Signal Management During Pharmacovigilance Inspections

Signal management is a natural inspection focus because it tests whether a pharmacovigilance system can recognise new safety information, evaluate it scientifically and translate conclusions into appropriate action. Inspectors can therefore use a small number of sampled signals to examine many parts of the system at once: data sources, scientific judgement, documentation, governance, QPPV oversight, regulatory interfaces and quality controls.

Regulatory Framework

The EU framework is derived from Directive 2001/83/EC, Regulation (EC) No 726/2004, Commission Implementing Regulation (EU) No 520/2012 as amended, GVP Module IX — Signal management, and GVP Module III — Pharmacovigilance inspections.

As of September 2026, GVP Module IX Rev. 1 remains published, but the legal framework was amended by Commission Implementing Regulation (EU) 2025/1466. EMA has ended the previous MAH EudraVigilance signal-detection pilot and has stated that Module IX will be revised for alignment. An inspection should therefore be prepared against the current legal and procedural framework, not against obsolete EVDAS operating assumptions.

What an Inspector Is Trying to Establish

The central question is whether the MAH operates an effective and controlled signal-management process.

An inspector may test whether the organisation can demonstrate:

• appropriate sources of safety information are considered;
• potential signals are handled through defined processes;
• validation and assessment decisions are scientifically justified;
• important issues are prioritised and escalated appropriately;
• decisions are traceable to the evidence considered;
• QPPV oversight is effective;
• outsourced work remains under MAH control; and
• conclusions are translated into RMP, product-information, PSUR/PBRER, communication or other actions where relevant.

Inspection therefore examines both scientific quality and system control.

The Reconstruction Test

A useful way to understand inspection methodology is the reconstruction test.

An inspector selects a signal and asks the organisation to reconstruct its history:

source → detection → validation → prioritisation → assessment → decision → action → follow-up or closure.

At each step, the inspector can compare three things:

1. what the procedure says should happen;
2. what the records show actually happened; and
3. what staff explain during interviews.

Misalignment between these three sources is often more important than the absence of a particular template.

Selecting Samples

Inspection sampling is risk-based and authority-dependent. An inspector may choose signals because they were important, delayed, closed without action, associated with a regulatory request, linked to product-information changes, or illustrative of a particular process or vendor interface.

The MAH should therefore not prepare only its "best" signal examples. Genuine readiness means that ordinary historical records are understandable and defensible.

Detection and Validation Evidence

For the early stages of a sampled signal, an inspector may explore:

• how the observation entered the process;
• which data sources were relevant;
• how the clinical concept or search strategy was defined;
• how validation was performed;
• why an observation progressed or was closed; and
• whether the record reflects the terminology and process in force at that time.

For legacy periods, this last point matters. Historical signal activity performed under the former EudraVigilance MAH pilot should be judged in the context of the requirements then applicable; current procedures should reflect the post-2025 framework.

Assessment and Scientific Rationale

Signal assessment is where inspectors can test whether the organisation's scientific reasoning is reproducible. A robust record should identify the question being assessed, the evidence considered, important limitations, alternative explanations and the reasoning supporting the conclusion.

An inspector may challenge:

• selective use of favourable evidence;
• unexplained exclusion of cases or studies;
• conclusions stronger than the available evidence;
• reliance on disproportionality as proof of causality;
• lack of consideration of exposure, background incidence or confounding where relevant;
• inadequate discussion of contradictory evidence; or
• closure rationales that cannot be reconstructed.

The objective is not that every inspector must agree with the original scientific judgement. The organisation should be able to show that the judgement was competent, balanced and based on the information reasonably available at the time.

Governance and Escalation

Signal governance is often examined because important safety decisions may cross organisational boundaries. Inspectors may ask:

• Who had authority to validate, assess, close or escalate the signal?
• Which decisions required multidisciplinary review?
• How were urgent issues handled?
• What happened when reviewers disagreed?
• How was the QPPV informed?
• Were actions assigned and completed?

A standing committee is not inherently required. The organisation must instead show that its chosen governance mechanism reliably delivers the required expertise, accountability and escalation.

QPPV Oversight

The QPPV should have sufficient authority and access to information to oversee the functioning of the pharmacovigilance system. During a signal-management inspection discussion, the QPPV may therefore be asked to explain:

• how significant signal matters become visible;
• how emerging or high-impact concerns are escalated;
• how signal-management performance is monitored;
• how outsourced activities are controlled;
• how important conclusions feed into the RMP and benefit-risk evaluation; and
• how the QPPV knows that agreed corrective or regulatory actions were implemented.

This does not mean the QPPV must personally perform routine detection, approve every assessment or sign every closure.

Outsourced Signal Activities

Where vendors perform literature review, analytics, case screening or assessment support, inspectors may trace the interface between vendor and MAH.

