably known or decided at the historical time. Inspectors may need to see the information available then, the applicable procedure, the assessment and the governance record.

This links directly to the evidence-chain principles developed in I5.

Potential Inspection Finding Patterns

The following categories are illustrative analytical patterns, not claims that each has been published as an inspection finding by an EU authority.

Material evidence omitted

Relevant safety information was available to the organisation but was not appropriately considered in a material assessment.

Conclusion unsupported by retained evidence

The organisation can state the conclusion but cannot reconstruct the analysis supporting it.

Uncertainty not addressed

Important limitations of the evidence are not reflected in the assessment or decision.

Decision process not followed

The applicable governance or escalation process was not applied to a material issue.

Cross-functional disagreement not resolved or documented

Material scientific disagreement existed but the basis for the final position cannot be reconstructed.

Delayed reassessment

New information that could materially affect the safety assessment did not trigger timely consideration.

Decision/action disconnect

The assessment identifies a material concern but the resulting action is not clearly linked to the conclusion.

Hindsight reconstruction

A current analysis is used to explain a historical decision without sufficient evidence of what was known and considered at the time.

Distinguishing a Scientific Difference From a Process Deficiency

This is one of the most important inspection distinctions in safety decision-making.

Suppose two qualified experts review the same evidence and reach different conclusions about the strength of a causal association. If the organisation's process permits appropriate scientific challenge, documents the relevant evidence and records the rationale for the adopted position, the disagreement does not necessarily demonstrate a deficiency.

By contrast, if one conclusion is adopted without considering material contradictory evidence, or if the organisation cannot establish why the final position was reached, the problem may be a process or governance deficiency even though the underlying scientific question remains uncertain.

The inspection focus is therefore on the quality and control of the decision process, not on requiring unanimity of scientific opinion.

Inspection Testing of a Decision

An inspector can test a material safety decision by moving backwards through the evidence chain:

Final decision
      ↓
Decision rationale
      ↓
Scientific assessment
      ↓
Evidence considered
      ↓
Source data
      ↓
Information available at the time

The inspector may then move forward again:

Decision
   ↓
Governance / approval
   ↓
Action
   ↓
Implementation
   ↓
Follow-up monitoring

This bidirectional approach can reveal whether the decision is supported both retrospectively by evidence and prospectively by appropriate action.

Sampling Decisions for Self-Inspection

An internal inspection should not sample only decisions that resulted in regulatory action. Decisions not to act can be equally informative because they test whether the organisation can demonstrate why continued monitoring or no change was appropriate.

A useful sample can include:

The sample should be adapted to the organisation's products and risk profile.

Documentation Versus Scientific Quality

Good documentation cannot rescue scientifically inadequate analysis. Conversely, scientifically strong work can become difficult to defend if the evidence and reasoning are not appropriately documented.

The two dimensions should therefore be assessed separately:

Scientific quality: Was the evidence appropriately evaluated?

Process quality: Was the assessment conducted, reviewed and governed appropriately?

Evidence quality: Can the organisation demonstrate what was done and why?

A mature system addresses all three.

CAPA When Decision-Making Controls Fail

If an inspection identifies a decision-making deficiency, CAPA should address the mechanism that allowed the failure.

For example, if material evidence was repeatedly omitted because no defined cross-functional evidence review existed, simply retraining assessors may not address the root cause. The process may need a clearer evidence-identification mechanism, defined review responsibilities or a governance control for material assessments.

The corrective response should be proportionate to the actual failure and should include scope assessment where the mechanism could have affected other decisions.

Effectiveness of Decision-Making CAPA

Effectiveness should test whether the relevant decision-making control now functions.

If the CAPA introduced enhanced review of material safety assessments, effectiveness should examine whether subsequent assessments demonstrate that the intended review actually occurred and identified relevant issues. If the CAPA addressed escalation, effectiveness should test whether subsequent qualifying events were escalated appropriately.

Counting completed training sessions or revised procedures is insufficient if the underlying decision-making weakness remains untested.

QPPV Inspection Questions

The following are illustrative questions, not an official regulatory checklist:

• How does the QPPV become aware of material changes in the benefit–risk profile?
• Which safety decisions require enhanced governance?
• How are material scientific disagreements handled?
• How do you demonstrate that relevant evidence was considered?
• How are important uncertainties recorded?
• What triggers reassessment when new evidence becomes available?
• Can you reconstruct why this decision was made at the time?
• How do you distinguish a scientific judgement from a procedural requirement?
• How are decisions to take no action governed?
• How do you verify that decisions led to the intended action?

Inspection Evidence Package

When a material decision is examined during inspection, the strongest evidence package is one that allows the decision to be reconstructed without relying on retrospective explanation.