Relevant evidence can include:

• allocation of responsibilities;
• data-flow and escalation arrangements;
• quality and performance monitoring;
• deviations and corrective actions;
• evidence that MAH reviewers challenge vendor outputs where appropriate; and
• ability to access the records required to reconstruct decisions.

A contract that assigns an activity does not transfer the MAH's pharmacovigilance responsibility.

Signal Outcomes and Downstream Implementation

A signal is not fully controlled merely because an assessment document is closed. Inspectors may follow the conclusion into downstream systems.

For example, a confirmed safety concern may require:

• product-information assessment and regulatory submission;
• RMP update or change to safety concerns;
• additional pharmacovigilance;
• risk-minimisation change;
• PSUR/PBRER discussion;
• safety communication; or
• continued monitoring under defined conditions.

The organisation should be able to demonstrate that these consequences were identified, assigned, implemented and reconciled.

Timeliness

Some signal-related actions have regulatory timelines; other time targets are organisation-specific controls. These should not be confused.

An inspector may assess whether:

• legally or procedurally required timelines were met;
• internal targets are appropriate to the risk;
• overdue activities are visible;
• significant delays are escalated; and
• deviations are assessed for impact rather than simply closed administratively.

A dashboard showing green status does not compensate for an inappropriate target or hidden backlog.

Documentation and Data Integrity

Signal records should be sufficiently controlled to support reconstruction. Depending on the system, evidence may include:

• signal register or workflow history;
• assessment documents and linked evidence;
• search strategies and data cut-off dates;
• version histories;
• meeting or decision records;
• regulatory correspondence;
• action trackers; and
• relevant system audit trails.

There is no requirement to create a separate "inspection pack" for every signal. The underlying records should already be reliable and retrievable as part of normal operation.

Illustrative Failure Modes

The following are potential failure modes, not reported inspection findings.

Practical Inspection Questions

An experienced internal reviewer can use questions such as:

1. Can we reconstruct several signals chosen at random rather than only prepared examples?
2. Does the procedure reflect the post-2025 EudraVigilance framework?
3. Are detection, validation, prioritisation and assessment used consistently as distinct concepts?
4. Can reviewers explain why signals were closed as well as why they were escalated?
5. Are evidence limitations and contradictory data visible in the assessment?
6. Do governance records show rationale rather than only attendance and conclusions?
7. Does the QPPV receive proportionate visibility of important matters?
8. Can vendor-derived conclusions be traced to underlying evidence?
9. Are regulatory and internal timelines distinguished correctly?
10. Can downstream product-information, RMP, PSUR/PBRER and risk-minimisation actions be traced back to the signal decision?
11. Are overdue actions and deviations visible and impact-assessed?
12. Does the system preserve historical records in the context of the requirements applicable at the time?

Relationship With Audit Readiness

Internal audit and inspection readiness overlap but are not identical. An audit programme may test the adequacy and effectiveness of the signal-management controls before a regulator examines them. Inspection readiness is the continuing state in which the underlying process and evidence are sufficiently reliable that external scrutiny does not require last-minute reconstruction.

The companion article Signal Management Audit Readiness focuses on how an organisation can test that state of control prospectively.

Key Takeaways

Inspectors use signal management to test both scientific judgement and pharmacovigilance-system control.

The most powerful inspection method is end-to-end reconstruction of real signals, comparing the procedure, the records and staff explanations.

A committee, dashboard, inspection pack or QPPV signature is not evidence of effectiveness by itself. What matters is whether important information is identified, evaluated, escalated and translated into appropriate action through a controlled and traceable process.

The 2025–2026 changes to EudraVigilance signal-detection obligations are particularly important. Current procedures should no longer present the former MAH EudraVigilance pilot model as a universal ongoing requirement.

References

1. European Medicines Agency. Guideline on good pharmacovigilance practices (GVP) Module IX — Signal management (Rev. 1). EMA/827661/2011 Rev. 1.
2. European Medicines Agency. Questions and answers on signal management. EMA/261758/2013 Rev. 5, updated 20 January 2026.
3. European Medicines Agency. Signal management. Current EMA procedural page, including conclusion of the MAH EudraVigilance signal-detection pilot.
4. European Medicines Agency. Guideline on good pharmacovigilance practices (GVP) Module III — Pharmacovigilance inspections. EMA/119871/2012 Rev. 1.
5. European Union. Commission Implementing Regulation (EU) No 520/2012, as amended by Commission Implementing Regulation (EU) 2025/1466.
6. European Union. Directive 2001/83/EC, as amended.
7. European Union. Regulation (EC) No 726/2004, as amended.

Regulatory Note

This article distinguishes regulatory inspection expectations from recommended internal preparation practices. As of 8 September 2026, GVP Module IX Rev. 1 remains published, but EMA has stated that it will be revised following the amended signal-management framework introduced by Commission Implementing Regulation (EU) 2025/1466. Current EMA signal-management Q&A and legal text should be checked for live inspections.