Depending on the decision, the evidence may include the triggering safety information, relevant analyses, assessment documents, meeting or governance records, approvals, correspondence, regulatory submissions, resulting actions and subsequent follow-up. Not every decision requires every type of record.

The evidence should nevertheless be coherent. A final conclusion should correspond to the assessment, the assessment should correspond to the evidence available at the time, and the resulting action should be consistent with the decision.

Decision Quality and Evidence Quality

An inspection should distinguish between the quality of a decision and the quality of evidence available to demonstrate that decision.

A decision may have been reasonable but poorly documented. Conversely, an extensively documented assessment may still contain weak scientific reasoning. These are different deficiencies and should lead to different remediation.

The organisation should avoid responding to an evidence problem by simply producing retrospective documentation that was not created at the relevant time. Retrospective reconstruction can be useful, but it should be identified as such and should not be presented as contemporaneous evidence.

Decision-Making After Regulatory Interaction

Regulatory authorities may request information, challenge an assessment or initiate a formal procedure. The MAH's subsequent internal assessment should clearly distinguish the authority's position from the company's own evaluation.

Where an authority requests additional analysis, the organisation should be able to demonstrate how the request was assessed, assigned and addressed. Where the authority's conclusion differs from the company's earlier position, the difference should not automatically be treated as evidence that the earlier assessment was deficient. The relevant circumstances and information available at each stage need to be considered.

Decisions Involving Risk-Minimisation

When a safety assessment results in a risk-minimisation measure, the decision process should connect the identified risk with the chosen intervention.

The choice of measure should be proportionate to the risk and appropriate to the objective being pursued. The organisation should be able to explain why the selected measure was considered appropriate and how its implementation and, where applicable, effectiveness will be evaluated.

An inspection can therefore move through two linked questions: why was this measure selected? and how does the organisation know whether it achieved its intended purpose?

This article does not replace the dedicated QPPV.com articles on risk-minimisation measures or their effectiveness.

Decisions Not to Change Product Information

A decision not to modify product information can be just as important to reconstruct as a decision to make a change. Such decisions may follow an assessment that the available evidence does not justify a change, that further evaluation is required or that existing information adequately addresses the concern.

The inspection question is not whether every safety concern resulted in a label change. It is whether the organisation applied the appropriate assessment and governance process and can demonstrate the basis for the outcome.

Decision-Making in Emerging Safety Issues

Emerging safety issues can create time pressure. The organisation may have to decide whether to escalate, seek additional evidence, implement interim measures or continue monitoring while the evidence develops.

The quality of the process is demonstrated by appropriate prioritisation, escalation, documentation and reassessment rather than by the absence of uncertainty.

An inspector may pay particular attention to the sequence of events because a later regulatory outcome can obscure the uncertainty that existed when the initial decision was made.

Common Weaknesses in Internal Review

Internal review can fail when it is designed around document completeness rather than decision quality. A checklist may confirm that an assessment contains the expected headings while failing to determine whether contradictory evidence was considered or whether the decision follows logically from the analysis.

A more effective review asks whether the decision question is clear, the relevant evidence has been considered, uncertainty is visible, the rationale is understandable and the action is linked to the conclusion.

These are recommended operational principles, not a prescribed EU checklist.

A Practical Decision-Quality Review

A proportionate internal review can use the following sequence:

The review should be proportionate to the materiality of the decision.

When a Decision Is Revisited

Reassessment should preserve the distinction between the original decision and the later decision. New evidence may legitimately change the conclusion.

A later change in position does not necessarily mean that the earlier decision was wrong. The inspection question is whether the earlier decision was reasonable and appropriately governed based on the evidence available at that time, and whether the new information was appropriately incorporated when it emerged.

This principle is particularly important in long-running safety issues where evidence accumulates gradually.

Root Cause When the Decision Process Fails

If an inspection identifies a decision-making deficiency, root-cause analysis should determine why the process did not detect or manage the problem.

Potential causes can include incomplete evidence identification, unclear responsibilities, inadequate escalation criteria, ineffective scientific review, weak documentation controls, insufficient cross-functional challenge or poor linkage between assessment and action. These are potential categories, not predetermined findings.

The root cause should be supported by evidence from the organisation's actual process.

CAPA and Effectiveness Verification

Corrective action should address the mechanism that failed. If the problem was inadequate evidence identification, a new document template may be insufficient. If the problem was weak governance, retraining individual assessors may not correct the system.

Effectiveness verification should then test the intended improvement using subsequent relevant decisions or another suitable evidence source. The test should be capable of detecting recurrence rather than merely confirming that the CAPA activities were completed.

This connects directly to the principles established in the CAPA effectiveness article in this series.

What an Inspector May Conclude From Strong Evidence

When the evidence chain is coherent, an organisation can demonstrate that safety information was identified, evaluated, challenged appropriately, translated into a decision and followed by proportionate action. Scientific uncertainty can remain without making the process defective.

This is an important feature of a mature pharmacovigilance system: good governance does not eliminate uncertainty; it makes uncertainty visible and manageable.

What an Inspector May Explore Further

Potential areas for further examination include:

• a material conclusion that cannot be reconstructed;
• contradictory evidence that is not addressed;
• repeated decisions reaching conclusions without documented rationale;
• significant safety information reaching the organisation but not the appropriate governance process;
• a decision to take no action without sufficient evidence of consideration;
• material new evidence without evidence of reassessment;
• a regulatory action that cannot be linked to the underlying safety assessment;
• recurring decision-process weaknesses despite previous CAPA.

These are illustrative inspection considerations, not reported inspection findings unless separately attributed to a published regulatory source.

Relationship to Other Inspection Domains

Decision-making is the point at which many pharmacovigilance processes converge. Case processing provides information; signal management identifies and evaluates potential concerns; aggregate reporting integrates evidence; risk management translates identified risks into measures; the quality system governs the process; and the QPPV provides pharmacovigilance oversight.

This article therefore does not replace those domain-specific articles. Instead, it provides the decision layer that connects them.

The complete inspection model can now be expressed as:

system → interface → data → evidence → evaluation → decision → action → monitoring → remediation.

Key Takeaways

• Pharmacovigilance inspection is concerned not only with whether activities occurred but with whether material safety decisions can be reconstructed and justified within the applicable process.
• Scientific uncertainty and legitimate scientific disagreement do not automatically constitute inspection deficiencies.
• A process deficiency may exist when material evidence is omitted, uncertainty is hidden, governance is bypassed or the rationale for a decision cannot be reconstructed.
• Benefit–risk decision-making should be considered in the context of the evidence available at the time rather than judged solely through hindsight.
• Decisions to take no action also require an appropriate evidentiary and governance basis.
• The QPPV provides oversight rather than personally performing every scientific assessment.
• A coherent evidence chain links source information, evaluation, decision, action and subsequent reassessment.
• CAPA should address the failed decision-making control, and effectiveness verification should test whether that control now works.
• Good governance does not require elimination of uncertainty; it requires uncertainty to be recognised, evaluated and appropriately managed.

References

1. European Commission. Regulation (EC) No 726/2004, as amended, establishing Union procedures for the authorisation and supervision of medicinal products and establishing the European Medicines Agency. urlEUR-Lex — Regulation 726/2004https://eur-lex.europa.eu/eli/reg/2004/726/oj
2. European Commission. Directive 2001/83/EC, as amended, on the Community code relating to medicinal products for human use. urlEUR-Lex — Directive 2001/83/EChttps://eur-lex.europa.eu/eli/dir/2001/83/oj
3. European Commission. Commission Implementing Regulation (EU) No 520/2012, as amended, on the performance of pharmacovigilance activities. urlEUR-Lex — Regulation 520/2012https://eur-lex.europa.eu/eli/reg_impl/2012/520/oj
4. European Medicines Agency. Good pharmacovigilance practices (GVP) Module I — Pharmacovigilance systems and their quality systems, current revision. urlEMA GVP Module Ihttps://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/good-pharmacovigilance-practices-gvp/gvp-modules
5. European Medicines Agency. GVP Module IX — Signal management, current revision. urlEMA GVP Module IXhttps://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/good-pharmacovigilance-practices-gvp/gvp-modules
6. European Medicines Agency. GVP Module VII — Periodic safety update report, current revision. urlEMA GVP Module VIIhttps://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/good-pharmacovigilance-practices-gvp/gvp-modules
7. European Medicines Agency. GVP Module XVI — Risk minimisation measures: selection of tools and effectiveness indicators, current revision. urlEMA GVP Module XVIhttps://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/good-pharmacovigilance-practices-gvp/gvp-modules
8. European Medicines Agency. GVP Module III — Pharmacovigilance inspections, current revision. urlEMA GVP Module IIIhttps://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation/good-pharmacovigilance-practices-gvp/gvp-modules
9. European Medicines Agency. Annual report of the Pharmacovigilance Inspectors' Working Group for 2024. Published information on EU pharmacovigilance inspection activity. urlEMA PhV IWG Annual Report 2024https://www.ema.europa.eu/en/documents/report/annual-report-pharmacovigilance-inspectors-working-group-2024_en.pdf

Regulatory Note

This article is an educational analysis of benefit–risk assessment and pharmacovigilance decision-making in the context of EU inspections. It distinguishes legal requirements and GVP guidance from recommended operational practice and illustrative inspection scenarios. The analytical categories and inspection questions not explicitly attributed to an authoritative source are hypothetical and must not be interpreted as published inspection findings or official regulatory checklists. Current legislation, GVP guidance, Union procedures, national requirements and product-specific obligations should be checked before using this material for operational or regulatory decisions.